1 <?xml version="1.0" encoding="utf-8"?> 2 <!-- Copyright (C) 2012 The Android Open Source Project 3 4 Licensed under the Apache License, Version 2.0 (the "License"); 5 you may not use this file except in compliance with the License. 6 You may obtain a copy of the License at 7 8 http://www.apache.org/licenses/LICENSE-2.0 9 10 Unless required by applicable law or agreed to in writing, software 11 distributed under the License is distributed on an "AS IS" BASIS, 12 WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 13 See the License for the specific language governing permissions and 14 limitations under the License. 15 --> 16 <metadata xmlns="http://schemas.android.com/service/camera/metadata/" 17 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 18 xsi:schemaLocation="http://schemas.android.com/service/camera/metadata/ metadata_properties.xsd"> 19 20 <tags> 21 <tag id="BC"> 22 Needed for backwards compatibility with old Java API 23 </tag> 24 <tag id="V1"> 25 New features for first camera 2 release (API1) 26 </tag> 27 <tag id="RAW"> 28 Needed for useful RAW image processing and DNG file support 29 </tag> 30 <tag id="HAL2"> 31 Entry is only used by camera device HAL 2.x 32 </tag> 33 <tag id="FULL"> 34 Entry is required for full hardware level devices, and optional for other hardware levels 35 </tag> 36 <tag id="DEPTH"> 37 Entry is required for the depth capability. 38 </tag> 39 <tag id="REPROC"> 40 Entry is required for the YUV or PRIVATE reprocessing capability. 41 </tag> 42 <tag id="FUTURE"> 43 Entry is under-specified and is not required for now. This is for book-keeping purpose, 44 do not implement or use it, it may be revised for future. 45 </tag> 46 </tags> 47 48 <types> 49 <typedef name="pairFloatFloat"> 50 <language name="java">android.util.Pair<Float,Float></language> 51 </typedef> 52 <typedef name="pairDoubleDouble"> 53 <language name="java">android.util.Pair<Double,Double></language> 54 </typedef> 55 <typedef name="rectangle"> 56 <language name="java">android.graphics.Rect</language> 57 </typedef> 58 <typedef name="size"> 59 <language name="java">android.util.Size</language> 60 </typedef> 61 <typedef name="string"> 62 <language name="java">String</language> 63 </typedef> 64 <typedef name="boolean"> 65 <language name="java">boolean</language> 66 </typedef> 67 <typedef name="imageFormat"> 68 <language name="java">int</language> 69 </typedef> 70 <typedef name="streamConfigurationMap"> 71 <language name="java">android.hardware.camera2.params.StreamConfigurationMap</language> 72 </typedef> 73 <typedef name="streamConfiguration"> 74 <language name="java">android.hardware.camera2.params.StreamConfiguration</language> 75 </typedef> 76 <typedef name="streamConfigurationDuration"> 77 <language name="java">android.hardware.camera2.params.StreamConfigurationDuration</language> 78 </typedef> 79 <typedef name="face"> 80 <language name="java">android.hardware.camera2.params.Face</language> 81 </typedef> 82 <typedef name="meteringRectangle"> 83 <language name="java">android.hardware.camera2.params.MeteringRectangle</language> 84 </typedef> 85 <typedef name="rangeFloat"> 86 <language name="java">android.util.Range<Float></language> 87 </typedef> 88 <typedef name="rangeInt"> 89 <language name="java">android.util.Range<Integer></language> 90 </typedef> 91 <typedef name="rangeLong"> 92 <language name="java">android.util.Range<Long></language> 93 </typedef> 94 <typedef name="colorSpaceTransform"> 95 <language name="java">android.hardware.camera2.params.ColorSpaceTransform</language> 96 </typedef> 97 <typedef name="rggbChannelVector"> 98 <language name="java">android.hardware.camera2.params.RggbChannelVector</language> 99 </typedef> 100 <typedef name="blackLevelPattern"> 101 <language name="java">android.hardware.camera2.params.BlackLevelPattern</language> 102 </typedef> 103 <typedef name="enumList"> 104 <language name="java">int</language> 105 </typedef> 106 <typedef name="sizeF"> 107 <language name="java">android.util.SizeF</language> 108 </typedef> 109 <typedef name="point"> 110 <language name="java">android.graphics.Point</language> 111 </typedef> 112 <typedef name="tonemapCurve"> 113 <language name="java">android.hardware.camera2.params.TonemapCurve</language> 114 </typedef> 115 <typedef name="lensShadingMap"> 116 <language name="java">android.hardware.camera2.params.LensShadingMap</language> 117 </typedef> 118 <typedef name="location"> 119 <language name="java">android.location.Location</language> 120 </typedef> 121 <typedef name="highSpeedVideoConfiguration"> 122 <language name="java">android.hardware.camera2.params.HighSpeedVideoConfiguration</language> 123 </typedef> 124 <typedef name="reprocessFormatsMap"> 125 <language name="java">android.hardware.camera2.params.ReprocessFormatsMap</language> 126 </typedef> 127 </types> 128 129 <namespace name="android"> 130 <section name="colorCorrection"> 131 <controls> 132 <entry name="mode" type="byte" visibility="public" enum="true" hwlevel="full"> 133 <enum> 134 <value>TRANSFORM_MATRIX 135 <notes>Use the android.colorCorrection.transform matrix 136 and android.colorCorrection.gains to do color conversion. 137 138 All advanced white balance adjustments (not specified 139 by our white balance pipeline) must be disabled. 140 141 If AWB is enabled with `android.control.awbMode != OFF`, then 142 TRANSFORM_MATRIX is ignored. The camera device will override 143 this value to either FAST or HIGH_QUALITY. 144 </notes> 145 </value> 146 <value>FAST 147 <notes>Color correction processing must not slow down 148 capture rate relative to sensor raw output. 149 150 Advanced white balance adjustments above and beyond 151 the specified white balance pipeline may be applied. 152 153 If AWB is enabled with `android.control.awbMode != OFF`, then 154 the camera device uses the last frame's AWB values 155 (or defaults if AWB has never been run). 156 </notes> 157 </value> 158 <value>HIGH_QUALITY 159 <notes>Color correction processing operates at improved 160 quality but the capture rate might be reduced (relative to sensor 161 raw output rate) 162 163 Advanced white balance adjustments above and beyond 164 the specified white balance pipeline may be applied. 165 166 If AWB is enabled with `android.control.awbMode != OFF`, then 167 the camera device uses the last frame's AWB values 168 (or defaults if AWB has never been run). 169 </notes> 170 </value> 171 </enum> 172 173 <description> 174 The mode control selects how the image data is converted from the 175 sensor's native color into linear sRGB color. 176 </description> 177 <details> 178 When auto-white balance (AWB) is enabled with android.control.awbMode, this 179 control is overridden by the AWB routine. When AWB is disabled, the 180 application controls how the color mapping is performed. 181 182 We define the expected processing pipeline below. For consistency 183 across devices, this is always the case with TRANSFORM_MATRIX. 184 185 When either FULL or HIGH_QUALITY is used, the camera device may 186 do additional processing but android.colorCorrection.gains and 187 android.colorCorrection.transform will still be provided by the 188 camera device (in the results) and be roughly correct. 189 190 Switching to TRANSFORM_MATRIX and using the data provided from 191 FAST or HIGH_QUALITY will yield a picture with the same white point 192 as what was produced by the camera device in the earlier frame. 193 194 The expected processing pipeline is as follows: 195 196  197 198 The white balance is encoded by two values, a 4-channel white-balance 199 gain vector (applied in the Bayer domain), and a 3x3 color transform 200 matrix (applied after demosaic). 201 202 The 4-channel white-balance gains are defined as: 203 204 android.colorCorrection.gains = [ R G_even G_odd B ] 205 206 where `G_even` is the gain for green pixels on even rows of the 207 output, and `G_odd` is the gain for green pixels on the odd rows. 208 These may be identical for a given camera device implementation; if 209 the camera device does not support a separate gain for even/odd green 210 channels, it will use the `G_even` value, and write `G_odd` equal to 211 `G_even` in the output result metadata. 212 213 The matrices for color transforms are defined as a 9-entry vector: 214 215 android.colorCorrection.transform = [ I0 I1 I2 I3 I4 I5 I6 I7 I8 ] 216 217 which define a transform from input sensor colors, `P_in = [ r g b ]`, 218 to output linear sRGB, `P_out = [ r' g' b' ]`, 219 220 with colors as follows: 221 222 r' = I0r + I1g + I2b 223 g' = I3r + I4g + I5b 224 b' = I6r + I7g + I8b 225 226 Both the input and output value ranges must match. Overflow/underflow 227 values are clipped to fit within the range. 228 </details> 229 <hal_details> 230 HAL must support both FAST and HIGH_QUALITY if color correction control is available 231 on the camera device, but the underlying implementation can be the same for both modes. 232 That is, if the highest quality implementation on the camera device does not slow down 233 capture rate, then FAST and HIGH_QUALITY should generate the same output. 234 </hal_details> 235 </entry> 236 <entry name="transform" type="rational" visibility="public" 237 type_notes="3x3 rational matrix in row-major order" 238 container="array" typedef="colorSpaceTransform" hwlevel="full"> 239 <array> 240 <size>3</size> 241 <size>3</size> 242 </array> 243 <description>A color transform matrix to use to transform 244 from sensor RGB color space to output linear sRGB color space. 245 </description> 246 <units>Unitless scale factors</units> 247 <details>This matrix is either set by the camera device when the request 248 android.colorCorrection.mode is not TRANSFORM_MATRIX, or 249 directly by the application in the request when the 250 android.colorCorrection.mode is TRANSFORM_MATRIX. 251 252 In the latter case, the camera device may round the matrix to account 253 for precision issues; the final rounded matrix should be reported back 254 in this matrix result metadata. The transform should keep the magnitude 255 of the output color values within `[0, 1.0]` (assuming input color 256 values is within the normalized range `[0, 1.0]`), or clipping may occur. 257 258 The valid range of each matrix element varies on different devices, but 259 values within [-1.5, 3.0] are guaranteed not to be clipped. 260 </details> 261 </entry> 262 <entry name="gains" type="float" visibility="public" 263 type_notes="A 1D array of floats for 4 color channel gains" 264 container="array" typedef="rggbChannelVector" hwlevel="full"> 265 <array> 266 <size>4</size> 267 </array> 268 <description>Gains applying to Bayer raw color channels for 269 white-balance.</description> 270 <units>Unitless gain factors</units> 271 <details> 272 These per-channel gains are either set by the camera device 273 when the request android.colorCorrection.mode is not 274 TRANSFORM_MATRIX, or directly by the application in the 275 request when the android.colorCorrection.mode is 276 TRANSFORM_MATRIX. 277 278 The gains in the result metadata are the gains actually 279 applied by the camera device to the current frame. 280 281 The valid range of gains varies on different devices, but gains 282 between [1.0, 3.0] are guaranteed not to be clipped. Even if a given 283 device allows gains below 1.0, this is usually not recommended because 284 this can create color artifacts. 285 </details> 286 <hal_details> 287 The 4-channel white-balance gains are defined in 288 the order of `[R G_even G_odd B]`, where `G_even` is the gain 289 for green pixels on even rows of the output, and `G_odd` 290 is the gain for green pixels on the odd rows. 291 292 If a HAL does not support a separate gain for even/odd green 293 channels, it must use the `G_even` value, and write 294 `G_odd` equal to `G_even` in the output result metadata. 295 </hal_details> 296 </entry> 297 <entry name="aberrationMode" type="byte" visibility="public" enum="true" hwlevel="legacy"> 298 <enum> 299 <value>OFF 300 <notes> 301 No aberration correction is applied. 302 </notes> 303 </value> 304 <value>FAST 305 <notes> 306 Aberration correction will not slow down capture rate 307 relative to sensor raw output. 308 </notes> 309 </value> 310 <value>HIGH_QUALITY 311 <notes> 312 Aberration correction operates at improved quality but the capture rate might be 313 reduced (relative to sensor raw output rate) 314 </notes> 315 </value> 316 </enum> 317 <description> 318 Mode of operation for the chromatic aberration correction algorithm. 319 </description> 320 <range>android.colorCorrection.availableAberrationModes</range> 321 <details> 322 Chromatic (color) aberration is caused by the fact that different wavelengths of light 323 can not focus on the same point after exiting from the lens. This metadata defines 324 the high level control of chromatic aberration correction algorithm, which aims to 325 minimize the chromatic artifacts that may occur along the object boundaries in an 326 image. 327 328 FAST/HIGH_QUALITY both mean that camera device determined aberration 329 correction will be applied. HIGH_QUALITY mode indicates that the camera device will 330 use the highest-quality aberration correction algorithms, even if it slows down 331 capture rate. FAST means the camera device will not slow down capture rate when 332 applying aberration correction. 333 334 LEGACY devices will always be in FAST mode. 335 </details> 336 </entry> 337 </controls> 338 <dynamic> 339 <clone entry="android.colorCorrection.mode" kind="controls"> 340 </clone> 341 <clone entry="android.colorCorrection.transform" kind="controls"> 342 </clone> 343 <clone entry="android.colorCorrection.gains" kind="controls"> 344 </clone> 345 <clone entry="android.colorCorrection.aberrationMode" kind="controls"> 346 </clone> 347 </dynamic> 348 <static> 349 <entry name="availableAberrationModes" type="byte" visibility="public" 350 type_notes="list of enums" container="array" typedef="enumList" hwlevel="legacy"> 351 <array> 352 <size>n</size> 353 </array> 354 <description> 355 List of aberration correction modes for android.colorCorrection.aberrationMode that are 356 supported by this camera device. 357 </description> 358 <range>Any value listed in android.colorCorrection.aberrationMode</range> 359 <details> 360 This key lists the valid modes for android.colorCorrection.aberrationMode. If no 361 aberration correction modes are available for a device, this list will solely include 362 OFF mode. All camera devices will support either OFF or FAST mode. 363 364 Camera devices that support the MANUAL_POST_PROCESSING capability will always list 365 OFF mode. This includes all FULL level devices. 366 367 LEGACY devices will always only support FAST mode. 368 </details> 369 <hal_details> 370 HAL must support both FAST and HIGH_QUALITY if chromatic aberration control is available 371 on the camera device, but the underlying implementation can be the same for both modes. 372 That is, if the highest quality implementation on the camera device does not slow down 373 capture rate, then FAST and HIGH_QUALITY will generate the same output. 374 </hal_details> 375 <tag id="V1" /> 376 </entry> 377 </static> 378 </section> 379 <section name="control"> 380 <controls> 381 <entry name="aeAntibandingMode" type="byte" visibility="public" 382 enum="true" hwlevel="legacy"> 383 <enum> 384 <value>OFF 385 <notes> 386 The camera device will not adjust exposure duration to 387 avoid banding problems. 388 </notes> 389 </value> 390 <value>50HZ 391 <notes> 392 The camera device will adjust exposure duration to 393 avoid banding problems with 50Hz illumination sources. 394 </notes> 395 </value> 396 <value>60HZ 397 <notes> 398 The camera device will adjust exposure duration to 399 avoid banding problems with 60Hz illumination 400 sources. 401 </notes> 402 </value> 403 <value>AUTO 404 <notes> 405 The camera device will automatically adapt its 406 antibanding routine to the current illumination 407 condition. This is the default mode if AUTO is 408 available on given camera device. 409 </notes> 410 </value> 411 </enum> 412 <description> 413 The desired setting for the camera device's auto-exposure 414 algorithm's antibanding compensation. 415 </description> 416 <range> 417 android.control.aeAvailableAntibandingModes 418 </range> 419 <details> 420 Some kinds of lighting fixtures, such as some fluorescent 421 lights, flicker at the rate of the power supply frequency 422 (60Hz or 50Hz, depending on country). While this is 423 typically not noticeable to a person, it can be visible to 424 a camera device. If a camera sets its exposure time to the 425 wrong value, the flicker may become visible in the 426 viewfinder as flicker or in a final captured image, as a 427 set of variable-brightness bands across the image. 428 429 Therefore, the auto-exposure routines of camera devices 430 include antibanding routines that ensure that the chosen 431 exposure value will not cause such banding. The choice of 432 exposure time depends on the rate of flicker, which the 433 camera device can detect automatically, or the expected 434 rate can be selected by the application using this 435 control. 436 437 A given camera device may not support all of the possible 438 options for the antibanding mode. The 439 android.control.aeAvailableAntibandingModes key contains 440 the available modes for a given camera device. 441 442 AUTO mode is the default if it is available on given 443 camera device. When AUTO mode is not available, the 444 default will be either 50HZ or 60HZ, and both 50HZ 445 and 60HZ will be available. 446 447 If manual exposure control is enabled (by setting 448 android.control.aeMode or android.control.mode to OFF), 449 then this setting has no effect, and the application must 450 ensure it selects exposure times that do not cause banding 451 issues. The android.statistics.sceneFlicker key can assist 452 the application in this. 453 </details> 454 <hal_details> 455 For all capture request templates, this field must be set 456 to AUTO if AUTO mode is available. If AUTO is not available, 457 the default must be either 50HZ or 60HZ, and both 50HZ and 458 60HZ must be available. 459 460 If manual exposure control is enabled (by setting 461 android.control.aeMode or android.control.mode to OFF), 462 then the exposure values provided by the application must not be 463 adjusted for antibanding. 464 </hal_details> 465 <tag id="BC" /> 466 </entry> 467 <entry name="aeExposureCompensation" type="int32" visibility="public" hwlevel="legacy"> 468 <description>Adjustment to auto-exposure (AE) target image 469 brightness.</description> 470 <units>Compensation steps</units> 471 <range>android.control.aeCompensationRange</range> 472 <details> 473 The adjustment is measured as a count of steps, with the 474 step size defined by android.control.aeCompensationStep and the 475 allowed range by android.control.aeCompensationRange. 476 477 For example, if the exposure value (EV) step is 0.333, '6' 478 will mean an exposure compensation of +2 EV; -3 will mean an 479 exposure compensation of -1 EV. One EV represents a doubling 480 of image brightness. Note that this control will only be 481 effective if android.control.aeMode `!=` OFF. This control 482 will take effect even when android.control.aeLock `== true`. 483 484 In the event of exposure compensation value being changed, camera device 485 may take several frames to reach the newly requested exposure target. 486 During that time, android.control.aeState field will be in the SEARCHING 487 state. Once the new exposure target is reached, android.control.aeState will 488 change from SEARCHING to either CONVERGED, LOCKED (if AE lock is enabled), or 489 FLASH_REQUIRED (if the scene is too dark for still capture). 490 </details> 491 <tag id="BC" /> 492 </entry> 493 <entry name="aeLock" type="byte" visibility="public" enum="true" 494 typedef="boolean" hwlevel="legacy"> 495 <enum> 496 <value>OFF 497 <notes>Auto-exposure lock is disabled; the AE algorithm 498 is free to update its parameters.</notes></value> 499 <value>ON 500 <notes>Auto-exposure lock is enabled; the AE algorithm 501 must not update the exposure and sensitivity parameters 502 while the lock is active. 503 504 android.control.aeExposureCompensation setting changes 505 will still take effect while auto-exposure is locked. 506 507 Some rare LEGACY devices may not support 508 this, in which case the value will always be overridden to OFF. 509 </notes></value> 510 </enum> 511 <description>Whether auto-exposure (AE) is currently locked to its latest 512 calculated values.</description> 513 <details> 514 When set to `true` (ON), the AE algorithm is locked to its latest parameters, 515 and will not change exposure settings until the lock is set to `false` (OFF). 516 517 Note that even when AE is locked, the flash may be fired if 518 the android.control.aeMode is ON_AUTO_FLASH / 519 ON_ALWAYS_FLASH / ON_AUTO_FLASH_REDEYE. 520 521 When android.control.aeExposureCompensation is changed, even if the AE lock 522 is ON, the camera device will still adjust its exposure value. 523 524 If AE precapture is triggered (see android.control.aePrecaptureTrigger) 525 when AE is already locked, the camera device will not change the exposure time 526 (android.sensor.exposureTime) and sensitivity (android.sensor.sensitivity) 527 parameters. The flash may be fired if the android.control.aeMode 528 is ON_AUTO_FLASH/ON_AUTO_FLASH_REDEYE and the scene is too dark. If the 529 android.control.aeMode is ON_ALWAYS_FLASH, the scene may become overexposed. 530 Similarly, AE precapture trigger CANCEL has no effect when AE is already locked. 531 532 When an AE precapture sequence is triggered, AE unlock will not be able to unlock 533 the AE if AE is locked by the camera device internally during precapture metering 534 sequence In other words, submitting requests with AE unlock has no effect for an 535 ongoing precapture metering sequence. Otherwise, the precapture metering sequence 536 will never succeed in a sequence of preview requests where AE lock is always set 537 to `false`. 538 539 Since the camera device has a pipeline of in-flight requests, the settings that 540 get locked do not necessarily correspond to the settings that were present in the 541 latest capture result received from the camera device, since additional captures 542 and AE updates may have occurred even before the result was sent out. If an 543 application is switching between automatic and manual control and wishes to eliminate 544 any flicker during the switch, the following procedure is recommended: 545 546 1. Starting in auto-AE mode: 547 2. Lock AE 548 3. Wait for the first result to be output that has the AE locked 549 4. Copy exposure settings from that result into a request, set the request to manual AE 550 5. Submit the capture request, proceed to run manual AE as desired. 551 552 See android.control.aeState for AE lock related state transition details. 553 </details> 554 <tag id="BC" /> 555 </entry> 556 <entry name="aeMode" type="byte" visibility="public" enum="true" hwlevel="legacy"> 557 <enum> 558 <value>OFF 559 <notes> 560 The camera device's autoexposure routine is disabled. 561 562 The application-selected android.sensor.exposureTime, 563 android.sensor.sensitivity and 564 android.sensor.frameDuration are used by the camera 565 device, along with android.flash.* fields, if there's 566 a flash unit for this camera device. 567 568 Note that auto-white balance (AWB) and auto-focus (AF) 569 behavior is device dependent when AE is in OFF mode. 570 To have consistent behavior across different devices, 571 it is recommended to either set AWB and AF to OFF mode 572 or lock AWB and AF before setting AE to OFF. 573 See android.control.awbMode, android.control.afMode, 574 android.control.awbLock, and android.control.afTrigger 575 for more details. 576 577 LEGACY devices do not support the OFF mode and will 578 override attempts to use this value to ON. 579 </notes> 580 </value> 581 <value>ON 582 <notes> 583 The camera device's autoexposure routine is active, 584 with no flash control. 585 586 The application's values for 587 android.sensor.exposureTime, 588 android.sensor.sensitivity, and 589 android.sensor.frameDuration are ignored. The 590 application has control over the various 591 android.flash.* fields. 592 </notes> 593 </value> 594 <value>ON_AUTO_FLASH 595 <notes> 596 Like ON, except that the camera device also controls 597 the camera's flash unit, firing it in low-light 598 conditions. 599 600 The flash may be fired during a precapture sequence 601 (triggered by android.control.aePrecaptureTrigger) and 602 may be fired for captures for which the 603 android.control.captureIntent field is set to 604 STILL_CAPTURE 605 </notes> 606 </value> 607 <value>ON_ALWAYS_FLASH 608 <notes> 609 Like ON, except that the camera device also controls 610 the camera's flash unit, always firing it for still 611 captures. 612 613 The flash may be fired during a precapture sequence 614 (triggered by android.control.aePrecaptureTrigger) and 615 will always be fired for captures for which the 616 android.control.captureIntent field is set to 617 STILL_CAPTURE 618 </notes> 619 </value> 620 <value>ON_AUTO_FLASH_REDEYE 621 <notes> 622 Like ON_AUTO_FLASH, but with automatic red eye 623 reduction. 624 625 If deemed necessary by the camera device, a red eye 626 reduction flash will fire during the precapture 627 sequence. 628 </notes> 629 </value> 630 </enum> 631 <description>The desired mode for the camera device's 632 auto-exposure routine.</description> 633 <range>android.control.aeAvailableModes</range> 634 <details> 635 This control is only effective if android.control.mode is 636 AUTO. 637 638 When set to any of the ON modes, the camera device's 639 auto-exposure routine is enabled, overriding the 640 application's selected exposure time, sensor sensitivity, 641 and frame duration (android.sensor.exposureTime, 642 android.sensor.sensitivity, and 643 android.sensor.frameDuration). If one of the FLASH modes 644 is selected, the camera device's flash unit controls are 645 also overridden. 646 647 The FLASH modes are only available if the camera device 648 has a flash unit (android.flash.info.available is `true`). 649 650 If flash TORCH mode is desired, this field must be set to 651 ON or OFF, and android.flash.mode set to TORCH. 652 653 When set to any of the ON modes, the values chosen by the 654 camera device auto-exposure routine for the overridden 655 fields for a given capture will be available in its 656 CaptureResult. 657 </details> 658 <tag id="BC" /> 659 </entry> 660 <entry name="aeRegions" type="int32" visibility="public" 661 optional="true" container="array" typedef="meteringRectangle"> 662 <array> 663 <size>5</size> 664 <size>area_count</size> 665 </array> 666 <description>List of metering areas to use for auto-exposure adjustment.</description> 667 <units>Pixel coordinates within android.sensor.info.activeArraySize</units> 668 <range>Coordinates must be between `[(0,0), (width, height))` of 669 android.sensor.info.activeArraySize</range> 670 <details> 671 Not available if android.control.maxRegionsAe is 0. 672 Otherwise will always be present. 673 674 The maximum number of regions supported by the device is determined by the value 675 of android.control.maxRegionsAe. 676 677 The coordinate system is based on the active pixel array, 678 with (0,0) being the top-left pixel in the active pixel array, and 679 (android.sensor.info.activeArraySize.width - 1, 680 android.sensor.info.activeArraySize.height - 1) being the 681 bottom-right pixel in the active pixel array. 682 683 The weight must be within `[0, 1000]`, and represents a weight 684 for every pixel in the area. This means that a large metering area 685 with the same weight as a smaller area will have more effect in 686 the metering result. Metering areas can partially overlap and the 687 camera device will add the weights in the overlap region. 688 689 The weights are relative to weights of other exposure metering regions, so if only one 690 region is used, all non-zero weights will have the same effect. A region with 0 691 weight is ignored. 692 693 If all regions have 0 weight, then no specific metering area needs to be used by the 694 camera device. 695 696 If the metering region is outside the used android.scaler.cropRegion returned in 697 capture result metadata, the camera device will ignore the sections outside the crop 698 region and output only the intersection rectangle as the metering region in the result 699 metadata. If the region is entirely outside the crop region, it will be ignored and 700 not reported in the result metadata. 701 </details> 702 <hal_details> 703 The HAL level representation of MeteringRectangle[] is a 704 int[5 * area_count]. 705 Every five elements represent a metering region of 706 (xmin, ymin, xmax, ymax, weight). 707 The rectangle is defined to be inclusive on xmin and ymin, but 708 exclusive on xmax and ymax. 709 </hal_details> 710 <tag id="BC" /> 711 </entry> 712 <entry name="aeTargetFpsRange" type="int32" visibility="public" 713 container="array" typedef="rangeInt" hwlevel="legacy"> 714 <array> 715 <size>2</size> 716 </array> 717 <description>Range over which the auto-exposure routine can 718 adjust the capture frame rate to maintain good 719 exposure.</description> 720 <units>Frames per second (FPS)</units> 721 <range>Any of the entries in android.control.aeAvailableTargetFpsRanges</range> 722 <details>Only constrains auto-exposure (AE) algorithm, not 723 manual control of android.sensor.exposureTime and 724 android.sensor.frameDuration.</details> 725 <tag id="BC" /> 726 </entry> 727 <entry name="aePrecaptureTrigger" type="byte" visibility="public" 728 enum="true" hwlevel="limited"> 729 <enum> 730 <value>IDLE 731 <notes>The trigger is idle.</notes> 732 </value> 733 <value>START 734 <notes>The precapture metering sequence will be started 735 by the camera device. 736 737 The exact effect of the precapture trigger depends on 738 the current AE mode and state.</notes> 739 </value> 740 <value>CANCEL 741 <notes>The camera device will cancel any currently active or completed 742 precapture metering sequence, the auto-exposure routine will return to its 743 initial state.</notes> 744 </value> 745 </enum> 746 <description>Whether the camera device will trigger a precapture 747 metering sequence when it processes this request.</description> 748 <details>This entry is normally set to IDLE, or is not 749 included at all in the request settings. When included and 750 set to START, the camera device will trigger the auto-exposure (AE) 751 precapture metering sequence. 752 753 When set to CANCEL, the camera device will cancel any active 754 precapture metering trigger, and return to its initial AE state. 755 If a precapture metering sequence is already completed, and the camera 756 device has implicitly locked the AE for subsequent still capture, the 757 CANCEL trigger will unlock the AE and return to its initial AE state. 758 759 The precapture sequence should be triggered before starting a 760 high-quality still capture for final metering decisions to 761 be made, and for firing pre-capture flash pulses to estimate 762 scene brightness and required final capture flash power, when 763 the flash is enabled. 764 765 Normally, this entry should be set to START for only a 766 single request, and the application should wait until the 767 sequence completes before starting a new one. 768 769 When a precapture metering sequence is finished, the camera device 770 may lock the auto-exposure routine internally to be able to accurately expose the 771 subsequent still capture image (`android.control.captureIntent == STILL_CAPTURE`). 772 For this case, the AE may not resume normal scan if no subsequent still capture is 773 submitted. To ensure that the AE routine restarts normal scan, the application should 774 submit a request with `android.control.aeLock == true`, followed by a request 775 with `android.control.aeLock == false`, if the application decides not to submit a 776 still capture request after the precapture sequence completes. Alternatively, for 777 API level 23 or newer devices, the CANCEL can be used to unlock the camera device 778 internally locked AE if the application doesn't submit a still capture request after 779 the AE precapture trigger. Note that, the CANCEL was added in API level 23, and must not 780 be used in devices that have earlier API levels. 781 782 The exact effect of auto-exposure (AE) precapture trigger 783 depends on the current AE mode and state; see 784 android.control.aeState for AE precapture state transition 785 details. 786 787 On LEGACY-level devices, the precapture trigger is not supported; 788 capturing a high-resolution JPEG image will automatically trigger a 789 precapture sequence before the high-resolution capture, including 790 potentially firing a pre-capture flash. 791 792 Using the precapture trigger and the auto-focus trigger android.control.afTrigger 793 simultaneously is allowed. However, since these triggers often require cooperation between 794 the auto-focus and auto-exposure routines (for example, the may need to be enabled for a 795 focus sweep), the camera device may delay acting on a later trigger until the previous 796 trigger has been fully handled. This may lead to longer intervals between the trigger and 797 changes to android.control.aeState indicating the start of the precapture sequence, for 798 example. 799 800 If both the precapture and the auto-focus trigger are activated on the same request, then 801 the camera device will complete them in the optimal order for that device. 802 </details> 803 <hal_details> 804 The HAL must support triggering the AE precapture trigger while an AF trigger is active 805 (and vice versa), or at the same time as the AF trigger. It is acceptable for the HAL to 806 treat these as two consecutive triggers, for example handling the AF trigger and then the 807 AE trigger. Or the HAL may choose to optimize the case with both triggers fired at once, 808 to minimize the latency for converging both focus and exposure/flash usage. 809 </hal_details> 810 <tag id="BC" /> 811 </entry> 812 <entry name="afMode" type="byte" visibility="public" enum="true" 813 hwlevel="legacy"> 814 <enum> 815 <value>OFF 816 <notes>The auto-focus routine does not control the lens; 817 android.lens.focusDistance is controlled by the 818 application.</notes></value> 819 <value>AUTO 820 <notes>Basic automatic focus mode. 821 822 In this mode, the lens does not move unless 823 the autofocus trigger action is called. When that trigger 824 is activated, AF will transition to ACTIVE_SCAN, then to 825 the outcome of the scan (FOCUSED or NOT_FOCUSED). 826 827 Always supported if lens is not fixed focus. 828 829 Use android.lens.info.minimumFocusDistance to determine if lens 830 is fixed-focus. 831 832 Triggering AF_CANCEL resets the lens position to default, 833 and sets the AF state to INACTIVE.</notes></value> 834 <value>MACRO 835 <notes>Close-up focusing mode. 836 837 In this mode, the lens does not move unless the 838 autofocus trigger action is called. When that trigger is 839 activated, AF will transition to ACTIVE_SCAN, then to 840 the outcome of the scan (FOCUSED or NOT_FOCUSED). This 841 mode is optimized for focusing on objects very close to 842 the camera. 843 844 When that trigger is activated, AF will transition to 845 ACTIVE_SCAN, then to the outcome of the scan (FOCUSED or 846 NOT_FOCUSED). Triggering cancel AF resets the lens 847 position to default, and sets the AF state to 848 INACTIVE.</notes></value> 849 <value>CONTINUOUS_VIDEO 850 <notes>In this mode, the AF algorithm modifies the lens 851 position continually to attempt to provide a 852 constantly-in-focus image stream. 853 854 The focusing behavior should be suitable for good quality 855 video recording; typically this means slower focus 856 movement and no overshoots. When the AF trigger is not 857 involved, the AF algorithm should start in INACTIVE state, 858 and then transition into PASSIVE_SCAN and PASSIVE_FOCUSED 859 states as appropriate. When the AF trigger is activated, 860 the algorithm should immediately transition into 861 AF_FOCUSED or AF_NOT_FOCUSED as appropriate, and lock the 862 lens position until a cancel AF trigger is received. 863 864 Once cancel is received, the algorithm should transition 865 back to INACTIVE and resume passive scan. Note that this 866 behavior is not identical to CONTINUOUS_PICTURE, since an 867 ongoing PASSIVE_SCAN must immediately be 868 canceled.</notes></value> 869 <value>CONTINUOUS_PICTURE 870 <notes>In this mode, the AF algorithm modifies the lens 871 position continually to attempt to provide a 872 constantly-in-focus image stream. 873 874 The focusing behavior should be suitable for still image 875 capture; typically this means focusing as fast as 876 possible. When the AF trigger is not involved, the AF 877 algorithm should start in INACTIVE state, and then 878 transition into PASSIVE_SCAN and PASSIVE_FOCUSED states as 879 appropriate as it attempts to maintain focus. When the AF 880 trigger is activated, the algorithm should finish its 881 PASSIVE_SCAN if active, and then transition into 882 AF_FOCUSED or AF_NOT_FOCUSED as appropriate, and lock the 883 lens position until a cancel AF trigger is received. 884 885 When the AF cancel trigger is activated, the algorithm 886 should transition back to INACTIVE and then act as if it 887 has just been started.</notes></value> 888 <value>EDOF 889 <notes>Extended depth of field (digital focus) mode. 890 891 The camera device will produce images with an extended 892 depth of field automatically; no special focusing 893 operations need to be done before taking a picture. 894 895 AF triggers are ignored, and the AF state will always be 896 INACTIVE.</notes></value> 897 </enum> 898 <description>Whether auto-focus (AF) is currently enabled, and what 899 mode it is set to.</description> 900 <range>android.control.afAvailableModes</range> 901 <details>Only effective if android.control.mode = AUTO and the lens is not fixed focus 902 (i.e. `android.lens.info.minimumFocusDistance > 0`). Also note that 903 when android.control.aeMode is OFF, the behavior of AF is device 904 dependent. It is recommended to lock AF by using android.control.afTrigger before 905 setting android.control.aeMode to OFF, or set AF mode to OFF when AE is OFF. 906 907 If the lens is controlled by the camera device auto-focus algorithm, 908 the camera device will report the current AF status in android.control.afState 909 in result metadata.</details> 910 <hal_details> 911 When afMode is AUTO or MACRO, the lens must not move until an AF trigger is sent in a 912 request (android.control.afTrigger `==` START). After an AF trigger, the afState will end 913 up with either FOCUSED_LOCKED or NOT_FOCUSED_LOCKED state (see 914 android.control.afState for detailed state transitions), which indicates that the lens is 915 locked and will not move. If camera movement (e.g. tilting camera) causes the lens to move 916 after the lens is locked, the HAL must compensate this movement appropriately such that 917 the same focal plane remains in focus. 918 919 When afMode is one of the continuous auto focus modes, the HAL is free to start a AF 920 scan whenever it's not locked. When the lens is locked after an AF trigger 921 (see android.control.afState for detailed state transitions), the HAL should maintain the 922 same lock behavior as above. 923 924 When afMode is OFF, the application controls focus manually. The accuracy of the 925 focus distance control depends on the android.lens.info.focusDistanceCalibration. 926 However, the lens must not move regardless of the camera movement for any focus distance 927 manual control. 928 929 To put this in concrete terms, if the camera has lens elements which may move based on 930 camera orientation or motion (e.g. due to gravity), then the HAL must drive the lens to 931 remain in a fixed position invariant to the camera's orientation or motion, for example, 932 by using accelerometer measurements in the lens control logic. This is a typical issue 933 that will arise on camera modules with open-loop VCMs. 934 </hal_details> 935 <tag id="BC" /> 936 </entry> 937 <entry name="afRegions" type="int32" visibility="public" 938 optional="true" container="array" typedef="meteringRectangle"> 939 <array> 940 <size>5</size> 941 <size>area_count</size> 942 </array> 943 <description>List of metering areas to use for auto-focus.</description> 944 <units>Pixel coordinates within android.sensor.info.activeArraySize</units> 945 <range>Coordinates must be between `[(0,0), (width, height))` of 946 android.sensor.info.activeArraySize</range> 947 <details> 948 Not available if android.control.maxRegionsAf is 0. 949 Otherwise will always be present. 950 951 The maximum number of focus areas supported by the device is determined by the value 952 of android.control.maxRegionsAf. 953 954 The coordinate system is based on the active pixel array, 955 with (0,0) being the top-left pixel in the active pixel array, and 956 (android.sensor.info.activeArraySize.width - 1, 957 android.sensor.info.activeArraySize.height - 1) being the 958 bottom-right pixel in the active pixel array. 959 960 The weight must be within `[0, 1000]`, and represents a weight 961 for every pixel in the area. This means that a large metering area 962 with the same weight as a smaller area will have more effect in 963 the metering result. Metering areas can partially overlap and the 964 camera device will add the weights in the overlap region. 965 966 The weights are relative to weights of other metering regions, so if only one region 967 is used, all non-zero weights will have the same effect. A region with 0 weight is 968 ignored. 969 970 If all regions have 0 weight, then no specific metering area needs to be used by the 971 camera device. 972 973 If the metering region is outside the used android.scaler.cropRegion returned in 974 capture result metadata, the camera device will ignore the sections outside the crop 975 region and output only the intersection rectangle as the metering region in the result 976 metadata. If the region is entirely outside the crop region, it will be ignored and 977 not reported in the result metadata. 978 </details> 979 <hal_details> 980 The HAL level representation of MeteringRectangle[] is a 981 int[5 * area_count]. 982 Every five elements represent a metering region of 983 (xmin, ymin, xmax, ymax, weight). 984 The rectangle is defined to be inclusive on xmin and ymin, but 985 exclusive on xmax and ymax. 986 </hal_details> 987 <tag id="BC" /> 988 </entry> 989 <entry name="afTrigger" type="byte" visibility="public" enum="true" 990 hwlevel="legacy"> 991 <enum> 992 <value>IDLE 993 <notes>The trigger is idle.</notes> 994 </value> 995 <value>START 996 <notes>Autofocus will trigger now.</notes> 997 </value> 998 <value>CANCEL 999 <notes>Autofocus will return to its initial 1000 state, and cancel any currently active trigger.</notes> 1001 </value> 1002 </enum> 1003 <description> 1004 Whether the camera device will trigger autofocus for this request. 1005 </description> 1006 <details>This entry is normally set to IDLE, or is not 1007 included at all in the request settings. 1008 1009 When included and set to START, the camera device will trigger the 1010 autofocus algorithm. If autofocus is disabled, this trigger has no effect. 1011 1012 When set to CANCEL, the camera device will cancel any active trigger, 1013 and return to its initial AF state. 1014 1015 Generally, applications should set this entry to START or CANCEL for only a 1016 single capture, and then return it to IDLE (or not set at all). Specifying 1017 START for multiple captures in a row means restarting the AF operation over 1018 and over again. 1019 1020 See android.control.afState for what the trigger means for each AF mode. 1021 1022 Using the autofocus trigger and the precapture trigger android.control.aePrecaptureTrigger 1023 simultaneously is allowed. However, since these triggers often require cooperation between 1024 the auto-focus and auto-exposure routines (for example, the may need to be enabled for a 1025 focus sweep), the camera device may delay acting on a later trigger until the previous 1026 trigger has been fully handled. This may lead to longer intervals between the trigger and 1027 changes to android.control.afState, for example. 1028 </details> 1029 <hal_details> 1030 The HAL must support triggering the AF trigger while an AE precapture trigger is active 1031 (and vice versa), or at the same time as the AE trigger. It is acceptable for the HAL to 1032 treat these as two consecutive triggers, for example handling the AF trigger and then the 1033 AE trigger. Or the HAL may choose to optimize the case with both triggers fired at once, 1034 to minimize the latency for converging both focus and exposure/flash usage. 1035 </hal_details> 1036 <tag id="BC" /> 1037 </entry> 1038 <entry name="awbLock" type="byte" visibility="public" enum="true" 1039 typedef="boolean" hwlevel="legacy"> 1040 <enum> 1041 <value>OFF 1042 <notes>Auto-white balance lock is disabled; the AWB 1043 algorithm is free to update its parameters if in AUTO 1044 mode.</notes></value> 1045 <value>ON 1046 <notes>Auto-white balance lock is enabled; the AWB 1047 algorithm will not update its parameters while the lock 1048 is active.</notes></value> 1049 </enum> 1050 <description>Whether auto-white balance (AWB) is currently locked to its 1051 latest calculated values.</description> 1052 <details> 1053 When set to `true` (ON), the AWB algorithm is locked to its latest parameters, 1054 and will not change color balance settings until the lock is set to `false` (OFF). 1055 1056 Since the camera device has a pipeline of in-flight requests, the settings that 1057 get locked do not necessarily correspond to the settings that were present in the 1058 latest capture result received from the camera device, since additional captures 1059 and AWB updates may have occurred even before the result was sent out. If an 1060 application is switching between automatic and manual control and wishes to eliminate 1061 any flicker during the switch, the following procedure is recommended: 1062 1063 1. Starting in auto-AWB mode: 1064 2. Lock AWB 1065 3. Wait for the first result to be output that has the AWB locked 1066 4. Copy AWB settings from that result into a request, set the request to manual AWB 1067 5. Submit the capture request, proceed to run manual AWB as desired. 1068 1069 Note that AWB lock is only meaningful when 1070 android.control.awbMode is in the AUTO mode; in other modes, 1071 AWB is already fixed to a specific setting. 1072 1073 Some LEGACY devices may not support ON; the value is then overridden to OFF. 1074 </details> 1075 <tag id="BC" /> 1076 </entry> 1077 <entry name="awbMode" type="byte" visibility="public" enum="true" 1078 hwlevel="legacy"> 1079 <enum> 1080 <value>OFF 1081 <notes> 1082 The camera device's auto-white balance routine is disabled. 1083 1084 The application-selected color transform matrix 1085 (android.colorCorrection.transform) and gains 1086 (android.colorCorrection.gains) are used by the camera 1087 device for manual white balance control. 1088 </notes> 1089 </value> 1090 <value>AUTO 1091 <notes> 1092 The camera device's auto-white balance routine is active. 1093 1094 The application's values for android.colorCorrection.transform 1095 and android.colorCorrection.gains are ignored. 1096 For devices that support the MANUAL_POST_PROCESSING capability, the 1097 values used by the camera device for the transform and gains 1098 will be available in the capture result for this request. 1099 </notes> 1100 </value> 1101 <value>INCANDESCENT 1102 <notes> 1103 The camera device's auto-white balance routine is disabled; 1104 the camera device uses incandescent light as the assumed scene 1105 illumination for white balance. 1106 1107 While the exact white balance transforms are up to the 1108 camera device, they will approximately match the CIE 1109 standard illuminant A. 1110 1111 The application's values for android.colorCorrection.transform 1112 and android.colorCorrection.gains are ignored. 1113 For devices that support the MANUAL_POST_PROCESSING capability, the 1114 values used by the camera device for the transform and gains 1115 will be available in the capture result for this request. 1116 </notes> 1117 </value> 1118 <value>FLUORESCENT 1119 <notes> 1120 The camera device's auto-white balance routine is disabled; 1121 the camera device uses fluorescent light as the assumed scene 1122 illumination for white balance. 1123 1124 While the exact white balance transforms are up to the 1125 camera device, they will approximately match the CIE 1126 standard illuminant F2. 1127 1128 The application's values for android.colorCorrection.transform 1129 and android.colorCorrection.gains are ignored. 1130 For devices that support the MANUAL_POST_PROCESSING capability, the 1131 values used by the camera device for the transform and gains 1132 will be available in the capture result for this request. 1133 </notes> 1134 </value> 1135 <value>WARM_FLUORESCENT 1136 <notes> 1137 The camera device's auto-white balance routine is disabled; 1138 the camera device uses warm fluorescent light as the assumed scene 1139 illumination for white balance. 1140 1141 While the exact white balance transforms are up to the 1142 camera device, they will approximately match the CIE 1143 standard illuminant F4. 1144 1145 The application's values for android.colorCorrection.transform 1146 and android.colorCorrection.gains are ignored. 1147 For devices that support the MANUAL_POST_PROCESSING capability, the 1148 values used by the camera device for the transform and gains 1149 will be available in the capture result for this request. 1150 </notes> 1151 </value> 1152 <value>DAYLIGHT 1153 <notes> 1154 The camera device's auto-white balance routine is disabled; 1155 the camera device uses daylight light as the assumed scene 1156 illumination for white balance. 1157 1158 While the exact white balance transforms are up to the 1159 camera device, they will approximately match the CIE 1160 standard illuminant D65. 1161 1162 The application's values for android.colorCorrection.transform 1163 and android.colorCorrection.gains are ignored. 1164 For devices that support the MANUAL_POST_PROCESSING capability, the 1165 values used by the camera device for the transform and gains 1166 will be available in the capture result for this request. 1167 </notes> 1168 </value> 1169 <value>CLOUDY_DAYLIGHT 1170 <notes> 1171 The camera device's auto-white balance routine is disabled; 1172 the camera device uses cloudy daylight light as the assumed scene 1173 illumination for white balance. 1174 1175 The application's values for android.colorCorrection.transform 1176 and android.colorCorrection.gains are ignored. 1177 For devices that support the MANUAL_POST_PROCESSING capability, the 1178 values used by the camera device for the transform and gains 1179 will be available in the capture result for this request. 1180 </notes> 1181 </value> 1182 <value>TWILIGHT 1183 <notes> 1184 The camera device's auto-white balance routine is disabled; 1185 the camera device uses twilight light as the assumed scene 1186 illumination for white balance. 1187 1188 The application's values for android.colorCorrection.transform 1189 and android.colorCorrection.gains are ignored. 1190 For devices that support the MANUAL_POST_PROCESSING capability, the 1191 values used by the camera device for the transform and gains 1192 will be available in the capture result for this request. 1193 </notes> 1194 </value> 1195 <value>SHADE 1196 <notes> 1197 The camera device's auto-white balance routine is disabled; 1198 the camera device uses shade light as the assumed scene 1199 illumination for white balance. 1200 1201 The application's values for android.colorCorrection.transform 1202 and android.colorCorrection.gains are ignored. 1203 For devices that support the MANUAL_POST_PROCESSING capability, the 1204 values used by the camera device for the transform and gains 1205 will be available in the capture result for this request. 1206 </notes> 1207 </value> 1208 </enum> 1209 <description>Whether auto-white balance (AWB) is currently setting the color 1210 transform fields, and what its illumination target 1211 is.</description> 1212 <range>android.control.awbAvailableModes</range> 1213 <details> 1214 This control is only effective if android.control.mode is AUTO. 1215 1216 When set to the ON mode, the camera device's auto-white balance 1217 routine is enabled, overriding the application's selected 1218 android.colorCorrection.transform, android.colorCorrection.gains and 1219 android.colorCorrection.mode. Note that when android.control.aeMode 1220 is OFF, the behavior of AWB is device dependent. It is recommened to 1221 also set AWB mode to OFF or lock AWB by using android.control.awbLock before 1222 setting AE mode to OFF. 1223 1224 When set to the OFF mode, the camera device's auto-white balance 1225 routine is disabled. The application manually controls the white 1226 balance by android.colorCorrection.transform, android.colorCorrection.gains 1227 and android.colorCorrection.mode. 1228 1229 When set to any other modes, the camera device's auto-white 1230 balance routine is disabled. The camera device uses each 1231 particular illumination target for white balance 1232 adjustment. The application's values for 1233 android.colorCorrection.transform, 1234 android.colorCorrection.gains and 1235 android.colorCorrection.mode are ignored. 1236 </details> 1237 <tag id="BC" /> 1238 </entry> 1239 <entry name="awbRegions" type="int32" visibility="public" 1240 optional="true" container="array" typedef="meteringRectangle"> 1241 <array> 1242 <size>5</size> 1243 <size>area_count</size> 1244 </array> 1245 <description>List of metering areas to use for auto-white-balance illuminant 1246 estimation.</description> 1247 <units>Pixel coordinates within android.sensor.info.activeArraySize</units> 1248 <range>Coordinates must be between `[(0,0), (width, height))` of 1249 android.sensor.info.activeArraySize</range> 1250 <details> 1251 Not available if android.control.maxRegionsAwb is 0. 1252 Otherwise will always be present. 1253 1254 The maximum number of regions supported by the device is determined by the value 1255 of android.control.maxRegionsAwb. 1256 1257 The coordinate system is based on the active pixel array, 1258 with (0,0) being the top-left pixel in the active pixel array, and 1259 (android.sensor.info.activeArraySize.width - 1, 1260 android.sensor.info.activeArraySize.height - 1) being the 1261 bottom-right pixel in the active pixel array. 1262 1263 The weight must range from 0 to 1000, and represents a weight 1264 for every pixel in the area. This means that a large metering area 1265 with the same weight as a smaller area will have more effect in 1266 the metering result. Metering areas can partially overlap and the 1267 camera device will add the weights in the overlap region. 1268 1269 The weights are relative to weights of other white balance metering regions, so if 1270 only one region is used, all non-zero weights will have the same effect. A region with 1271 0 weight is ignored. 1272 1273 If all regions have 0 weight, then no specific metering area needs to be used by the 1274 camera device. 1275 1276 If the metering region is outside the used android.scaler.cropRegion returned in 1277 capture result metadata, the camera device will ignore the sections outside the crop 1278 region and output only the intersection rectangle as the metering region in the result 1279 metadata. If the region is entirely outside the crop region, it will be ignored and 1280 not reported in the result metadata. 1281 </details> 1282 <hal_details> 1283 The HAL level representation of MeteringRectangle[] is a 1284 int[5 * area_count]. 1285 Every five elements represent a metering region of 1286 (xmin, ymin, xmax, ymax, weight). 1287 The rectangle is defined to be inclusive on xmin and ymin, but 1288 exclusive on xmax and ymax. 1289 </hal_details> 1290 <tag id="BC" /> 1291 </entry> 1292 <entry name="captureIntent" type="byte" visibility="public" enum="true" 1293 hwlevel="legacy"> 1294 <enum> 1295 <value>CUSTOM 1296 <notes>The goal of this request doesn't fall into the other 1297 categories. The camera device will default to preview-like 1298 behavior.</notes></value> 1299 <value>PREVIEW 1300 <notes>This request is for a preview-like use case. 1301 1302 The precapture trigger may be used to start off a metering 1303 w/flash sequence. 1304 </notes></value> 1305 <value>STILL_CAPTURE 1306 <notes>This request is for a still capture-type 1307 use case. 1308 1309 If the flash unit is under automatic control, it may fire as needed. 1310 </notes></value> 1311 <value>VIDEO_RECORD 1312 <notes>This request is for a video recording 1313 use case.</notes></value> 1314 <value>VIDEO_SNAPSHOT 1315 <notes>This request is for a video snapshot (still 1316 image while recording video) use case. 1317 1318 The camera device should take the highest-quality image 1319 possible (given the other settings) without disrupting the 1320 frame rate of video recording. </notes></value> 1321 <value>ZERO_SHUTTER_LAG 1322 <notes>This request is for a ZSL usecase; the 1323 application will stream full-resolution images and 1324 reprocess one or several later for a final 1325 capture. 1326 </notes></value> 1327 <value>MANUAL 1328 <notes>This request is for manual capture use case where 1329 the applications want to directly control the capture parameters. 1330 1331 For example, the application may wish to manually control 1332 android.sensor.exposureTime, android.sensor.sensitivity, etc. 1333 </notes></value> 1334 </enum> 1335 <description>Information to the camera device 3A (auto-exposure, 1336 auto-focus, auto-white balance) routines about the purpose 1337 of this capture, to help the camera device to decide optimal 3A 1338 strategy.</description> 1339 <details>This control (except for MANUAL) is only effective if 1340 `android.control.mode != OFF` and any 3A routine is active. 1341 1342 ZERO_SHUTTER_LAG will be supported if android.request.availableCapabilities 1343 contains PRIVATE_REPROCESSING or YUV_REPROCESSING. MANUAL will be supported if 1344 android.request.availableCapabilities contains MANUAL_SENSOR. Other intent values are 1345 always supported. 1346 </details> 1347 <tag id="BC" /> 1348 </entry> 1349 <entry name="effectMode" type="byte" visibility="public" enum="true" 1350 hwlevel="legacy"> 1351 <enum> 1352 <value>OFF 1353 <notes> 1354 No color effect will be applied. 1355 </notes> 1356 </value> 1357 <value optional="true">MONO 1358 <notes> 1359 A "monocolor" effect where the image is mapped into 1360 a single color. 1361 1362 This will typically be grayscale. 1363 </notes> 1364 </value> 1365 <value optional="true">NEGATIVE 1366 <notes> 1367 A "photo-negative" effect where the image's colors 1368 are inverted. 1369 </notes> 1370 </value> 1371 <value optional="true">SOLARIZE 1372 <notes> 1373 A "solarisation" effect (Sabattier effect) where the 1374 image is wholly or partially reversed in 1375 tone. 1376 </notes> 1377 </value> 1378 <value optional="true">SEPIA 1379 <notes> 1380 A "sepia" effect where the image is mapped into warm 1381 gray, red, and brown tones. 1382 </notes> 1383 </value> 1384 <value optional="true">POSTERIZE 1385 <notes> 1386 A "posterization" effect where the image uses 1387 discrete regions of tone rather than a continuous 1388 gradient of tones. 1389 </notes> 1390 </value> 1391 <value optional="true">WHITEBOARD 1392 <notes> 1393 A "whiteboard" effect where the image is typically displayed 1394 as regions of white, with black or grey details. 1395 </notes> 1396 </value> 1397 <value optional="true">BLACKBOARD 1398 <notes> 1399 A "blackboard" effect where the image is typically displayed 1400 as regions of black, with white or grey details. 1401 </notes> 1402 </value> 1403 <value optional="true">AQUA 1404 <notes> 1405 An "aqua" effect where a blue hue is added to the image. 1406 </notes> 1407 </value> 1408 </enum> 1409 <description>A special color effect to apply.</description> 1410 <range>android.control.availableEffects</range> 1411 <details> 1412 When this mode is set, a color effect will be applied 1413 to images produced by the camera device. The interpretation 1414 and implementation of these color effects is left to the 1415 implementor of the camera device, and should not be 1416 depended on to be consistent (or present) across all 1417 devices. 1418 </details> 1419 <tag id="BC" /> 1420 </entry> 1421 <entry name="mode" type="byte" visibility="public" enum="true" 1422 hwlevel="legacy"> 1423 <enum> 1424 <value>OFF 1425 <notes>Full application control of pipeline. 1426 1427 All control by the device's metering and focusing (3A) 1428 routines is disabled, and no other settings in 1429 android.control.* have any effect, except that 1430 android.control.captureIntent may be used by the camera 1431 device to select post-processing values for processing 1432 blocks that do not allow for manual control, or are not 1433 exposed by the camera API. 1434 1435 However, the camera device's 3A routines may continue to 1436 collect statistics and update their internal state so that 1437 when control is switched to AUTO mode, good control values 1438 can be immediately applied. 1439 </notes></value> 1440 <value>AUTO 1441 <notes>Use settings for each individual 3A routine. 1442 1443 Manual control of capture parameters is disabled. All 1444 controls in android.control.* besides sceneMode take 1445 effect.</notes></value> 1446 <value optional="true">USE_SCENE_MODE 1447 <notes>Use a specific scene mode. 1448 1449 Enabling this disables control.aeMode, control.awbMode and 1450 control.afMode controls; the camera device will ignore 1451 those settings while USE_SCENE_MODE is active (except for 1452 FACE_PRIORITY scene mode). Other control entries are still active. 1453 This setting can only be used if scene mode is supported (i.e. 1454 android.control.availableSceneModes 1455 contain some modes other than DISABLED).</notes></value> 1456 <value optional="true">OFF_KEEP_STATE 1457 <notes>Same as OFF mode, except that this capture will not be 1458 used by camera device background auto-exposure, auto-white balance and 1459 auto-focus algorithms (3A) to update their statistics. 1460 1461 Specifically, the 3A routines are locked to the last 1462 values set from a request with AUTO, OFF, or 1463 USE_SCENE_MODE, and any statistics or state updates 1464 collected from manual captures with OFF_KEEP_STATE will be 1465 discarded by the camera device. 1466 </notes></value> 1467 </enum> 1468 <description>Overall mode of 3A (auto-exposure, auto-white-balance, auto-focus) control 1469 routines.</description> 1470 <range>android.control.availableModes</range> 1471 <details> 1472 This is a top-level 3A control switch. When set to OFF, all 3A control 1473 by the camera device is disabled. The application must set the fields for 1474 capture parameters itself. 1475 1476 When set to AUTO, the individual algorithm controls in 1477 android.control.* are in effect, such as android.control.afMode. 1478 1479 When set to USE_SCENE_MODE, the individual controls in 1480 android.control.* are mostly disabled, and the camera device implements 1481 one of the scene mode settings (such as ACTION, SUNSET, or PARTY) 1482 as it wishes. The camera device scene mode 3A settings are provided by 1483 {@link android.hardware.camera2.CaptureResult capture results}. 1484 1485 When set to OFF_KEEP_STATE, it is similar to OFF mode, the only difference 1486 is that this frame will not be used by camera device background 3A statistics 1487 update, as if this frame is never captured. This mode can be used in the scenario 1488 where the application doesn't want a 3A manual control capture to affect 1489 the subsequent auto 3A capture results. 1490 </details> 1491 <tag id="BC" /> 1492 </entry> 1493 <entry name="sceneMode" type="byte" visibility="public" enum="true" 1494 hwlevel="legacy"> 1495 <enum> 1496 <value id="0">DISABLED 1497 <notes> 1498 Indicates that no scene modes are set for a given capture request. 1499 </notes> 1500 </value> 1501 <value>FACE_PRIORITY 1502 <notes>If face detection support exists, use face 1503 detection data for auto-focus, auto-white balance, and 1504 auto-exposure routines. 1505 1506 If face detection statistics are disabled 1507 (i.e. android.statistics.faceDetectMode is set to OFF), 1508 this should still operate correctly (but will not return 1509 face detection statistics to the framework). 1510 1511 Unlike the other scene modes, android.control.aeMode, 1512 android.control.awbMode, and android.control.afMode 1513 remain active when FACE_PRIORITY is set. 1514 </notes> 1515 </value> 1516 <value optional="true">ACTION 1517 <notes> 1518 Optimized for photos of quickly moving objects. 1519 1520 Similar to SPORTS. 1521 </notes> 1522 </value> 1523 <value optional="true">PORTRAIT 1524 <notes> 1525 Optimized for still photos of people. 1526 </notes> 1527 </value> 1528 <value optional="true">LANDSCAPE 1529 <notes> 1530 Optimized for photos of distant macroscopic objects. 1531 </notes> 1532 </value> 1533 <value optional="true">NIGHT 1534 <notes> 1535 Optimized for low-light settings. 1536 </notes> 1537 </value> 1538 <value optional="true">NIGHT_PORTRAIT 1539 <notes> 1540 Optimized for still photos of people in low-light 1541 settings. 1542 </notes> 1543 </value> 1544 <value optional="true">THEATRE 1545 <notes> 1546 Optimized for dim, indoor settings where flash must 1547 remain off. 1548 </notes> 1549 </value> 1550 <value optional="true">BEACH 1551 <notes> 1552 Optimized for bright, outdoor beach settings. 1553 </notes> 1554 </value> 1555 <value optional="true">SNOW 1556 <notes> 1557 Optimized for bright, outdoor settings containing snow. 1558 </notes> 1559 </value> 1560 <value optional="true">SUNSET 1561 <notes> 1562 Optimized for scenes of the setting sun. 1563 </notes> 1564 </value> 1565 <value optional="true">STEADYPHOTO 1566 <notes> 1567 Optimized to avoid blurry photos due to small amounts of 1568 device motion (for example: due to hand shake). 1569 </notes> 1570 </value> 1571 <value optional="true">FIREWORKS 1572 <notes> 1573 Optimized for nighttime photos of fireworks. 1574 </notes> 1575 </value> 1576 <value optional="true">SPORTS 1577 <notes> 1578 Optimized for photos of quickly moving people. 1579 1580 Similar to ACTION. 1581 </notes> 1582 </value> 1583 <value optional="true">PARTY 1584 <notes> 1585 Optimized for dim, indoor settings with multiple moving 1586 people. 1587 </notes> 1588 </value> 1589 <value optional="true">CANDLELIGHT 1590 <notes> 1591 Optimized for dim settings where the main light source 1592 is a flame. 1593 </notes> 1594 </value> 1595 <value optional="true">BARCODE 1596 <notes> 1597 Optimized for accurately capturing a photo of barcode 1598 for use by camera applications that wish to read the 1599 barcode value. 1600 </notes> 1601 </value> 1602 <value deprecated="true" optional="true">HIGH_SPEED_VIDEO 1603 <notes> 1604 This is deprecated, please use {@link 1605 android.hardware.camera2.CameraDevice#createConstrainedHighSpeedCaptureSession} 1606 and {@link 1607 android.hardware.camera2.CameraConstrainedHighSpeedCaptureSession#createHighSpeedRequestList} 1608 for high speed video recording. 1609 1610 Optimized for high speed video recording (frame rate >=60fps) use case. 1611 1612 The supported high speed video sizes and fps ranges are specified in 1613 android.control.availableHighSpeedVideoConfigurations. To get desired 1614 output frame rates, the application is only allowed to select video size 1615 and fps range combinations listed in this static metadata. The fps range 1616 can be control via android.control.aeTargetFpsRange. 1617 1618 In this mode, the camera device will override aeMode, awbMode, and afMode to 1619 ON, ON, and CONTINUOUS_VIDEO, respectively. All post-processing block mode 1620 controls will be overridden to be FAST. Therefore, no manual control of capture 1621 and post-processing parameters is possible. All other controls operate the 1622 same as when android.control.mode == AUTO. This means that all other 1623 android.control.* fields continue to work, such as 1624 1625 * android.control.aeTargetFpsRange 1626 * android.control.aeExposureCompensation 1627 * android.control.aeLock 1628 * android.control.awbLock 1629 * android.control.effectMode 1630 * android.control.aeRegions 1631 * android.control.afRegions 1632 * android.control.awbRegions 1633 * android.control.afTrigger 1634 * android.control.aePrecaptureTrigger 1635 1636 Outside of android.control.*, the following controls will work: 1637 1638 * android.flash.mode (automatic flash for still capture will not work since aeMode is ON) 1639 * android.lens.opticalStabilizationMode (if it is supported) 1640 * android.scaler.cropRegion 1641 * android.statistics.faceDetectMode 1642 1643 For high speed recording use case, the actual maximum supported frame rate may 1644 be lower than what camera can output, depending on the destination Surfaces for 1645 the image data. For example, if the destination surface is from video encoder, 1646 the application need check if the video encoder is capable of supporting the 1647 high frame rate for a given video size, or it will end up with lower recording 1648 frame rate. If the destination surface is from preview window, the preview frame 1649 rate will be bounded by the screen refresh rate. 1650 1651 The camera device will only support up to 2 output high speed streams 1652 (processed non-stalling format defined in android.request.maxNumOutputStreams) 1653 in this mode. This control will be effective only if all of below conditions are true: 1654 1655 * The application created no more than maxNumHighSpeedStreams processed non-stalling 1656 format output streams, where maxNumHighSpeedStreams is calculated as 1657 min(2, android.request.maxNumOutputStreams[Processed (but not-stalling)]). 1658 * The stream sizes are selected from the sizes reported by 1659 android.control.availableHighSpeedVideoConfigurations. 1660 * No processed non-stalling or raw streams are configured. 1661 1662 When above conditions are NOT satistied, the controls of this mode and 1663 android.control.aeTargetFpsRange will be ignored by the camera device, 1664 the camera device will fall back to android.control.mode `==` AUTO, 1665 and the returned capture result metadata will give the fps range choosen 1666 by the camera device. 1667 1668 Switching into or out of this mode may trigger some camera ISP/sensor 1669 reconfigurations, which may introduce extra latency. It is recommended that 1670 the application avoids unnecessary scene mode switch as much as possible. 1671 </notes> 1672 </value> 1673 <value optional="true">HDR 1674 <notes> 1675 Turn on a device-specific high dynamic range (HDR) mode. 1676 1677 In this scene mode, the camera device captures images 1678 that keep a larger range of scene illumination levels 1679 visible in the final image. For example, when taking a 1680 picture of a object in front of a bright window, both 1681 the object and the scene through the window may be 1682 visible when using HDR mode, while in normal AUTO mode, 1683 one or the other may be poorly exposed. As a tradeoff, 1684 HDR mode generally takes much longer to capture a single 1685 image, has no user control, and may have other artifacts 1686 depending on the HDR method used. 1687 1688 Therefore, HDR captures operate at a much slower rate 1689 than regular captures. 1690 1691 In this mode, on LIMITED or FULL devices, when a request 1692 is made with a android.control.captureIntent of 1693 STILL_CAPTURE, the camera device will capture an image 1694 using a high dynamic range capture technique. On LEGACY 1695 devices, captures that target a JPEG-format output will 1696 be captured with HDR, and the capture intent is not 1697 relevant. 1698 1699 The HDR capture may involve the device capturing a burst 1700 of images internally and combining them into one, or it 1701 may involve the device using specialized high dynamic 1702 range capture hardware. In all cases, a single image is 1703 produced in response to a capture request submitted 1704 while in HDR mode. 1705 1706 Since substantial post-processing is generally needed to 1707 produce an HDR image, only YUV, PRIVATE, and JPEG 1708 outputs are supported for LIMITED/FULL device HDR 1709 captures, and only JPEG outputs are supported for LEGACY 1710 HDR captures. Using a RAW output for HDR capture is not 1711 supported. 1712 1713 Some devices may also support always-on HDR, which 1714 applies HDR processing at full frame rate. For these 1715 devices, intents other than STILL_CAPTURE will also 1716 produce an HDR output with no frame rate impact compared 1717 to normal operation, though the quality may be lower 1718 than for STILL_CAPTURE intents. 1719 1720 If SCENE_MODE_HDR is used with unsupported output types 1721 or capture intents, the images captured will be as if 1722 the SCENE_MODE was not enabled at all. 1723 </notes> 1724 </value> 1725 <value optional="true" hidden="true">FACE_PRIORITY_LOW_LIGHT 1726 <notes>Same as FACE_PRIORITY scene mode, except that the camera 1727 device will choose higher sensitivity values (android.sensor.sensitivity) 1728 under low light conditions. 1729 1730 The camera device may be tuned to expose the images in a reduced 1731 sensitivity range to produce the best quality images. For example, 1732 if the android.sensor.info.sensitivityRange gives range of [100, 1600], 1733 the camera device auto-exposure routine tuning process may limit the actual 1734 exposure sensitivity range to [100, 1200] to ensure that the noise level isn't 1735 exessive in order to preserve the image quality. Under this situation, the image under 1736 low light may be under-exposed when the sensor max exposure time (bounded by the 1737 android.control.aeTargetFpsRange when android.control.aeMode is one of the 1738 ON_* modes) and effective max sensitivity are reached. This scene mode allows the 1739 camera device auto-exposure routine to increase the sensitivity up to the max 1740 sensitivity specified by android.sensor.info.sensitivityRange when the scene is too 1741 dark and the max exposure time is reached. The captured images may be noisier 1742 compared with the images captured in normal FACE_PRIORITY mode; therefore, it is 1743 recommended that the application only use this scene mode when it is capable of 1744 reducing the noise level of the captured images. 1745 1746 Unlike the other scene modes, android.control.aeMode, 1747 android.control.awbMode, and android.control.afMode 1748 remain active when FACE_PRIORITY_LOW_LIGHT is set. 1749 </notes> 1750 </value> 1751 <value optional="true" hidden="true" id="100">DEVICE_CUSTOM_START 1752 <notes> 1753 Scene mode values within the range of 1754 `[DEVICE_CUSTOM_START, DEVICE_CUSTOM_END]` are reserved for device specific 1755 customized scene modes. 1756 </notes> 1757 </value> 1758 <value optional="true" hidden="true" id="127">DEVICE_CUSTOM_END 1759 <notes> 1760 Scene mode values within the range of 1761 `[DEVICE_CUSTOM_START, DEVICE_CUSTOM_END]` are reserved for device specific 1762 customized scene modes. 1763 </notes> 1764 </value> 1765 </enum> 1766 <description> 1767 Control for which scene mode is currently active. 1768 </description> 1769 <range>android.control.availableSceneModes</range> 1770 <details> 1771 Scene modes are custom camera modes optimized for a certain set of conditions and 1772 capture settings. 1773 1774 This is the mode that that is active when 1775 `android.control.mode == USE_SCENE_MODE`. Aside from FACE_PRIORITY, these modes will 1776 disable android.control.aeMode, android.control.awbMode, and android.control.afMode 1777 while in use. 1778 1779 The interpretation and implementation of these scene modes is left 1780 to the implementor of the camera device. Their behavior will not be 1781 consistent across all devices, and any given device may only implement 1782 a subset of these modes. 1783 </details> 1784 <hal_details> 1785 HAL implementations that include scene modes are expected to provide 1786 the per-scene settings to use for android.control.aeMode, 1787 android.control.awbMode, and android.control.afMode in 1788 android.control.sceneModeOverrides. 1789 1790 For HIGH_SPEED_VIDEO mode, if it is included in android.control.availableSceneModes, 1791 the HAL must list supported video size and fps range in 1792 android.control.availableHighSpeedVideoConfigurations. For a given size, e.g. 1793 1280x720, if the HAL has two different sensor configurations for normal streaming 1794 mode and high speed streaming, when this scene mode is set/reset in a sequence of capture 1795 requests, the HAL may have to switch between different sensor modes. 1796 This mode is deprecated in HAL3.3, to support high speed video recording, please implement 1797 android.control.availableHighSpeedVideoConfigurations and CONSTRAINED_HIGH_SPEED_VIDEO 1798 capbility defined in android.request.availableCapabilities. 1799 </hal_details> 1800 <tag id="BC" /> 1801 </entry> 1802 <entry name="videoStabilizationMode" type="byte" visibility="public" 1803 enum="true" hwlevel="legacy"> 1804 <enum> 1805 <value>OFF 1806 <notes> 1807 Video stabilization is disabled. 1808 </notes></value> 1809 <value>ON 1810 <notes> 1811 Video stabilization is enabled. 1812 </notes></value> 1813 </enum> 1814 <description>Whether video stabilization is 1815 active.</description> 1816 <details> 1817 Video stabilization automatically warps images from 1818 the camera in order to stabilize motion between consecutive frames. 1819 1820 If enabled, video stabilization can modify the 1821 android.scaler.cropRegion to keep the video stream stabilized. 1822 1823 Switching between different video stabilization modes may take several 1824 frames to initialize, the camera device will report the current mode 1825 in capture result metadata. For example, When "ON" mode is requested, 1826 the video stabilization modes in the first several capture results may 1827 still be "OFF", and it will become "ON" when the initialization is 1828 done. 1829 1830 In addition, not all recording sizes or frame rates may be supported for 1831 stabilization by a device that reports stabilization support. It is guaranteed 1832 that an output targeting a MediaRecorder or MediaCodec will be stabilized if 1833 the recording resolution is less than or equal to 1920 x 1080 (width less than 1834 or equal to 1920, height less than or equal to 1080), and the recording 1835 frame rate is less than or equal to 30fps. At other sizes, the CaptureResult 1836 android.control.videoStabilizationMode field will return 1837 OFF if the recording output is not stabilized, or if there are no output 1838 Surface types that can be stabilized. 1839 1840 If a camera device supports both this mode and OIS 1841 (android.lens.opticalStabilizationMode), turning both modes on may 1842 produce undesirable interaction, so it is recommended not to enable 1843 both at the same time. 1844 </details> 1845 <tag id="BC" /> 1846 </entry> 1847 </controls> 1848 <static> 1849 <entry name="aeAvailableAntibandingModes" type="byte" visibility="public" 1850 type_notes="list of enums" container="array" typedef="enumList" 1851 hwlevel="legacy"> 1852 <array> 1853 <size>n</size> 1854 </array> 1855 <description> 1856 List of auto-exposure antibanding modes for android.control.aeAntibandingMode that are 1857 supported by this camera device. 1858 </description> 1859 <range>Any value listed in android.control.aeAntibandingMode</range> 1860 <details> 1861 Not all of the auto-exposure anti-banding modes may be 1862 supported by a given camera device. This field lists the 1863 valid anti-banding modes that the application may request 1864 for this camera device with the 1865 android.control.aeAntibandingMode control. 1866 </details> 1867 <tag id="BC" /> 1868 </entry> 1869 <entry name="aeAvailableModes" type="byte" visibility="public" 1870 type_notes="list of enums" container="array" typedef="enumList" 1871 hwlevel="legacy"> 1872 <array> 1873 <size>n</size> 1874 </array> 1875 <description> 1876 List of auto-exposure modes for android.control.aeMode that are supported by this camera 1877 device. 1878 </description> 1879 <range>Any value listed in android.control.aeMode</range> 1880 <details> 1881 Not all the auto-exposure modes may be supported by a 1882 given camera device, especially if no flash unit is 1883 available. This entry lists the valid modes for 1884 android.control.aeMode for this camera device. 1885 1886 All camera devices support ON, and all camera devices with flash 1887 units support ON_AUTO_FLASH and ON_ALWAYS_FLASH. 1888 1889 FULL mode camera devices always support OFF mode, 1890 which enables application control of camera exposure time, 1891 sensitivity, and frame duration. 1892 1893 LEGACY mode camera devices never support OFF mode. 1894 LIMITED mode devices support OFF if they support the MANUAL_SENSOR 1895 capability. 1896 </details> 1897 <tag id="BC" /> 1898 </entry> 1899 <entry name="aeAvailableTargetFpsRanges" type="int32" visibility="public" 1900 type_notes="list of pairs of frame rates" 1901 container="array" typedef="rangeInt" 1902 hwlevel="legacy"> 1903 <array> 1904 <size>2</size> 1905 <size>n</size> 1906 </array> 1907 <description>List of frame rate ranges for android.control.aeTargetFpsRange supported by 1908 this camera device.</description> 1909 <units>Frames per second (FPS)</units> 1910 <details> 1911 For devices at the LEGACY level or above: 1912 1913 * For constant-framerate recording, for each normal 1914 {@link android.media.CamcorderProfile CamcorderProfile}, that is, a 1915 {@link android.media.CamcorderProfile CamcorderProfile} that has 1916 {@link android.media.CamcorderProfile#quality quality} in 1917 the range [{@link android.media.CamcorderProfile#QUALITY_LOW QUALITY_LOW}, 1918 {@link android.media.CamcorderProfile#QUALITY_2160P QUALITY_2160P}], if the profile is 1919 supported by the device and has 1920 {@link android.media.CamcorderProfile#videoFrameRate videoFrameRate} `x`, this list will 1921 always include (`x`,`x`). 1922 1923 * Also, a camera device must either not support any 1924 {@link android.media.CamcorderProfile CamcorderProfile}, 1925 or support at least one 1926 normal {@link android.media.CamcorderProfile CamcorderProfile} that has 1927 {@link android.media.CamcorderProfile#videoFrameRate videoFrameRate} `x` >= 24. 1928 1929 For devices at the LIMITED level or above: 1930 1931 * For YUV_420_888 burst capture use case, this list will always include (`min`, `max`) 1932 and (`max`, `max`) where `min` <= 15 and `max` = the maximum output frame rate of the 1933 maximum YUV_420_888 output size. 1934 </details> 1935 <tag id="BC" /> 1936 </entry> 1937 <entry name="aeCompensationRange" type="int32" visibility="public" 1938 container="array" typedef="rangeInt" 1939 hwlevel="legacy"> 1940 <array> 1941 <size>2</size> 1942 </array> 1943 <description>Maximum and minimum exposure compensation values for 1944 android.control.aeExposureCompensation, in counts of android.control.aeCompensationStep, 1945 that are supported by this camera device.</description> 1946 <range> 1947 Range [0,0] indicates that exposure compensation is not supported. 1948 1949 For LIMITED and FULL devices, range must follow below requirements if exposure 1950 compensation is supported (`range != [0, 0]`): 1951 1952 `Min.exposure compensation * android.control.aeCompensationStep <= -2 EV` 1953 1954 `Max.exposure compensation * android.control.aeCompensationStep >= 2 EV` 1955 1956 LEGACY devices may support a smaller range than this. 1957 </range> 1958 <tag id="BC" /> 1959 </entry> 1960 <entry name="aeCompensationStep" type="rational" visibility="public" 1961 hwlevel="legacy"> 1962 <description>Smallest step by which the exposure compensation 1963 can be changed.</description> 1964 <units>Exposure Value (EV)</units> 1965 <details> 1966 This is the unit for android.control.aeExposureCompensation. For example, if this key has 1967 a value of `1/2`, then a setting of `-2` for android.control.aeExposureCompensation means 1968 that the target EV offset for the auto-exposure routine is -1 EV. 1969 1970 One unit of EV compensation changes the brightness of the captured image by a factor 1971 of two. +1 EV doubles the image brightness, while -1 EV halves the image brightness. 1972 </details> 1973 <hal_details> 1974 This must be less than or equal to 1/2. 1975 </hal_details> 1976 <tag id="BC" /> 1977 </entry> 1978 <entry name="afAvailableModes" type="byte" visibility="public" 1979 type_notes="List of enums" container="array" typedef="enumList" 1980 hwlevel="legacy"> 1981 <array> 1982 <size>n</size> 1983 </array> 1984 <description> 1985 List of auto-focus (AF) modes for android.control.afMode that are 1986 supported by this camera device. 1987 </description> 1988 <range>Any value listed in android.control.afMode</range> 1989 <details> 1990 Not all the auto-focus modes may be supported by a 1991 given camera device. This entry lists the valid modes for 1992 android.control.afMode for this camera device. 1993 1994 All LIMITED and FULL mode camera devices will support OFF mode, and all 1995 camera devices with adjustable focuser units 1996 (`android.lens.info.minimumFocusDistance > 0`) will support AUTO mode. 1997 1998 LEGACY devices will support OFF mode only if they support 1999 focusing to infinity (by also setting android.lens.focusDistance to 2000 `0.0f`). 2001 </details> 2002 <tag id="BC" /> 2003 </entry> 2004 <entry name="availableEffects" type="byte" visibility="public" 2005 type_notes="List of enums (android.control.effectMode)." container="array" 2006 typedef="enumList" hwlevel="legacy"> 2007 <array> 2008 <size>n</size> 2009 </array> 2010 <description> 2011 List of color effects for android.control.effectMode that are supported by this camera 2012 device. 2013 </description> 2014 <range>Any value listed in android.control.effectMode</range> 2015 <details> 2016 This list contains the color effect modes that can be applied to 2017 images produced by the camera device. 2018 Implementations are not expected to be consistent across all devices. 2019 If no color effect modes are available for a device, this will only list 2020 OFF. 2021 2022 A color effect will only be applied if 2023 android.control.mode != OFF. OFF is always included in this list. 2024 2025 This control has no effect on the operation of other control routines such 2026 as auto-exposure, white balance, or focus. 2027 </details> 2028 <tag id="BC" /> 2029 </entry> 2030 <entry name="availableSceneModes" type="byte" visibility="public" 2031 type_notes="List of enums (android.control.sceneMode)." 2032 container="array" typedef="enumList" hwlevel="legacy"> 2033 <array> 2034 <size>n</size> 2035 </array> 2036 <description> 2037 List of scene modes for android.control.sceneMode that are supported by this camera 2038 device. 2039 </description> 2040 <range>Any value listed in android.control.sceneMode</range> 2041 <details> 2042 This list contains scene modes that can be set for the camera device. 2043 Only scene modes that have been fully implemented for the 2044 camera device may be included here. Implementations are not expected 2045 to be consistent across all devices. 2046 2047 If no scene modes are supported by the camera device, this 2048 will be set to DISABLED. Otherwise DISABLED will not be listed. 2049 2050 FACE_PRIORITY is always listed if face detection is 2051 supported (i.e.`android.statistics.info.maxFaceCount > 2052 0`). 2053 </details> 2054 <tag id="BC" /> 2055 </entry> 2056 <entry name="availableVideoStabilizationModes" type="byte" 2057 visibility="public" type_notes="List of enums." container="array" 2058 typedef="enumList" hwlevel="legacy"> 2059 <array> 2060 <size>n</size> 2061 </array> 2062 <description> 2063 List of video stabilization modes for android.control.videoStabilizationMode 2064 that are supported by this camera device. 2065 </description> 2066 <range>Any value listed in android.control.videoStabilizationMode</range> 2067 <details> 2068 OFF will always be listed. 2069 </details> 2070 <tag id="BC" /> 2071 </entry> 2072 <entry name="awbAvailableModes" type="byte" visibility="public" 2073 type_notes="List of enums" 2074 container="array" typedef="enumList" hwlevel="legacy"> 2075 <array> 2076 <size>n</size> 2077 </array> 2078 <description> 2079 List of auto-white-balance modes for android.control.awbMode that are supported by this 2080 camera device. 2081 </description> 2082 <range>Any value listed in android.control.awbMode</range> 2083 <details> 2084 Not all the auto-white-balance modes may be supported by a 2085 given camera device. This entry lists the valid modes for 2086 android.control.awbMode for this camera device. 2087 2088 All camera devices will support ON mode. 2089 2090 Camera devices that support the MANUAL_POST_PROCESSING capability will always support OFF 2091 mode, which enables application control of white balance, by using 2092 android.colorCorrection.transform and android.colorCorrection.gains 2093 (android.colorCorrection.mode must be set to TRANSFORM_MATRIX). This includes all FULL 2094 mode camera devices. 2095 </details> 2096 <tag id="BC" /> 2097 </entry> 2098 <entry name="maxRegions" type="int32" visibility="ndk_public" 2099 container="array" hwlevel="legacy"> 2100 <array> 2101 <size>3</size> 2102 </array> 2103 <description> 2104 List of the maximum number of regions that can be used for metering in 2105 auto-exposure (AE), auto-white balance (AWB), and auto-focus (AF); 2106 this corresponds to the the maximum number of elements in 2107 android.control.aeRegions, android.control.awbRegions, 2108 and android.control.afRegions. 2109 </description> 2110 <range> 2111 Value must be &gt;= 0 for each element. For full-capability devices 2112 this value must be &gt;= 1 for AE and AF. The order of the elements is: 2113 `(AE, AWB, AF)`.</range> 2114 <tag id="BC" /> 2115 </entry> 2116 <entry name="maxRegionsAe" type="int32" visibility="java_public" 2117 synthetic="true" hwlevel="legacy"> 2118 <description> 2119 The maximum number of metering regions that can be used by the auto-exposure (AE) 2120 routine. 2121 </description> 2122 <range>Value will be &gt;= 0. For FULL-capability devices, this 2123 value will be &gt;= 1. 2124 </range> 2125 <details> 2126 This corresponds to the the maximum allowed number of elements in 2127 android.control.aeRegions. 2128 </details> 2129 <hal_details>This entry is private to the framework. Fill in 2130 maxRegions to have this entry be automatically populated. 2131 </hal_details> 2132 </entry> 2133 <entry name="maxRegionsAwb" type="int32" visibility="java_public" 2134 synthetic="true" hwlevel="legacy"> 2135 <description> 2136 The maximum number of metering regions that can be used by the auto-white balance (AWB) 2137 routine. 2138 </description> 2139 <range>Value will be &gt;= 0. 2140 </range> 2141 <details> 2142 This corresponds to the the maximum allowed number of elements in 2143 android.control.awbRegions. 2144 </details> 2145 <hal_details>This entry is private to the framework. Fill in 2146 maxRegions to have this entry be automatically populated. 2147 </hal_details> 2148 </entry> 2149 <entry name="maxRegionsAf" type="int32" visibility="java_public" 2150 synthetic="true" hwlevel="legacy"> 2151 <description> 2152 The maximum number of metering regions that can be used by the auto-focus (AF) routine. 2153 </description> 2154 <range>Value will be &gt;= 0. For FULL-capability devices, this 2155 value will be &gt;= 1. 2156 </range> 2157 <details> 2158 This corresponds to the the maximum allowed number of elements in 2159 android.control.afRegions. 2160 </details> 2161 <hal_details>This entry is private to the framework. Fill in 2162 maxRegions to have this entry be automatically populated. 2163 </hal_details> 2164 </entry> 2165 <entry name="sceneModeOverrides" type="byte" visibility="system" 2166 container="array" hwlevel="limited"> 2167 <array> 2168 <size>3</size> 2169 <size>length(availableSceneModes)</size> 2170 </array> 2171 <description> 2172 Ordered list of auto-exposure, auto-white balance, and auto-focus 2173 settings to use with each available scene mode. 2174 </description> 2175 <range> 2176 For each available scene mode, the list must contain three 2177 entries containing the android.control.aeMode, 2178 android.control.awbMode, and android.control.afMode values used 2179 by the camera device. The entry order is `(aeMode, awbMode, afMode)` 2180 where aeMode has the lowest index position. 2181 </range> 2182 <details> 2183 When a scene mode is enabled, the camera device is expected 2184 to override android.control.aeMode, android.control.awbMode, 2185 and android.control.afMode with its preferred settings for 2186 that scene mode. 2187 2188 The order of this list matches that of availableSceneModes, 2189 with 3 entries for each mode. The overrides listed 2190 for FACE_PRIORITY and FACE_PRIORITY_LOW_LIGHT (if supported) are ignored, 2191 since for that mode the application-set android.control.aeMode, 2192 android.control.awbMode, and android.control.afMode values are 2193 used instead, matching the behavior when android.control.mode 2194 is set to AUTO. It is recommended that the FACE_PRIORITY and 2195 FACE_PRIORITY_LOW_LIGHT (if supported) overrides should be set to 0. 2196 2197 For example, if availableSceneModes contains 2198 `(FACE_PRIORITY, ACTION, NIGHT)`, then the camera framework 2199 expects sceneModeOverrides to have 9 entries formatted like: 2200 `(0, 0, 0, ON_AUTO_FLASH, AUTO, CONTINUOUS_PICTURE, 2201 ON_AUTO_FLASH, INCANDESCENT, AUTO)`. 2202 </details> 2203 <hal_details> 2204 To maintain backward compatibility, this list will be made available 2205 in the static metadata of the camera service. The camera service will 2206 use these values to set android.control.aeMode, 2207 android.control.awbMode, and android.control.afMode when using a scene 2208 mode other than FACE_PRIORITY and FACE_PRIORITY_LOW_LIGHT (if supported). 2209 </hal_details> 2210 <tag id="BC" /> 2211 </entry> 2212 </static> 2213 <dynamic> 2214 <entry name="aePrecaptureId" type="int32" visibility="system" deprecated="true"> 2215 <description>The ID sent with the latest 2216 CAMERA2_TRIGGER_PRECAPTURE_METERING call</description> 2217 <details>Must be 0 if no 2218 CAMERA2_TRIGGER_PRECAPTURE_METERING trigger received yet 2219 by HAL. Always updated even if AE algorithm ignores the 2220 trigger</details> 2221 </entry> 2222 <clone entry="android.control.aeAntibandingMode" kind="controls"> 2223 </clone> 2224 <clone entry="android.control.aeExposureCompensation" kind="controls"> 2225 </clone> 2226 <clone entry="android.control.aeLock" kind="controls"> 2227 </clone> 2228 <clone entry="android.control.aeMode" kind="controls"> 2229 </clone> 2230 <clone entry="android.control.aeRegions" kind="controls"> 2231 </clone> 2232 <clone entry="android.control.aeTargetFpsRange" kind="controls"> 2233 </clone> 2234 <clone entry="android.control.aePrecaptureTrigger" kind="controls"> 2235 </clone> 2236 <entry name="aeState" type="byte" visibility="public" enum="true" 2237 hwlevel="limited"> 2238 <enum> 2239 <value>INACTIVE 2240 <notes>AE is off or recently reset. 2241 2242 When a camera device is opened, it starts in 2243 this state. This is a transient state, the camera device may skip reporting 2244 this state in capture result.</notes></value> 2245 <value>SEARCHING 2246 <notes>AE doesn't yet have a good set of control values 2247 for the current scene. 2248 2249 This is a transient state, the camera device may skip 2250 reporting this state in capture result.</notes></value> 2251 <value>CONVERGED 2252 <notes>AE has a good set of control values for the 2253 current scene.</notes></value> 2254 <value>LOCKED 2255 <notes>AE has been locked.</notes></value> 2256 <value>FLASH_REQUIRED 2257 <notes>AE has a good set of control values, but flash 2258 needs to be fired for good quality still 2259 capture.</notes></value> 2260 <value>PRECAPTURE 2261 <notes>AE has been asked to do a precapture sequence 2262 and is currently executing it. 2263 2264 Precapture can be triggered through setting 2265 android.control.aePrecaptureTrigger to START. Currently 2266 active and completed (if it causes camera device internal AE lock) precapture 2267 metering sequence can be canceled through setting 2268 android.control.aePrecaptureTrigger to CANCEL. 2269 2270 Once PRECAPTURE completes, AE will transition to CONVERGED 2271 or FLASH_REQUIRED as appropriate. This is a transient 2272 state, the camera device may skip reporting this state in 2273 capture result.</notes></value> 2274 </enum> 2275 <description>Current state of the auto-exposure (AE) algorithm.</description> 2276 <details>Switching between or enabling AE modes (android.control.aeMode) always 2277 resets the AE state to INACTIVE. Similarly, switching between android.control.mode, 2278 or android.control.sceneMode if `android.control.mode == USE_SCENE_MODE` resets all 2279 the algorithm states to INACTIVE. 2280 2281 The camera device can do several state transitions between two results, if it is 2282 allowed by the state transition table. For example: INACTIVE may never actually be 2283 seen in a result. 2284 2285 The state in the result is the state for this image (in sync with this image): if 2286 AE state becomes CONVERGED, then the image data associated with this result should 2287 be good to use. 2288 2289 Below are state transition tables for different AE modes. 2290 2291 State | Transition Cause | New State | Notes 2292 :------------:|:----------------:|:---------:|:-----------------------: 2293 INACTIVE | | INACTIVE | Camera device auto exposure algorithm is disabled 2294 2295 When android.control.aeMode is AE_MODE_ON_*: 2296 2297 State | Transition Cause | New State | Notes 2298 :-------------:|:--------------------------------------------:|:--------------:|:-----------------: 2299 INACTIVE | Camera device initiates AE scan | SEARCHING | Values changing 2300 INACTIVE | android.control.aeLock is ON | LOCKED | Values locked 2301 SEARCHING | Camera device finishes AE scan | CONVERGED | Good values, not changing 2302 SEARCHING | Camera device finishes AE scan | FLASH_REQUIRED | Converged but too dark w/o flash 2303 SEARCHING | android.control.aeLock is ON | LOCKED | Values locked 2304 CONVERGED | Camera device initiates AE scan | SEARCHING | Values changing 2305 CONVERGED | android.control.aeLock is ON | LOCKED | Values locked 2306 FLASH_REQUIRED | Camera device initiates AE scan | SEARCHING | Values changing 2307 FLASH_REQUIRED | android.control.aeLock is ON | LOCKED | Values locked 2308 LOCKED | android.control.aeLock is OFF | SEARCHING | Values not good after unlock 2309 LOCKED | android.control.aeLock is OFF | CONVERGED | Values good after unlock 2310 LOCKED | android.control.aeLock is OFF | FLASH_REQUIRED | Exposure good, but too dark 2311 PRECAPTURE | Sequence done. android.control.aeLock is OFF | CONVERGED | Ready for high-quality capture 2312 PRECAPTURE | Sequence done. android.control.aeLock is ON | LOCKED | Ready for high-quality capture 2313 LOCKED | aeLock is ON and aePrecaptureTrigger is START | LOCKED | Precapture trigger is ignored when AE is already locked 2314 LOCKED | aeLock is ON and aePrecaptureTrigger is CANCEL| LOCKED | Precapture trigger is ignored when AE is already locked 2315 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is START | PRECAPTURE | Start AE precapture metering sequence 2316 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is CANCEL| INACTIVE | Currently active precapture metering sequence is canceled 2317 2318 For the above table, the camera device may skip reporting any state changes that happen 2319 without application intervention (i.e. mode switch, trigger, locking). Any state that 2320 can be skipped in that manner is called a transient state. 2321 2322 For example, for above AE modes (AE_MODE_ON_*), in addition to the state transitions 2323 listed in above table, it is also legal for the camera device to skip one or more 2324 transient states between two results. See below table for examples: 2325 2326 State | Transition Cause | New State | Notes 2327 :-------------:|:-----------------------------------------------------------:|:--------------:|:-----------------: 2328 INACTIVE | Camera device finished AE scan | CONVERGED | Values are already good, transient states are skipped by camera device. 2329 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is START, sequence done | FLASH_REQUIRED | Converged but too dark w/o flash after a precapture sequence, transient states are skipped by camera device. 2330 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is START, sequence done | CONVERGED | Converged after a precapture sequence, transient states are skipped by camera device. 2331 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is CANCEL, converged | FLASH_REQUIRED | Converged but too dark w/o flash after a precapture sequence is canceled, transient states are skipped by camera device. 2332 Any state (excluding LOCKED) | android.control.aePrecaptureTrigger is CANCEL, converged | CONVERGED | Converged after a precapture sequenceis canceled, transient states are skipped by camera device. 2333 CONVERGED | Camera device finished AE scan | FLASH_REQUIRED | Converged but too dark w/o flash after a new scan, transient states are skipped by camera device. 2334 FLASH_REQUIRED | Camera device finished AE scan | CONVERGED | Converged after a new scan, transient states are skipped by camera device. 2335 </details> 2336 </entry> 2337 <clone entry="android.control.afMode" kind="controls"> 2338 </clone> 2339 <clone entry="android.control.afRegions" kind="controls"> 2340 </clone> 2341 <clone entry="android.control.afTrigger" kind="controls"> 2342 </clone> 2343 <entry name="afState" type="byte" visibility="public" enum="true" 2344 hwlevel="legacy"> 2345 <enum> 2346 <value>INACTIVE 2347 <notes>AF is off or has not yet tried to scan/been asked 2348 to scan. 2349 2350 When a camera device is opened, it starts in this 2351 state. This is a transient state, the camera device may 2352 skip reporting this state in capture 2353 result.</notes></value> 2354 <value>PASSIVE_SCAN 2355 <notes>AF is currently performing an AF scan initiated the 2356 camera device in a continuous autofocus mode. 2357 2358 Only used by CONTINUOUS_* AF modes. This is a transient 2359 state, the camera device may skip reporting this state in 2360 capture result.</notes></value> 2361 <value>PASSIVE_FOCUSED 2362 <notes>AF currently believes it is in focus, but may 2363 restart scanning at any time. 2364 2365 Only used by CONTINUOUS_* AF modes. This is a transient 2366 state, the camera device may skip reporting this state in 2367 capture result.</notes></value> 2368 <value>ACTIVE_SCAN 2369 <notes>AF is performing an AF scan because it was 2370 triggered by AF trigger. 2371 2372 Only used by AUTO or MACRO AF modes. This is a transient 2373 state, the camera device may skip reporting this state in 2374 capture result.</notes></value> 2375 <value>FOCUSED_LOCKED 2376 <notes>AF believes it is focused correctly and has locked 2377 focus. 2378 2379 This state is reached only after an explicit START AF trigger has been 2380 sent (android.control.afTrigger), when good focus has been obtained. 2381 2382 The lens will remain stationary until the AF mode (android.control.afMode) is changed or 2383 a new AF trigger is sent to the camera device (android.control.afTrigger). 2384 </notes></value> 2385 <value>NOT_FOCUSED_LOCKED 2386 <notes>AF has failed to focus successfully and has locked 2387 focus. 2388 2389 This state is reached only after an explicit START AF trigger has been 2390 sent (android.control.afTrigger), when good focus cannot be obtained. 2391 2392 The lens will remain stationary until the AF mode (android.control.afMode) is changed or 2393 a new AF trigger is sent to the camera device (android.control.afTrigger). 2394 </notes></value> 2395 <value>PASSIVE_UNFOCUSED 2396 <notes>AF finished a passive scan without finding focus, 2397 and may restart scanning at any time. 2398 2399 Only used by CONTINUOUS_* AF modes. This is a transient state, the camera 2400 device may skip reporting this state in capture result. 2401 2402 LEGACY camera devices do not support this state. When a passive 2403 scan has finished, it will always go to PASSIVE_FOCUSED. 2404 </notes></value> 2405 </enum> 2406 <description>Current state of auto-focus (AF) algorithm.</description> 2407 <details> 2408 Switching between or enabling AF modes (android.control.afMode) always 2409 resets the AF state to INACTIVE. Similarly, switching between android.control.mode, 2410 or android.control.sceneMode if `android.control.mode == USE_SCENE_MODE` resets all 2411 the algorithm states to INACTIVE. 2412 2413 The camera device can do several state transitions between two results, if it is 2414 allowed by the state transition table. For example: INACTIVE may never actually be 2415 seen in a result. 2416 2417 The state in the result is the state for this image (in sync with this image): if 2418 AF state becomes FOCUSED, then the image data associated with this result should 2419 be sharp. 2420 2421 Below are state transition tables for different AF modes. 2422 2423 When android.control.afMode is AF_MODE_OFF or AF_MODE_EDOF: 2424 2425 State | Transition Cause | New State | Notes 2426 :------------:|:----------------:|:---------:|:-----------: 2427 INACTIVE | | INACTIVE | Never changes 2428 2429 When android.control.afMode is AF_MODE_AUTO or AF_MODE_MACRO: 2430 2431 State | Transition Cause | New State | Notes 2432 :-----------------:|:----------------:|:------------------:|:--------------: 2433 INACTIVE | AF_TRIGGER | ACTIVE_SCAN | Start AF sweep, Lens now moving 2434 ACTIVE_SCAN | AF sweep done | FOCUSED_LOCKED | Focused, Lens now locked 2435 ACTIVE_SCAN | AF sweep done | NOT_FOCUSED_LOCKED | Not focused, Lens now locked 2436 ACTIVE_SCAN | AF_CANCEL | INACTIVE | Cancel/reset AF, Lens now locked 2437 FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Cancel/reset AF 2438 FOCUSED_LOCKED | AF_TRIGGER | ACTIVE_SCAN | Start new sweep, Lens now moving 2439 NOT_FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Cancel/reset AF 2440 NOT_FOCUSED_LOCKED | AF_TRIGGER | ACTIVE_SCAN | Start new sweep, Lens now moving 2441 Any state | Mode change | INACTIVE | 2442 2443 For the above table, the camera device may skip reporting any state changes that happen 2444 without application intervention (i.e. mode switch, trigger, locking). Any state that 2445 can be skipped in that manner is called a transient state. 2446 2447 For example, for these AF modes (AF_MODE_AUTO and AF_MODE_MACRO), in addition to the 2448 state transitions listed in above table, it is also legal for the camera device to skip 2449 one or more transient states between two results. See below table for examples: 2450 2451 State | Transition Cause | New State | Notes 2452 :-----------------:|:----------------:|:------------------:|:--------------: 2453 INACTIVE | AF_TRIGGER | FOCUSED_LOCKED | Focus is already good or good after a scan, lens is now locked. 2454 INACTIVE | AF_TRIGGER | NOT_FOCUSED_LOCKED | Focus failed after a scan, lens is now locked. 2455 FOCUSED_LOCKED | AF_TRIGGER | FOCUSED_LOCKED | Focus is already good or good after a scan, lens is now locked. 2456 NOT_FOCUSED_LOCKED | AF_TRIGGER | FOCUSED_LOCKED | Focus is good after a scan, lens is not locked. 2457 2458 2459 When android.control.afMode is AF_MODE_CONTINUOUS_VIDEO: 2460 2461 State | Transition Cause | New State | Notes 2462 :-----------------:|:-----------------------------------:|:------------------:|:--------------: 2463 INACTIVE | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2464 INACTIVE | AF_TRIGGER | NOT_FOCUSED_LOCKED | AF state query, Lens now locked 2465 PASSIVE_SCAN | Camera device completes current scan| PASSIVE_FOCUSED | End AF scan, Lens now locked 2466 PASSIVE_SCAN | Camera device fails current scan | PASSIVE_UNFOCUSED | End AF scan, Lens now locked 2467 PASSIVE_SCAN | AF_TRIGGER | FOCUSED_LOCKED | Immediate transition, if focus is good. Lens now locked 2468 PASSIVE_SCAN | AF_TRIGGER | NOT_FOCUSED_LOCKED | Immediate transition, if focus is bad. Lens now locked 2469 PASSIVE_SCAN | AF_CANCEL | INACTIVE | Reset lens position, Lens now locked 2470 PASSIVE_FOCUSED | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2471 PASSIVE_UNFOCUSED | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2472 PASSIVE_FOCUSED | AF_TRIGGER | FOCUSED_LOCKED | Immediate transition, lens now locked 2473 PASSIVE_UNFOCUSED | AF_TRIGGER | NOT_FOCUSED_LOCKED | Immediate transition, lens now locked 2474 FOCUSED_LOCKED | AF_TRIGGER | FOCUSED_LOCKED | No effect 2475 FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Restart AF scan 2476 NOT_FOCUSED_LOCKED | AF_TRIGGER | NOT_FOCUSED_LOCKED | No effect 2477 NOT_FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Restart AF scan 2478 2479 When android.control.afMode is AF_MODE_CONTINUOUS_PICTURE: 2480 2481 State | Transition Cause | New State | Notes 2482 :-----------------:|:------------------------------------:|:------------------:|:--------------: 2483 INACTIVE | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2484 INACTIVE | AF_TRIGGER | NOT_FOCUSED_LOCKED | AF state query, Lens now locked 2485 PASSIVE_SCAN | Camera device completes current scan | PASSIVE_FOCUSED | End AF scan, Lens now locked 2486 PASSIVE_SCAN | Camera device fails current scan | PASSIVE_UNFOCUSED | End AF scan, Lens now locked 2487 PASSIVE_SCAN | AF_TRIGGER | FOCUSED_LOCKED | Eventual transition once the focus is good. Lens now locked 2488 PASSIVE_SCAN | AF_TRIGGER | NOT_FOCUSED_LOCKED | Eventual transition if cannot find focus. Lens now locked 2489 PASSIVE_SCAN | AF_CANCEL | INACTIVE | Reset lens position, Lens now locked 2490 PASSIVE_FOCUSED | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2491 PASSIVE_UNFOCUSED | Camera device initiates new scan | PASSIVE_SCAN | Start AF scan, Lens now moving 2492 PASSIVE_FOCUSED | AF_TRIGGER | FOCUSED_LOCKED | Immediate trans. Lens now locked 2493 PASSIVE_UNFOCUSED | AF_TRIGGER | NOT_FOCUSED_LOCKED | Immediate trans. Lens now locked 2494 FOCUSED_LOCKED | AF_TRIGGER | FOCUSED_LOCKED | No effect 2495 FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Restart AF scan 2496 NOT_FOCUSED_LOCKED | AF_TRIGGER | NOT_FOCUSED_LOCKED | No effect 2497 NOT_FOCUSED_LOCKED | AF_CANCEL | INACTIVE | Restart AF scan 2498 2499 When switch between AF_MODE_CONTINUOUS_* (CAF modes) and AF_MODE_AUTO/AF_MODE_MACRO 2500 (AUTO modes), the initial INACTIVE or PASSIVE_SCAN states may be skipped by the 2501 camera device. When a trigger is included in a mode switch request, the trigger 2502 will be evaluated in the context of the new mode in the request. 2503 See below table for examples: 2504 2505 State | Transition Cause | New State | Notes 2506 :-----------:|:--------------------------------------:|:----------------------------------------:|:--------------: 2507 any state | CAF-->AUTO mode switch | INACTIVE | Mode switch without trigger, initial state must be INACTIVE 2508 any state | CAF-->AUTO mode switch with AF_TRIGGER | trigger-reachable states from INACTIVE | Mode switch with trigger, INACTIVE is skipped 2509 any state | AUTO-->CAF mode switch | passively reachable states from INACTIVE | Mode switch without trigger, passive transient state is skipped 2510 </details> 2511 </entry> 2512 <entry name="afTriggerId" type="int32" visibility="system" deprecated="true"> 2513 <description>The ID sent with the latest 2514 CAMERA2_TRIGGER_AUTOFOCUS call</description> 2515 <details>Must be 0 if no CAMERA2_TRIGGER_AUTOFOCUS trigger 2516 received yet by HAL. Always updated even if AF algorithm 2517 ignores the trigger</details> 2518 </entry> 2519 <clone entry="android.control.awbLock" kind="controls"> 2520 </clone> 2521 <clone entry="android.control.awbMode" kind="controls"> 2522 </clone> 2523 <clone entry="android.control.awbRegions" kind="controls"> 2524 </clone> 2525 <clone entry="android.control.captureIntent" kind="controls"> 2526 </clone> 2527 <entry name="awbState" type="byte" visibility="public" enum="true" 2528 hwlevel="limited"> 2529 <enum> 2530 <value>INACTIVE 2531 <notes>AWB is not in auto mode, or has not yet started metering. 2532 2533 When a camera device is opened, it starts in this 2534 state. This is a transient state, the camera device may 2535 skip reporting this state in capture 2536 result.</notes></value> 2537 <value>SEARCHING 2538 <notes>AWB doesn't yet have a good set of control 2539 values for the current scene. 2540 2541 This is a transient state, the camera device 2542 may skip reporting this state in capture result.</notes></value> 2543 <value>CONVERGED 2544 <notes>AWB has a good set of control values for the 2545 current scene.</notes></value> 2546 <value>LOCKED 2547 <notes>AWB has been locked. 2548 </notes></value> 2549 </enum> 2550 <description>Current state of auto-white balance (AWB) algorithm.</description> 2551 <details>Switching between or enabling AWB modes (android.control.awbMode) always 2552 resets the AWB state to INACTIVE. Similarly, switching between android.control.mode, 2553 or android.control.sceneMode if `android.control.mode == USE_SCENE_MODE` resets all 2554 the algorithm states to INACTIVE. 2555 2556 The camera device can do several state transitions between two results, if it is 2557 allowed by the state transition table. So INACTIVE may never actually be seen in 2558 a result. 2559 2560 The state in the result is the state for this image (in sync with this image): if 2561 AWB state becomes CONVERGED, then the image data associated with this result should 2562 be good to use. 2563 2564 Below are state transition tables for different AWB modes. 2565 2566 When `android.control.awbMode != AWB_MODE_AUTO`: 2567 2568 State | Transition Cause | New State | Notes 2569 :------------:|:----------------:|:---------:|:-----------------------: 2570 INACTIVE | |INACTIVE |Camera device auto white balance algorithm is disabled 2571 2572 When android.control.awbMode is AWB_MODE_AUTO: 2573 2574 State | Transition Cause | New State | Notes 2575 :-------------:|:--------------------------------:|:-------------:|:-----------------: 2576 INACTIVE | Camera device initiates AWB scan | SEARCHING | Values changing 2577 INACTIVE | android.control.awbLock is ON | LOCKED | Values locked 2578 SEARCHING | Camera device finishes AWB scan | CONVERGED | Good values, not changing 2579 SEARCHING | android.control.awbLock is ON | LOCKED | Values locked 2580 CONVERGED | Camera device initiates AWB scan | SEARCHING | Values changing 2581 CONVERGED | android.control.awbLock is ON | LOCKED | Values locked 2582 LOCKED | android.control.awbLock is OFF | SEARCHING | Values not good after unlock 2583 2584 For the above table, the camera device may skip reporting any state changes that happen 2585 without application intervention (i.e. mode switch, trigger, locking). Any state that 2586 can be skipped in that manner is called a transient state. 2587 2588 For example, for this AWB mode (AWB_MODE_AUTO), in addition to the state transitions 2589 listed in above table, it is also legal for the camera device to skip one or more 2590 transient states between two results. See below table for examples: 2591 2592 State | Transition Cause | New State | Notes 2593 :-------------:|:--------------------------------:|:-------------:|:-----------------: 2594 INACTIVE | Camera device finished AWB scan | CONVERGED | Values are already good, transient states are skipped by camera device. 2595 LOCKED | android.control.awbLock is OFF | CONVERGED | Values good after unlock, transient states are skipped by camera device. 2596 </details> 2597 </entry> 2598 <clone entry="android.control.effectMode" kind="controls"> 2599 </clone> 2600 <clone entry="android.control.mode" kind="controls"> 2601 </clone> 2602 <clone entry="android.control.sceneMode" kind="controls"> 2603 </clone> 2604 <clone entry="android.control.videoStabilizationMode" kind="controls"> 2605 </clone> 2606 </dynamic> 2607 <static> 2608 <entry name="availableHighSpeedVideoConfigurations" type="int32" visibility="hidden" 2609 container="array" typedef="highSpeedVideoConfiguration" hwlevel="limited"> 2610 <array> 2611 <size>5</size> 2612 <size>n</size> 2613 </array> 2614 <description> 2615 List of available high speed video size, fps range and max batch size configurations 2616 supported by the camera device, in the format of (width, height, fps_min, fps_max, batch_size_max). 2617 </description> 2618 <range> 2619 For each configuration, the fps_max &gt;= 120fps. 2620 </range> 2621 <details> 2622 When CONSTRAINED_HIGH_SPEED_VIDEO is supported in android.request.availableCapabilities, 2623 this metadata will list the supported high speed video size, fps range and max batch size 2624 configurations. All the sizes listed in this configuration will be a subset of the sizes 2625 reported by {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputSizes} 2626 for processed non-stalling formats. 2627 2628 For the high speed video use case, the application must 2629 select the video size and fps range from this metadata to configure the recording and 2630 preview streams and setup the recording requests. For example, if the application intends 2631 to do high speed recording, it can select the maximum size reported by this metadata to 2632 configure output streams. Once the size is selected, application can filter this metadata 2633 by selected size and get the supported fps ranges, and use these fps ranges to setup the 2634 recording requests. Note that for the use case of multiple output streams, application 2635 must select one unique size from this metadata to use (e.g., preview and recording streams 2636 must have the same size). Otherwise, the high speed capture session creation will fail. 2637 2638 The min and max fps will be multiple times of 30fps. 2639 2640 High speed video streaming extends significant performance pressue to camera hardware, 2641 to achieve efficient high speed streaming, the camera device may have to aggregate 2642 multiple frames together and send to camera device for processing where the request 2643 controls are same for all the frames in this batch. Max batch size indicates 2644 the max possible number of frames the camera device will group together for this high 2645 speed stream configuration. This max batch size will be used to generate a high speed 2646 recording request list by 2647 {@link android.hardware.camera2.CameraConstrainedHighSpeedCaptureSession#createHighSpeedRequestList}. 2648 The max batch size for each configuration will satisfy below conditions: 2649 2650 * Each max batch size will be a divisor of its corresponding fps_max / 30. For example, 2651 if max_fps is 300, max batch size will only be 1, 2, 5, or 10. 2652 * The camera device may choose smaller internal batch size for each configuration, but 2653 the actual batch size will be a divisor of max batch size. For example, if the max batch 2654 size is 8, the actual batch size used by camera device will only be 1, 2, 4, or 8. 2655 * The max batch size in each configuration entry must be no larger than 32. 2656 2657 The camera device doesn't have to support batch mode to achieve high speed video recording, 2658 in such case, batch_size_max will be reported as 1 in each configuration entry. 2659 2660 This fps ranges in this configuration list can only be used to create requests 2661 that are submitted to a high speed camera capture session created by 2662 {@link android.hardware.camera2.CameraDevice#createConstrainedHighSpeedCaptureSession}. 2663 The fps ranges reported in this metadata must not be used to setup capture requests for 2664 normal capture session, or it will cause request error. 2665 </details> 2666 <hal_details> 2667 All the sizes listed in this configuration will be a subset of the sizes reported by 2668 android.scaler.availableStreamConfigurations for processed non-stalling output formats. 2669 Note that for all high speed video configurations, HAL must be able to support a minimum 2670 of two streams, though the application might choose to configure just one stream. 2671 2672 The HAL may support multiple sensor modes for high speed outputs, for example, 120fps 2673 sensor mode and 120fps recording, 240fps sensor mode for 240fps recording. The application 2674 usually starts preview first, then starts recording. To avoid sensor mode switch caused 2675 stutter when starting recording as much as possible, the application may want to ensure 2676 the same sensor mode is used for preview and recording. Therefore, The HAL must advertise 2677 the variable fps range [30, fps_max] for each fixed fps range in this configuration list. 2678 For example, if the HAL advertises [120, 120] and [240, 240], the HAL must also advertise 2679 [30, 120] and [30, 240] for each configuration. In doing so, if the application intends to 2680 do 120fps recording, it can select [30, 120] to start preview, and [120, 120] to start 2681 recording. For these variable fps ranges, it's up to the HAL to decide the actual fps 2682 values that are suitable for smooth preview streaming. If the HAL sees different max_fps 2683 values that fall into different sensor modes in a sequence of requests, the HAL must 2684 switch the sensor mode as quick as possible to minimize the mode switch caused stutter. 2685 </hal_details> 2686 <tag id="V1" /> 2687 </entry> 2688 <entry name="aeLockAvailable" type="byte" visibility="public" enum="true" 2689 typedef="boolean" hwlevel="legacy"> 2690 <enum> 2691 <value>FALSE</value> 2692 <value>TRUE</value> 2693 </enum> 2694 <description>Whether the camera device supports android.control.aeLock</description> 2695 <details> 2696 Devices with MANUAL_SENSOR capability or BURST_CAPTURE capability will always 2697 list `true`. This includes FULL devices. 2698 </details> 2699 <tag id="BC"/> 2700 </entry> 2701 <entry name="awbLockAvailable" type="byte" visibility="public" enum="true" 2702 typedef="boolean" hwlevel="legacy"> 2703 <enum> 2704 <value>FALSE</value> 2705 <value>TRUE</value> 2706 </enum> 2707 <description>Whether the camera device supports android.control.awbLock</description> 2708 <details> 2709 Devices with MANUAL_POST_PROCESSING capability or BURST_CAPTURE capability will 2710 always list `true`. This includes FULL devices. 2711 </details> 2712 <tag id="BC"/> 2713 </entry> 2714 <entry name="availableModes" type="byte" visibility="public" 2715 type_notes="List of enums (android.control.mode)." container="array" 2716 typedef="enumList" hwlevel="legacy"> 2717 <array> 2718 <size>n</size> 2719 </array> 2720 <description> 2721 List of control modes for android.control.mode that are supported by this camera 2722 device. 2723 </description> 2724 <range>Any value listed in android.control.mode</range> 2725 <details> 2726 This list contains control modes that can be set for the camera device. 2727 LEGACY mode devices will always support AUTO mode. LIMITED and FULL 2728 devices will always support OFF, AUTO modes. 2729 </details> 2730 </entry> 2731 <entry name="postRawSensitivityBoostRange" type="int32" visibility="public" 2732 type_notes="Range of supported post RAW sensitivitiy boosts" 2733 container="array" typedef="rangeInt"> 2734 <array> 2735 <size>2</size> 2736 </array> 2737 <description>Range of boosts for android.control.postRawSensitivityBoost supported 2738 by this camera device. 2739 </description> 2740 <units>ISO arithmetic units, the same as android.sensor.sensitivity</units> 2741 <details> 2742 Devices support post RAW sensitivity boost will advertise 2743 android.control.postRawSensitivityBoost key for controling 2744 post RAW sensitivity boost. 2745 2746 This key will be `null` for devices that do not support any RAW format 2747 outputs. For devices that do support RAW format outputs, this key will always 2748 present, and if a device does not support post RAW sensitivity boost, it will 2749 list `(100, 100)` in this key. 2750 </details> 2751 <hal_details> 2752 This key is added in HAL3.4. For HAL3.3 or earlier devices, camera framework will 2753 generate this key as `(100, 100)` if device supports any of RAW output formats. 2754 All HAL3.4 and above devices should list this key if device supports any of RAW 2755 output formats. 2756 </hal_details> 2757 </entry> 2758 </static> 2759 <controls> 2760 <entry name="postRawSensitivityBoost" type="int32" visibility="public"> 2761 <description>The amount of additional sensitivity boost applied to output images 2762 after RAW sensor data is captured. 2763 </description> 2764 <units>ISO arithmetic units, the same as android.sensor.sensitivity</units> 2765 <range>android.control.postRawSensitivityBoostRange</range> 2766 <details> 2767 Some camera devices support additional digital sensitivity boosting in the 2768 camera processing pipeline after sensor RAW image is captured. 2769 Such a boost will be applied to YUV/JPEG format output images but will not 2770 have effect on RAW output formats like RAW_SENSOR, RAW10, RAW12 or RAW_OPAQUE. 2771 2772 This key will be `null` for devices that do not support any RAW format 2773 outputs. For devices that do support RAW format outputs, this key will always 2774 present, and if a device does not support post RAW sensitivity boost, it will 2775 list `100` in this key. 2776 2777 If the camera device cannot apply the exact boost requested, it will reduce the 2778 boost to the nearest supported value. 2779 The final boost value used will be available in the output capture result. 2780 2781 For devices that support post RAW sensitivity boost, the YUV/JPEG output images 2782 of such device will have the total sensitivity of 2783 `android.sensor.sensitivity * android.control.postRawSensitivityBoost / 100` 2784 The sensitivity of RAW format images will always be `android.sensor.sensitivity` 2785 2786 This control is only effective if android.control.aeMode or android.control.mode is set to 2787 OFF; otherwise the auto-exposure algorithm will override this value. 2788 </details> 2789 </entry> 2790 </controls> 2791 <dynamic> 2792 <clone entry="android.control.postRawSensitivityBoost" kind="controls"> 2793 </clone> 2794 </dynamic> 2795 </section> 2796 <section name="demosaic"> 2797 <controls> 2798 <entry name="mode" type="byte" enum="true"> 2799 <enum> 2800 <value>FAST 2801 <notes>Minimal or no slowdown of frame rate compared to 2802 Bayer RAW output.</notes></value> 2803 <value>HIGH_QUALITY 2804 <notes>Improved processing quality but the frame rate might be slowed down 2805 relative to raw output.</notes></value> 2806 </enum> 2807 <description>Controls the quality of the demosaicing 2808 processing.</description> 2809 <tag id="FUTURE" /> 2810 </entry> 2811 </controls> 2812 </section> 2813 <section name="edge"> 2814 <controls> 2815 <entry name="mode" type="byte" visibility="public" enum="true" hwlevel="full"> 2816 <enum> 2817 <value>OFF 2818 <notes>No edge enhancement is applied.</notes></value> 2819 <value>FAST 2820 <notes>Apply edge enhancement at a quality level that does not slow down frame rate 2821 relative to sensor output. It may be the same as OFF if edge enhancement will 2822 slow down frame rate relative to sensor.</notes></value> 2823 <value>HIGH_QUALITY 2824 <notes>Apply high-quality edge enhancement, at a cost of possibly reduced output frame rate. 2825 </notes></value> 2826 <value optional="true">ZERO_SHUTTER_LAG 2827 <notes>Edge enhancement is applied at different levels for different output streams, 2828 based on resolution. Streams at maximum recording resolution (see {@link 2829 android.hardware.camera2.CameraDevice#createCaptureSession}) or below have 2830 edge enhancement applied, while higher-resolution streams have no edge enhancement 2831 applied. The level of edge enhancement for low-resolution streams is tuned so that 2832 frame rate is not impacted, and the quality is equal to or better than FAST (since it 2833 is only applied to lower-resolution outputs, quality may improve from FAST). 2834 2835 This mode is intended to be used by applications operating in a zero-shutter-lag mode 2836 with YUV or PRIVATE reprocessing, where the application continuously captures 2837 high-resolution intermediate buffers into a circular buffer, from which a final image is 2838 produced via reprocessing when a user takes a picture. For such a use case, the 2839 high-resolution buffers must not have edge enhancement applied to maximize efficiency of 2840 preview and to avoid double-applying enhancement when reprocessed, while low-resolution 2841 buffers (used for recording or preview, generally) need edge enhancement applied for 2842 reasonable preview quality. 2843 2844 This mode is guaranteed to be supported by devices that support either the 2845 YUV_REPROCESSING or PRIVATE_REPROCESSING capabilities 2846 (android.request.availableCapabilities lists either of those capabilities) and it will 2847 be the default mode for CAMERA3_TEMPLATE_ZERO_SHUTTER_LAG template. 2848 </notes></value> 2849 </enum> 2850 <description>Operation mode for edge 2851 enhancement.</description> 2852 <range>android.edge.availableEdgeModes</range> 2853 <details>Edge enhancement improves sharpness and details in the captured image. OFF means 2854 no enhancement will be applied by the camera device. 2855 2856 FAST/HIGH_QUALITY both mean camera device determined enhancement 2857 will be applied. HIGH_QUALITY mode indicates that the 2858 camera device will use the highest-quality enhancement algorithms, 2859 even if it slows down capture rate. FAST means the camera device will 2860 not slow down capture rate when applying edge enhancement. FAST may be the same as OFF if 2861 edge enhancement will slow down capture rate. Every output stream will have a similar 2862 amount of enhancement applied. 2863 2864 ZERO_SHUTTER_LAG is meant to be used by applications that maintain a continuous circular 2865 buffer of high-resolution images during preview and reprocess image(s) from that buffer 2866 into a final capture when triggered by the user. In this mode, the camera device applies 2867 edge enhancement to low-resolution streams (below maximum recording resolution) to 2868 maximize preview quality, but does not apply edge enhancement to high-resolution streams, 2869 since those will be reprocessed later if necessary. 2870 2871 For YUV_REPROCESSING, these FAST/HIGH_QUALITY modes both mean that the camera 2872 device will apply FAST/HIGH_QUALITY YUV-domain edge enhancement, respectively. 2873 The camera device may adjust its internal edge enhancement parameters for best 2874 image quality based on the android.reprocess.effectiveExposureFactor, if it is set. 2875 </details> 2876 <hal_details> 2877 For YUV_REPROCESSING The HAL can use android.reprocess.effectiveExposureFactor to 2878 adjust the internal edge enhancement reduction parameters appropriately to get the best 2879 quality images. 2880 </hal_details> 2881 <tag id="V1" /> 2882 <tag id="REPROC" /> 2883 </entry> 2884 <entry name="strength" type="byte"> 2885 <description>Control the amount of edge enhancement 2886 applied to the images</description> 2887 <units>1-10; 10 is maximum sharpening</units> 2888 <tag id="FUTURE" /> 2889 </entry> 2890 </controls> 2891 <static> 2892 <entry name="availableEdgeModes" type="byte" visibility="public" 2893 type_notes="list of enums" container="array" typedef="enumList" 2894 hwlevel="full"> 2895 <array> 2896 <size>n</size> 2897 </array> 2898 <description> 2899 List of edge enhancement modes for android.edge.mode that are supported by this camera 2900 device. 2901 </description> 2902 <range>Any value listed in android.edge.mode</range> 2903 <details> 2904 Full-capability camera devices must always support OFF; camera devices that support 2905 YUV_REPROCESSING or PRIVATE_REPROCESSING will list ZERO_SHUTTER_LAG; all devices will 2906 list FAST. 2907 </details> 2908 <hal_details> 2909 HAL must support both FAST and HIGH_QUALITY if edge enhancement control is available 2910 on the camera device, but the underlying implementation can be the same for both modes. 2911 That is, if the highest quality implementation on the camera device does not slow down 2912 capture rate, then FAST and HIGH_QUALITY will generate the same output. 2913 </hal_details> 2914 <tag id="V1" /> 2915 <tag id="REPROC" /> 2916 </entry> 2917 </static> 2918 <dynamic> 2919 <clone entry="android.edge.mode" kind="controls"> 2920 <tag id="V1" /> 2921 <tag id="REPROC" /> 2922 </clone> 2923 </dynamic> 2924 </section> 2925 <section name="flash"> 2926 <controls> 2927 <entry name="firingPower" type="byte"> 2928 <description>Power for flash firing/torch</description> 2929 <units>10 is max power; 0 is no flash. Linear</units> 2930 <range>0 - 10</range> 2931 <details>Power for snapshot may use a different scale than 2932 for torch mode. Only one entry for torch mode will be 2933 used</details> 2934 <tag id="FUTURE" /> 2935 </entry> 2936 <entry name="firingTime" type="int64"> 2937 <description>Firing time of flash relative to start of 2938 exposure</description> 2939 <units>nanoseconds</units> 2940 <range>0-(exposure time-flash duration)</range> 2941 <details>Clamped to (0, exposure time - flash 2942 duration).</details> 2943 <tag id="FUTURE" /> 2944 </entry> 2945 <entry name="mode" type="byte" visibility="public" enum="true" hwlevel="legacy"> 2946 <enum> 2947 <value>OFF 2948 <notes> 2949 Do not fire the flash for this capture. 2950 </notes> 2951 </value> 2952 <value>SINGLE 2953 <notes> 2954 If the flash is available and charged, fire flash 2955 for this capture. 2956 </notes> 2957 </value> 2958 <value>TORCH 2959 <notes> 2960 Transition flash to continuously on. 2961 </notes> 2962 </value> 2963 </enum> 2964 <description>The desired mode for for the camera device's flash control.</description> 2965 <details> 2966 This control is only effective when flash unit is available 2967 (`android.flash.info.available == true`). 2968 2969 When this control is used, the android.control.aeMode must be set to ON or OFF. 2970 Otherwise, the camera device auto-exposure related flash control (ON_AUTO_FLASH, 2971 ON_ALWAYS_FLASH, or ON_AUTO_FLASH_REDEYE) will override this control. 2972 2973 When set to OFF, the camera device will not fire flash for this capture. 2974 2975 When set to SINGLE, the camera device will fire flash regardless of the camera 2976 device's auto-exposure routine's result. When used in still capture case, this 2977 control should be used along with auto-exposure (AE) precapture metering sequence 2978 (android.control.aePrecaptureTrigger), otherwise, the image may be incorrectly exposed. 2979 2980 When set to TORCH, the flash will be on continuously. This mode can be used 2981 for use cases such as preview, auto-focus assist, still capture, or video recording. 2982 2983 The flash status will be reported by android.flash.state in the capture result metadata. 2984 </details> 2985 <tag id="BC" /> 2986 </entry> 2987 </controls> 2988 <static> 2989 <namespace name="info"> 2990 <entry name="available" type="byte" visibility="public" enum="true" 2991 typedef="boolean" hwlevel="legacy"> 2992 <enum> 2993 <value>FALSE</value> 2994 <value>TRUE</value> 2995 </enum> 2996 <description>Whether this camera device has a 2997 flash unit.</description> 2998 <details> 2999 Will be `false` if no flash is available. 3000 3001 If there is no flash unit, none of the flash controls do 3002 anything.</details> 3003 <tag id="BC" /> 3004 </entry> 3005 <entry name="chargeDuration" type="int64"> 3006 <description>Time taken before flash can fire 3007 again</description> 3008 <units>nanoseconds</units> 3009 <range>0-1e9</range> 3010 <details>1 second too long/too short for recharge? Should 3011 this be power-dependent?</details> 3012 <tag id="FUTURE" /> 3013 </entry> 3014 </namespace> 3015 <entry name="colorTemperature" type="byte"> 3016 <description>The x,y whitepoint of the 3017 flash</description> 3018 <units>pair of floats</units> 3019 <range>0-1 for both</range> 3020 <tag id="FUTURE" /> 3021 </entry> 3022 <entry name="maxEnergy" type="byte"> 3023 <description>Max energy output of the flash for a full 3024 power single flash</description> 3025 <units>lumen-seconds</units> 3026 <range>&gt;= 0</range> 3027 <tag id="FUTURE" /> 3028 </entry> 3029 </static> 3030 <dynamic> 3031 <clone entry="android.flash.firingPower" kind="controls"> 3032 </clone> 3033 <clone entry="android.flash.firingTime" kind="controls"> 3034 </clone> 3035 <clone entry="android.flash.mode" kind="controls"></clone> 3036 <entry name="state" type="byte" visibility="public" enum="true" 3037 hwlevel="limited"> 3038 <enum> 3039 <value>UNAVAILABLE 3040 <notes>No flash on camera.</notes></value> 3041 <value>CHARGING 3042 <notes>Flash is charging and cannot be fired.</notes></value> 3043 <value>READY 3044 <notes>Flash is ready to fire.</notes></value> 3045 <value>FIRED 3046 <notes>Flash fired for this capture.</notes></value> 3047 <value>PARTIAL 3048 <notes>Flash partially illuminated this frame. 3049 3050 This is usually due to the next or previous frame having 3051 the flash fire, and the flash spilling into this capture 3052 due to hardware limitations.</notes></value> 3053 </enum> 3054 <description>Current state of the flash 3055 unit.</description> 3056 <details> 3057 When the camera device doesn't have flash unit 3058 (i.e. `android.flash.info.available == false`), this state will always be UNAVAILABLE. 3059 Other states indicate the current flash status. 3060 3061 In certain conditions, this will be available on LEGACY devices: 3062 3063 * Flash-less cameras always return UNAVAILABLE. 3064 * Using android.control.aeMode `==` ON_ALWAYS_FLASH 3065 will always return FIRED. 3066 * Using android.flash.mode `==` TORCH 3067 will always return FIRED. 3068 3069 In all other conditions the state will not be available on 3070 LEGACY devices (i.e. it will be `null`). 3071 </details> 3072 </entry> 3073 </dynamic> 3074 </section> 3075 <section name="hotPixel"> 3076 <controls> 3077 <entry name="mode" type="byte" visibility="public" enum="true"> 3078 <enum> 3079 <value>OFF 3080 <notes> 3081 No hot pixel correction is applied. 3082 3083 The frame rate must not be reduced relative to sensor raw output 3084 for this option. 3085 3086 The hotpixel map may be returned in android.statistics.hotPixelMap. 3087 </notes> 3088 </value> 3089 <value>FAST 3090 <notes> 3091 Hot pixel correction is applied, without reducing frame 3092 rate relative to sensor raw output. 3093 3094 The hotpixel map may be returned in android.statistics.hotPixelMap. 3095 </notes> 3096 </value> 3097 <value>HIGH_QUALITY 3098 <notes> 3099 High-quality hot pixel correction is applied, at a cost 3100 of possibly reduced frame rate relative to sensor raw output. 3101 3102 The hotpixel map may be returned in android.statistics.hotPixelMap. 3103 </notes> 3104 </value> 3105 </enum> 3106 <description> 3107 Operational mode for hot pixel correction. 3108 </description> 3109 <range>android.hotPixel.availableHotPixelModes</range> 3110 <details> 3111 Hotpixel correction interpolates out, or otherwise removes, pixels 3112 that do not accurately measure the incoming light (i.e. pixels that 3113 are stuck at an arbitrary value or are oversensitive). 3114 </details> 3115 <tag id="V1" /> 3116 <tag id="RAW" /> 3117 </entry> 3118 </controls> 3119 <static> 3120 <entry name="availableHotPixelModes" type="byte" visibility="public" 3121 type_notes="list of enums" container="array" typedef="enumList"> 3122 <array> 3123 <size>n</size> 3124 </array> 3125 <description> 3126 List of hot pixel correction modes for android.hotPixel.mode that are supported by this 3127 camera device. 3128 </description> 3129 <range>Any value listed in android.hotPixel.mode</range> 3130 <details> 3131 FULL mode camera devices will always support FAST. 3132 </details> 3133 <hal_details> 3134 To avoid performance issues, there will be significantly fewer hot 3135 pixels than actual pixels on the camera sensor. 3136 HAL must support both FAST and HIGH_QUALITY if hot pixel correction control is available 3137 on the camera device, but the underlying implementation can be the same for both modes. 3138 That is, if the highest quality implementation on the camera device does not slow down 3139 capture rate, then FAST and HIGH_QUALITY will generate the same output. 3140 </hal_details> 3141 <tag id="V1" /> 3142 <tag id="RAW" /> 3143 </entry> 3144 </static> 3145 <dynamic> 3146 <clone entry="android.hotPixel.mode" kind="controls"> 3147 <tag id="V1" /> 3148 <tag id="RAW" /> 3149 </clone> 3150 </dynamic> 3151 </section> 3152 <section name="jpeg"> 3153 <controls> 3154 <entry name="gpsLocation" type="byte" visibility="java_public" synthetic="true" 3155 typedef="location" hwlevel="legacy"> 3156 <description> 3157 A location object to use when generating image GPS metadata. 3158 </description> 3159 <details> 3160 Setting a location object in a request will include the GPS coordinates of the location 3161 into any JPEG images captured based on the request. These coordinates can then be 3162 viewed by anyone who receives the JPEG image. 3163 </details> 3164 </entry> 3165 <entry name="gpsCoordinates" type="double" visibility="ndk_public" 3166 type_notes="latitude, longitude, altitude. First two in degrees, the third in meters" 3167 container="array" hwlevel="legacy"> 3168 <array> 3169 <size>3</size> 3170 </array> 3171 <description>GPS coordinates to include in output JPEG 3172 EXIF.</description> 3173 <range>(-180 - 180], [-90,90], [-inf, inf]</range> 3174 <tag id="BC" /> 3175 </entry> 3176 <entry name="gpsProcessingMethod" type="byte" visibility="ndk_public" 3177 typedef="string" hwlevel="legacy"> 3178 <description>32 characters describing GPS algorithm to 3179 include in EXIF.</description> 3180 <units>UTF-8 null-terminated string</units> 3181 <tag id="BC" /> 3182 </entry> 3183 <entry name="gpsTimestamp" type="int64" visibility="ndk_public" hwlevel="legacy"> 3184 <description>Time GPS fix was made to include in 3185 EXIF.</description> 3186 <units>UTC in seconds since January 1, 1970</units> 3187 <tag id="BC" /> 3188 </entry> 3189 <entry name="orientation" type="int32" visibility="public" hwlevel="legacy"> 3190 <description>The orientation for a JPEG image.</description> 3191 <units>Degrees in multiples of 90</units> 3192 <range>0, 90, 180, 270</range> 3193 <details> 3194 The clockwise rotation angle in degrees, relative to the orientation 3195 to the camera, that the JPEG picture needs to be rotated by, to be viewed 3196 upright. 3197 3198 Camera devices may either encode this value into the JPEG EXIF header, or 3199 rotate the image data to match this orientation. When the image data is rotated, 3200 the thumbnail data will also be rotated. 3201 3202 Note that this orientation is relative to the orientation of the camera sensor, given 3203 by android.sensor.orientation. 3204 3205 To translate from the device orientation given by the Android sensor APIs, the following 3206 sample code may be used: 3207 3208 private int getJpegOrientation(CameraCharacteristics c, int deviceOrientation) { 3209 if (deviceOrientation == android.view.OrientationEventListener.ORIENTATION_UNKNOWN) return 0; 3210 int sensorOrientation = c.get(CameraCharacteristics.SENSOR_ORIENTATION); 3211 3212 // Round device orientation to a multiple of 90 3213 deviceOrientation = (deviceOrientation + 45) / 90 * 90; 3214 3215 // Reverse device orientation for front-facing cameras 3216 boolean facingFront = c.get(CameraCharacteristics.LENS_FACING) == CameraCharacteristics.LENS_FACING_FRONT; 3217 if (facingFront) deviceOrientation = -deviceOrientation; 3218 3219 // Calculate desired JPEG orientation relative to camera orientation to make 3220 // the image upright relative to the device orientation 3221 int jpegOrientation = (sensorOrientation + deviceOrientation + 360) % 360; 3222 3223 return jpegOrientation; 3224 } 3225 </details> 3226 <tag id="BC" /> 3227 </entry> 3228 <entry name="quality" type="byte" visibility="public" hwlevel="legacy"> 3229 <description>Compression quality of the final JPEG 3230 image.</description> 3231 <range>1-100; larger is higher quality</range> 3232 <details>85-95 is typical usage range.</details> 3233 <tag id="BC" /> 3234 </entry> 3235 <entry name="thumbnailQuality" type="byte" visibility="public" hwlevel="legacy"> 3236 <description>Compression quality of JPEG 3237 thumbnail.</description> 3238 <range>1-100; larger is higher quality</range> 3239 <tag id="BC" /> 3240 </entry> 3241 <entry name="thumbnailSize" type="int32" visibility="public" 3242 container="array" typedef="size" hwlevel="legacy"> 3243 <array> 3244 <size>2</size> 3245 </array> 3246 <description>Resolution of embedded JPEG thumbnail.</description> 3247 <range>android.jpeg.availableThumbnailSizes</range> 3248 <details>When set to (0, 0) value, the JPEG EXIF will not contain thumbnail, 3249 but the captured JPEG will still be a valid image. 3250 3251 For best results, when issuing a request for a JPEG image, the thumbnail size selected 3252 should have the same aspect ratio as the main JPEG output. 3253 3254 If the thumbnail image aspect ratio differs from the JPEG primary image aspect 3255 ratio, the camera device creates the thumbnail by cropping it from the primary image. 3256 For example, if the primary image has 4:3 aspect ratio, the thumbnail image has 3257 16:9 aspect ratio, the primary image will be cropped vertically (letterbox) to 3258 generate the thumbnail image. The thumbnail image will always have a smaller Field 3259 Of View (FOV) than the primary image when aspect ratios differ. 3260 3261 When an android.jpeg.orientation of non-zero degree is requested, 3262 the camera device will handle thumbnail rotation in one of the following ways: 3263 3264 * Set the {@link android.media.ExifInterface#TAG_ORIENTATION EXIF orientation flag} 3265 and keep jpeg and thumbnail image data unrotated. 3266 * Rotate the jpeg and thumbnail image data and not set 3267 {@link android.media.ExifInterface#TAG_ORIENTATION EXIF orientation flag}. In this 3268 case, LIMITED or FULL hardware level devices will report rotated thumnail size in 3269 capture result, so the width and height will be interchanged if 90 or 270 degree 3270 orientation is requested. LEGACY device will always report unrotated thumbnail 3271 size. 3272 </details> 3273 <hal_details> 3274 The HAL must not squeeze or stretch the downscaled primary image to generate thumbnail. 3275 The cropping must be done on the primary jpeg image rather than the sensor active array. 3276 The stream cropping rule specified by "S5. Cropping" in camera3.h doesn't apply to the 3277 thumbnail image cropping. 3278 </hal_details> 3279 <tag id="BC" /> 3280 </entry> 3281 </controls> 3282 <static> 3283 <entry name="availableThumbnailSizes" type="int32" visibility="public" 3284 container="array" typedef="size" hwlevel="legacy"> 3285 <array> 3286 <size>2</size> 3287 <size>n</size> 3288 </array> 3289 <description>List of JPEG thumbnail sizes for android.jpeg.thumbnailSize supported by this 3290 camera device.</description> 3291 <details> 3292 This list will include at least one non-zero resolution, plus `(0,0)` for indicating no 3293 thumbnail should be generated. 3294 3295 Below condiditions will be satisfied for this size list: 3296 3297 * The sizes will be sorted by increasing pixel area (width x height). 3298 If several resolutions have the same area, they will be sorted by increasing width. 3299 * The aspect ratio of the largest thumbnail size will be same as the 3300 aspect ratio of largest JPEG output size in android.scaler.availableStreamConfigurations. 3301 The largest size is defined as the size that has the largest pixel area 3302 in a given size list. 3303 * Each output JPEG size in android.scaler.availableStreamConfigurations will have at least 3304 one corresponding size that has the same aspect ratio in availableThumbnailSizes, 3305 and vice versa. 3306 * All non-`(0, 0)` sizes will have non-zero widths and heights.</details> 3307 <tag id="BC" /> 3308 </entry> 3309 <entry name="maxSize" type="int32" visibility="system"> 3310 <description>Maximum size in bytes for the compressed 3311 JPEG buffer</description> 3312 <range>Must be large enough to fit any JPEG produced by 3313 the camera</range> 3314 <details>This is used for sizing the gralloc buffers for 3315 JPEG</details> 3316 </entry> 3317 </static> 3318 <dynamic> 3319 <clone entry="android.jpeg.gpsLocation" kind="controls"> 3320 </clone> 3321 <clone entry="android.jpeg.gpsCoordinates" kind="controls"> 3322 </clone> 3323 <clone entry="android.jpeg.gpsProcessingMethod" 3324 kind="controls"></clone> 3325 <clone entry="android.jpeg.gpsTimestamp" kind="controls"> 3326 </clone> 3327 <clone entry="android.jpeg.orientation" kind="controls"> 3328 </clone> 3329 <clone entry="android.jpeg.quality" kind="controls"> 3330 </clone> 3331 <entry name="size" type="int32"> 3332 <description>The size of the compressed JPEG image, in 3333 bytes</description> 3334 <range>&gt;= 0</range> 3335 <details>If no JPEG output is produced for the request, 3336 this must be 0. 3337 3338 Otherwise, this describes the real size of the compressed 3339 JPEG image placed in the output stream. More specifically, 3340 if android.jpeg.maxSize = 1000000, and a specific capture 3341 has android.jpeg.size = 500000, then the output buffer from 3342 the JPEG stream will be 1000000 bytes, of which the first 3343 500000 make up the real data.</details> 3344 <tag id="FUTURE" /> 3345 </entry> 3346 <clone entry="android.jpeg.thumbnailQuality" 3347 kind="controls"></clone> 3348 <clone entry="android.jpeg.thumbnailSize" kind="controls"> 3349 </clone> 3350 </dynamic> 3351 </section> 3352 <section name="lens"> 3353 <controls> 3354 <entry name="aperture" type="float" visibility="public" hwlevel="full"> 3355 <description>The desired lens aperture size, as a ratio of lens focal length to the 3356 effective aperture diameter.</description> 3357 <units>The f-number (f/N)</units> 3358 <range>android.lens.info.availableApertures</range> 3359 <details>Setting this value is only supported on the camera devices that have a variable 3360 aperture lens. 3361 3362 When this is supported and android.control.aeMode is OFF, 3363 this can be set along with android.sensor.exposureTime, 3364 android.sensor.sensitivity, and android.sensor.frameDuration 3365 to achieve manual exposure control. 3366 3367 The requested aperture value may take several frames to reach the 3368 requested value; the camera device will report the current (intermediate) 3369 aperture size in capture result metadata while the aperture is changing. 3370 While the aperture is still changing, android.lens.state will be set to MOVING. 3371 3372 When this is supported and android.control.aeMode is one of 3373 the ON modes, this will be overridden by the camera device 3374 auto-exposure algorithm, the overridden values are then provided 3375 back to the user in the corresponding result.</details> 3376 <tag id="V1" /> 3377 </entry> 3378 <entry name="filterDensity" type="float" visibility="public" hwlevel="full"> 3379 <description> 3380 The desired setting for the lens neutral density filter(s). 3381 </description> 3382 <units>Exposure Value (EV)</units> 3383 <range>android.lens.info.availableFilterDensities</range> 3384 <details> 3385 This control will not be supported on most camera devices. 3386 3387 Lens filters are typically used to lower the amount of light the 3388 sensor is exposed to (measured in steps of EV). As used here, an EV 3389 step is the standard logarithmic representation, which are 3390 non-negative, and inversely proportional to the amount of light 3391 hitting the sensor. For example, setting this to 0 would result 3392 in no reduction of the incoming light, and setting this to 2 would 3393 mean that the filter is set to reduce incoming light by two stops 3394 (allowing 1/4 of the prior amount of light to the sensor). 3395 3396 It may take several frames before the lens filter density changes 3397 to the requested value. While the filter density is still changing, 3398 android.lens.state will be set to MOVING. 3399 </details> 3400 <tag id="V1" /> 3401 </entry> 3402 <entry name="focalLength" type="float" visibility="public" hwlevel="legacy"> 3403 <description> 3404 The desired lens focal length; used for optical zoom. 3405 </description> 3406 <units>Millimeters</units> 3407 <range>android.lens.info.availableFocalLengths</range> 3408 <details> 3409 This setting controls the physical focal length of the camera 3410 device's lens. Changing the focal length changes the field of 3411 view of the camera device, and is usually used for optical zoom. 3412 3413 Like android.lens.focusDistance and android.lens.aperture, this 3414 setting won't be applied instantaneously, and it may take several 3415 frames before the lens can change to the requested focal length. 3416 While the focal length is still changing, android.lens.state will 3417 be set to MOVING. 3418 3419 Optical zoom will not be supported on most devices. 3420 </details> 3421 <tag id="V1" /> 3422 </entry> 3423 <entry name="focusDistance" type="float" visibility="public" hwlevel="full"> 3424 <description>Desired distance to plane of sharpest focus, 3425 measured from frontmost surface of the lens.</description> 3426 <units>See android.lens.info.focusDistanceCalibration for details</units> 3427 <range>&gt;= 0</range> 3428 <details> 3429 This control can be used for setting manual focus, on devices that support 3430 the MANUAL_SENSOR capability and have a variable-focus lens (see 3431 android.lens.info.minimumFocusDistance). 3432 3433 A value of `0.0f` means infinity focus. The value set will be clamped to 3434 `[0.0f, android.lens.info.minimumFocusDistance]`. 3435 3436 Like android.lens.focalLength, this setting won't be applied 3437 instantaneously, and it may take several frames before the lens 3438 can move to the requested focus distance. While the lens is still moving, 3439 android.lens.state will be set to MOVING. 3440 3441 LEGACY devices support at most setting this to `0.0f` 3442 for infinity focus. 3443 </details> 3444 <tag id="BC" /> 3445 <tag id="V1" /> 3446 </entry> 3447 <entry name="opticalStabilizationMode" type="byte" visibility="public" 3448 enum="true" hwlevel="limited"> 3449 <enum> 3450 <value>OFF 3451 <notes>Optical stabilization is unavailable.</notes> 3452 </value> 3453 <value optional="true">ON 3454 <notes>Optical stabilization is enabled.</notes> 3455 </value> 3456 </enum> 3457 <description> 3458 Sets whether the camera device uses optical image stabilization (OIS) 3459 when capturing images. 3460 </description> 3461 <range>android.lens.info.availableOpticalStabilization</range> 3462 <details> 3463 OIS is used to compensate for motion blur due to small 3464 movements of the camera during capture. Unlike digital image 3465 stabilization (android.control.videoStabilizationMode), OIS 3466 makes use of mechanical elements to stabilize the camera 3467 sensor, and thus allows for longer exposure times before 3468 camera shake becomes apparent. 3469 3470 Switching between different optical stabilization modes may take several 3471 frames to initialize, the camera device will report the current mode in 3472 capture result metadata. For example, When "ON" mode is requested, the 3473 optical stabilization modes in the first several capture results may still 3474 be "OFF", and it will become "ON" when the initialization is done. 3475 3476 If a camera device supports both OIS and digital image stabilization 3477 (android.control.videoStabilizationMode), turning both modes on may produce undesirable 3478 interaction, so it is recommended not to enable both at the same time. 3479 3480 Not all devices will support OIS; see 3481 android.lens.info.availableOpticalStabilization for 3482 available controls. 3483 </details> 3484 <tag id="V1" /> 3485 </entry> 3486 </controls> 3487 <static> 3488 <namespace name="info"> 3489 <entry name="availableApertures" type="float" visibility="public" 3490 container="array" hwlevel="full"> 3491 <array> 3492 <size>n</size> 3493 </array> 3494 <description>List of aperture size values for android.lens.aperture that are 3495 supported by this camera device.</description> 3496 <units>The aperture f-number</units> 3497 <details>If the camera device doesn't support a variable lens aperture, 3498 this list will contain only one value, which is the fixed aperture size. 3499 3500 If the camera device supports a variable aperture, the aperture values 3501 in this list will be sorted in ascending order.</details> 3502 <tag id="V1" /> 3503 </entry> 3504 <entry name="availableFilterDensities" type="float" visibility="public" 3505 container="array" hwlevel="full"> 3506 <array> 3507 <size>n</size> 3508 </array> 3509 <description> 3510 List of neutral density filter values for 3511 android.lens.filterDensity that are supported by this camera device. 3512 </description> 3513 <units>Exposure value (EV)</units> 3514 <range> 3515 Values are &gt;= 0 3516 </range> 3517 <details> 3518 If a neutral density filter is not supported by this camera device, 3519 this list will contain only 0. Otherwise, this list will include every 3520 filter density supported by the camera device, in ascending order. 3521 </details> 3522 <tag id="V1" /> 3523 </entry> 3524 <entry name="availableFocalLengths" type="float" visibility="public" 3525 type_notes="The list of available focal lengths" 3526 container="array" hwlevel="legacy"> 3527 <array> 3528 <size>n</size> 3529 </array> 3530 <description> 3531 List of focal lengths for android.lens.focalLength that are supported by this camera 3532 device. 3533 </description> 3534 <units>Millimeters</units> 3535 <range> 3536 Values are &gt; 0 3537 </range> 3538 <details> 3539 If optical zoom is not supported, this list will only contain 3540 a single value corresponding to the fixed focal length of the 3541 device. Otherwise, this list will include every focal length supported 3542 by the camera device, in ascending order. 3543 </details> 3544 <tag id="BC" /> 3545 <tag id="V1" /> 3546 </entry> 3547 <entry name="availableOpticalStabilization" type="byte" 3548 visibility="public" type_notes="list of enums" container="array" 3549 typedef="enumList" hwlevel="limited"> 3550 <array> 3551 <size>n</size> 3552 </array> 3553 <description> 3554 List of optical image stabilization (OIS) modes for 3555 android.lens.opticalStabilizationMode that are supported by this camera device. 3556 </description> 3557 <range>Any value listed in android.lens.opticalStabilizationMode</range> 3558 <details> 3559 If OIS is not supported by a given camera device, this list will 3560 contain only OFF. 3561 </details> 3562 <tag id="V1" /> 3563 </entry> 3564 <entry name="hyperfocalDistance" type="float" visibility="public" optional="true" 3565 hwlevel="limited"> 3566 <description>Hyperfocal distance for this lens.</description> 3567 <units>See android.lens.info.focusDistanceCalibration for details</units> 3568 <range>If lens is fixed focus, &gt;= 0. If lens has focuser unit, the value is 3569 within `(0.0f, android.lens.info.minimumFocusDistance]`</range> 3570 <details> 3571 If the lens is not fixed focus, the camera device will report this 3572 field when android.lens.info.focusDistanceCalibration is APPROXIMATE or CALIBRATED. 3573 </details> 3574 </entry> 3575 <entry name="minimumFocusDistance" type="float" visibility="public" optional="true" 3576 hwlevel="limited"> 3577 <description>Shortest distance from frontmost surface 3578 of the lens that can be brought into sharp focus.</description> 3579 <units>See android.lens.info.focusDistanceCalibration for details</units> 3580 <range>&gt;= 0</range> 3581 <details>If the lens is fixed-focus, this will be 3582 0.</details> 3583 <hal_details>Mandatory for FULL devices; LIMITED devices 3584 must always set this value to 0 for fixed-focus; and may omit 3585 the minimum focus distance otherwise. 3586 3587 This field is also mandatory for all devices advertising 3588 the MANUAL_SENSOR capability.</hal_details> 3589 <tag id="V1" /> 3590 </entry> 3591 <entry name="shadingMapSize" type="int32" visibility="ndk_public" 3592 type_notes="width and height (N, M) of lens shading map provided by the camera device." 3593 container="array" typedef="size" hwlevel="full"> 3594 <array> 3595 <size>2</size> 3596 </array> 3597 <description>Dimensions of lens shading map.</description> 3598 <range>Both values &gt;= 1</range> 3599 <details> 3600 The map should be on the order of 30-40 rows and columns, and 3601 must be smaller than 64x64. 3602 </details> 3603 <tag id="V1" /> 3604 </entry> 3605 <entry name="focusDistanceCalibration" type="byte" visibility="public" 3606 enum="true" hwlevel="limited"> 3607 <enum> 3608 <value>UNCALIBRATED 3609 <notes> 3610 The lens focus distance is not accurate, and the units used for 3611 android.lens.focusDistance do not correspond to any physical units. 3612 3613 Setting the lens to the same focus distance on separate occasions may 3614 result in a different real focus distance, depending on factors such 3615 as the orientation of the device, the age of the focusing mechanism, 3616 and the device temperature. The focus distance value will still be 3617 in the range of `[0, android.lens.info.minimumFocusDistance]`, where 0 3618 represents the farthest focus. 3619 </notes> 3620 </value> 3621 <value>APPROXIMATE 3622 <notes> 3623 The lens focus distance is measured in diopters. 3624 3625 However, setting the lens to the same focus distance 3626 on separate occasions may result in a different real 3627 focus distance, depending on factors such as the 3628 orientation of the device, the age of the focusing 3629 mechanism, and the device temperature. 3630 </notes> 3631 </value> 3632 <value>CALIBRATED 3633 <notes> 3634 The lens focus distance is measured in diopters, and 3635 is calibrated. 3636 3637 The lens mechanism is calibrated so that setting the 3638 same focus distance is repeatable on multiple 3639 occasions with good accuracy, and the focus distance 3640 corresponds to the real physical distance to the plane 3641 of best focus. 3642 </notes> 3643 </value> 3644 </enum> 3645 <description>The lens focus distance calibration quality.</description> 3646 <details> 3647 The lens focus distance calibration quality determines the reliability of 3648 focus related metadata entries, i.e. android.lens.focusDistance, 3649 android.lens.focusRange, android.lens.info.hyperfocalDistance, and 3650 android.lens.info.minimumFocusDistance. 3651 3652 APPROXIMATE and CALIBRATED devices report the focus metadata in 3653 units of diopters (1/meter), so `0.0f` represents focusing at infinity, 3654 and increasing positive numbers represent focusing closer and closer 3655 to the camera device. The focus distance control also uses diopters 3656 on these devices. 3657 3658 UNCALIBRATED devices do not use units that are directly comparable 3659 to any real physical measurement, but `0.0f` still represents farthest 3660 focus, and android.lens.info.minimumFocusDistance represents the 3661 nearest focus the device can achieve. 3662 </details> 3663 <hal_details> 3664 For devices advertise APPROXIMATE quality or higher, diopters 0 (infinity 3665 focus) must work. When autofocus is disabled (android.control.afMode == OFF) 3666 and the lens focus distance is set to 0 diopters 3667 (android.lens.focusDistance == 0), the lens will move to focus at infinity 3668 and is stably focused at infinity even if the device tilts. It may take the 3669 lens some time to move; during the move the lens state should be MOVING and 3670 the output diopter value should be changing toward 0. 3671 </hal_details> 3672 <tag id="V1" /> 3673 </entry> 3674 </namespace> 3675 <entry name="facing" type="byte" visibility="public" enum="true" hwlevel="legacy"> 3676 <enum> 3677 <value>FRONT 3678 <notes> 3679 The camera device faces the same direction as the device's screen. 3680 </notes></value> 3681 <value>BACK 3682 <notes> 3683 The camera device faces the opposite direction as the device's screen. 3684 </notes></value> 3685 <value>EXTERNAL 3686 <notes> 3687 The camera device is an external camera, and has no fixed facing relative to the 3688 device's screen. 3689 </notes></value> 3690 </enum> 3691 <description>Direction the camera faces relative to 3692 device screen.</description> 3693 </entry> 3694 <entry name="poseRotation" type="float" visibility="public" 3695 container="array"> 3696 <array> 3697 <size>4</size> 3698 </array> 3699 <description> 3700 The orientation of the camera relative to the sensor 3701 coordinate system. 3702 </description> 3703 <units> 3704 Quaternion coefficients 3705 </units> 3706 <details> 3707 The four coefficients that describe the quaternion 3708 rotation from the Android sensor coordinate system to a 3709 camera-aligned coordinate system where the X-axis is 3710 aligned with the long side of the image sensor, the Y-axis 3711 is aligned with the short side of the image sensor, and 3712 the Z-axis is aligned with the optical axis of the sensor. 3713 3714 To convert from the quaternion coefficients `(x,y,z,w)` 3715 to the axis of rotation `(a_x, a_y, a_z)` and rotation 3716 amount `theta`, the following formulas can be used: 3717 3718 theta = 2 * acos(w) 3719 a_x = x / sin(theta/2) 3720 a_y = y / sin(theta/2) 3721 a_z = z / sin(theta/2) 3722 3723 To create a 3x3 rotation matrix that applies the rotation 3724 defined by this quaternion, the following matrix can be 3725 used: 3726 3727 R = [ 1 - 2y^2 - 2z^2, 2xy - 2zw, 2xz + 2yw, 3728 2xy + 2zw, 1 - 2x^2 - 2z^2, 2yz - 2xw, 3729 2xz - 2yw, 2yz + 2xw, 1 - 2x^2 - 2y^2 ] 3730 3731 This matrix can then be used to apply the rotation to a 3732 column vector point with 3733 3734 `p' = Rp` 3735 3736 where `p` is in the device sensor coordinate system, and 3737 `p'` is in the camera-oriented coordinate system. 3738 </details> 3739 <tag id="DEPTH" /> 3740 </entry> 3741 <entry name="poseTranslation" type="float" visibility="public" 3742 container="array"> 3743 <array> 3744 <size>3</size> 3745 </array> 3746 <description>Position of the camera optical center.</description> 3747 <units>Meters</units> 3748 <details> 3749 The position of the camera device's lens optical center, 3750 as a three-dimensional vector `(x,y,z)`, relative to the 3751 optical center of the largest camera device facing in the 3752 same direction as this camera, in the {@link 3753 android.hardware.SensorEvent Android sensor coordinate 3754 axes}. Note that only the axis definitions are shared with 3755 the sensor coordinate system, but not the origin. 3756 3757 If this device is the largest or only camera device with a 3758 given facing, then this position will be `(0, 0, 0)`; a 3759 camera device with a lens optical center located 3 cm from 3760 the main sensor along the +X axis (to the right from the 3761 user's perspective) will report `(0.03, 0, 0)`. 3762 3763 To transform a pixel coordinates between two cameras 3764 facing the same direction, first the source camera 3765 android.lens.radialDistortion must be corrected for. Then 3766 the source camera android.lens.intrinsicCalibration needs 3767 to be applied, followed by the android.lens.poseRotation 3768 of the source camera, the translation of the source camera 3769 relative to the destination camera, the 3770 android.lens.poseRotation of the destination camera, and 3771 finally the inverse of android.lens.intrinsicCalibration 3772 of the destination camera. This obtains a 3773 radial-distortion-free coordinate in the destination 3774 camera pixel coordinates. 3775 3776 To compare this against a real image from the destination 3777 camera, the destination camera image then needs to be 3778 corrected for radial distortion before comparison or 3779 sampling. 3780 </details> 3781 <tag id="DEPTH" /> 3782 </entry> 3783 </static> 3784 <dynamic> 3785 <clone entry="android.lens.aperture" kind="controls"> 3786 <tag id="V1" /> 3787 </clone> 3788 <clone entry="android.lens.filterDensity" kind="controls"> 3789 <tag id="V1" /> 3790 </clone> 3791 <clone entry="android.lens.focalLength" kind="controls"> 3792 <tag id="BC" /> 3793 </clone> 3794 <clone entry="android.lens.focusDistance" kind="controls"> 3795 <details>Should be zero for fixed-focus cameras</details> 3796 <tag id="BC" /> 3797 </clone> 3798 <entry name="focusRange" type="float" visibility="public" 3799 type_notes="Range of scene distances that are in focus" 3800 container="array" typedef="pairFloatFloat" hwlevel="limited"> 3801 <array> 3802 <size>2</size> 3803 </array> 3804 <description>The range of scene distances that are in 3805 sharp focus (depth of field).</description> 3806 <units>A pair of focus distances in diopters: (near, 3807 far); see android.lens.info.focusDistanceCalibration for details.</units> 3808 <range>&gt;=0</range> 3809 <details>If variable focus not supported, can still report 3810 fixed depth of field range</details> 3811 <tag id="BC" /> 3812 </entry> 3813 <clone entry="android.lens.opticalStabilizationMode" 3814 kind="controls"> 3815 <tag id="V1" /> 3816 </clone> 3817 <entry name="state" type="byte" visibility="public" enum="true" hwlevel="limited"> 3818 <enum> 3819 <value>STATIONARY 3820 <notes> 3821 The lens parameters (android.lens.focalLength, android.lens.focusDistance, 3822 android.lens.filterDensity and android.lens.aperture) are not changing. 3823 </notes> 3824 </value> 3825 <value>MOVING 3826 <notes> 3827 One or several of the lens parameters 3828 (android.lens.focalLength, android.lens.focusDistance, 3829 android.lens.filterDensity or android.lens.aperture) is 3830 currently changing. 3831 </notes> 3832 </value> 3833 </enum> 3834 <description>Current lens status.</description> 3835 <details> 3836 For lens parameters android.lens.focalLength, android.lens.focusDistance, 3837 android.lens.filterDensity and android.lens.aperture, when changes are requested, 3838 they may take several frames to reach the requested values. This state indicates 3839 the current status of the lens parameters. 3840 3841 When the state is STATIONARY, the lens parameters are not changing. This could be 3842 either because the parameters are all fixed, or because the lens has had enough 3843 time to reach the most recently-requested values. 3844 If all these lens parameters are not changable for a camera device, as listed below: 3845 3846 * Fixed focus (`android.lens.info.minimumFocusDistance == 0`), which means 3847 android.lens.focusDistance parameter will always be 0. 3848 * Fixed focal length (android.lens.info.availableFocalLengths contains single value), 3849 which means the optical zoom is not supported. 3850 * No ND filter (android.lens.info.availableFilterDensities contains only 0). 3851 * Fixed aperture (android.lens.info.availableApertures contains single value). 3852 3853 Then this state will always be STATIONARY. 3854 3855 When the state is MOVING, it indicates that at least one of the lens parameters 3856 is changing. 3857 </details> 3858 <tag id="V1" /> 3859 </entry> 3860 <clone entry="android.lens.poseRotation" kind="static"> 3861 </clone> 3862 <clone entry="android.lens.poseTranslation" kind="static"> 3863 </clone> 3864 </dynamic> 3865 <static> 3866 <entry name="intrinsicCalibration" type="float" visibility="public" 3867 container="array"> 3868 <array> 3869 <size>5</size> 3870 </array> 3871 <description> 3872 The parameters for this camera device's intrinsic 3873 calibration. 3874 </description> 3875 <units> 3876 Pixels in the 3877 android.sensor.info.preCorrectionActiveArraySize 3878 coordinate system. 3879 </units> 3880 <details> 3881 The five calibration parameters that describe the 3882 transform from camera-centric 3D coordinates to sensor 3883 pixel coordinates: 3884 3885 [f_x, f_y, c_x, c_y, s] 3886 3887 Where `f_x` and `f_y` are the horizontal and vertical 3888 focal lengths, `[c_x, c_y]` is the position of the optical 3889 axis, and `s` is a skew parameter for the sensor plane not 3890 being aligned with the lens plane. 3891 3892 These are typically used within a transformation matrix K: 3893 3894 K = [ f_x, s, c_x, 3895 0, f_y, c_y, 3896 0 0, 1 ] 3897 3898 which can then be combined with the camera pose rotation 3899 `R` and translation `t` (android.lens.poseRotation and 3900 android.lens.poseTranslation, respective) to calculate the 3901 complete transform from world coordinates to pixel 3902 coordinates: 3903 3904 P = [ K 0 * [ R t 3905 0 1 ] 0 1 ] 3906 3907 and with `p_w` being a point in the world coordinate system 3908 and `p_s` being a point in the camera active pixel array 3909 coordinate system, and with the mapping including the 3910 homogeneous division by z: 3911 3912 p_h = (x_h, y_h, z_h) = P p_w 3913 p_s = p_h / z_h 3914 3915 so `[x_s, y_s]` is the pixel coordinates of the world 3916 point, `z_s = 1`, and `w_s` is a measurement of disparity 3917 (depth) in pixel coordinates. 3918 3919 Note that the coordinate system for this transform is the 3920 android.sensor.info.preCorrectionActiveArraySize system, 3921 where `(0,0)` is the top-left of the 3922 preCorrectionActiveArraySize rectangle. Once the pose and 3923 intrinsic calibration transforms have been applied to a 3924 world point, then the android.lens.radialDistortion 3925 transform needs to be applied, and the result adjusted to 3926 be in the android.sensor.info.activeArraySize coordinate 3927 system (where `(0, 0)` is the top-left of the 3928 activeArraySize rectangle), to determine the final pixel 3929 coordinate of the world point for processed (non-RAW) 3930 output buffers. 3931 </details> 3932 <tag id="DEPTH" /> 3933 </entry> 3934 <entry name="radialDistortion" type="float" visibility="public" 3935 container="array"> 3936 <array> 3937 <size>6</size> 3938 </array> 3939 <description> 3940 The correction coefficients to correct for this camera device's 3941 radial and tangential lens distortion. 3942 </description> 3943 <units> 3944 Unitless coefficients. 3945 </units> 3946 <details> 3947 Four radial distortion coefficients `[kappa_0, kappa_1, kappa_2, 3948 kappa_3]` and two tangential distortion coefficients 3949 `[kappa_4, kappa_5]` that can be used to correct the 3950 lens's geometric distortion with the mapping equations: 3951 3952 x_c = x_i * ( kappa_0 + kappa_1 * r^2 + kappa_2 * r^4 + kappa_3 * r^6 ) + 3953 kappa_4 * (2 * x_i * y_i) + kappa_5 * ( r^2 + 2 * x_i^2 ) 3954 y_c = y_i * ( kappa_0 + kappa_1 * r^2 + kappa_2 * r^4 + kappa_3 * r^6 ) + 3955 kappa_5 * (2 * x_i * y_i) + kappa_4 * ( r^2 + 2 * y_i^2 ) 3956 3957 Here, `[x_c, y_c]` are the coordinates to sample in the 3958 input image that correspond to the pixel values in the 3959 corrected image at the coordinate `[x_i, y_i]`: 3960 3961 correctedImage(x_i, y_i) = sample_at(x_c, y_c, inputImage) 3962 3963 The pixel coordinates are defined in a normalized 3964 coordinate system related to the 3965 android.lens.intrinsicCalibration calibration fields. 3966 Both `[x_i, y_i]` and `[x_c, y_c]` have `(0,0)` at the 3967 lens optical center `[c_x, c_y]`. The maximum magnitudes 3968 of both x and y coordinates are normalized to be 1 at the 3969 edge further from the optical center, so the range 3970 for both dimensions is `-1 <= x <= 1`. 3971 3972 Finally, `r` represents the radial distance from the 3973 optical center, `r^2 = x_i^2 + y_i^2`, and its magnitude 3974 is therefore no larger than `|r| <= sqrt(2)`. 3975 3976 The distortion model used is the Brown-Conrady model. 3977 </details> 3978 <tag id="DEPTH" /> 3979 </entry> 3980 </static> 3981 <dynamic> 3982 <clone entry="android.lens.intrinsicCalibration" kind="static"> 3983 </clone> 3984 <clone entry="android.lens.radialDistortion" kind="static"> 3985 </clone> 3986 </dynamic> 3987 </section> 3988 <section name="noiseReduction"> 3989 <controls> 3990 <entry name="mode" type="byte" visibility="public" enum="true" hwlevel="full"> 3991 <enum> 3992 <value>OFF 3993 <notes>No noise reduction is applied.</notes></value> 3994 <value>FAST 3995 <notes>Noise reduction is applied without reducing frame rate relative to sensor 3996 output. It may be the same as OFF if noise reduction will reduce frame rate 3997 relative to sensor.</notes></value> 3998 <value>HIGH_QUALITY 3999 <notes>High-quality noise reduction is applied, at the cost of possibly reduced frame 4000 rate relative to sensor output.</notes></value> 4001 <value optional="true">MINIMAL 4002 <notes>MINIMAL noise reduction is applied without reducing frame rate relative to 4003 sensor output. </notes></value> 4004 <value optional="true">ZERO_SHUTTER_LAG 4005 4006 <notes>Noise reduction is applied at different levels for different output streams, 4007 based on resolution. Streams at maximum recording resolution (see {@link 4008 android.hardware.camera2.CameraDevice#createCaptureSession}) or below have noise 4009 reduction applied, while higher-resolution streams have MINIMAL (if supported) or no 4010 noise reduction applied (if MINIMAL is not supported.) The degree of noise reduction 4011 for low-resolution streams is tuned so that frame rate is not impacted, and the quality 4012 is equal to or better than FAST (since it is only applied to lower-resolution outputs, 4013 quality may improve from FAST). 4014 4015 This mode is intended to be used by applications operating in a zero-shutter-lag mode 4016 with YUV or PRIVATE reprocessing, where the application continuously captures 4017 high-resolution intermediate buffers into a circular buffer, from which a final image is 4018 produced via reprocessing when a user takes a picture. For such a use case, the 4019 high-resolution buffers must not have noise reduction applied to maximize efficiency of 4020 preview and to avoid over-applying noise filtering when reprocessing, while 4021 low-resolution buffers (used for recording or preview, generally) need noise reduction 4022 applied for reasonable preview quality. 4023 4024 This mode is guaranteed to be supported by devices that support either the 4025 YUV_REPROCESSING or PRIVATE_REPROCESSING capabilities 4026 (android.request.availableCapabilities lists either of those capabilities) and it will 4027 be the default mode for CAMERA3_TEMPLATE_ZERO_SHUTTER_LAG template. 4028 </notes></value> 4029 </enum> 4030 <description>Mode of operation for the noise reduction algorithm.</description> 4031 <range>android.noiseReduction.availableNoiseReductionModes</range> 4032 <details>The noise reduction algorithm attempts to improve image quality by removing 4033 excessive noise added by the capture process, especially in dark conditions. 4034 4035 OFF means no noise reduction will be applied by the camera device, for both raw and 4036 YUV domain. 4037 4038 MINIMAL means that only sensor raw domain basic noise reduction is enabled ,to remove 4039 demosaicing or other processing artifacts. For YUV_REPROCESSING, MINIMAL is same as OFF. 4040 This mode is optional, may not be support by all devices. The application should check 4041 android.noiseReduction.availableNoiseReductionModes before using it. 4042 4043 FAST/HIGH_QUALITY both mean camera device determined noise filtering 4044 will be applied. HIGH_QUALITY mode indicates that the camera device 4045 will use the highest-quality noise filtering algorithms, 4046 even if it slows down capture rate. FAST means the camera device will not 4047 slow down capture rate when applying noise filtering. FAST may be the same as MINIMAL if 4048 MINIMAL is listed, or the same as OFF if any noise filtering will slow down capture rate. 4049 Every output stream will have a similar amount of enhancement applied. 4050 4051 ZERO_SHUTTER_LAG is meant to be used by applications that maintain a continuous circular 4052 buffer of high-resolution images during preview and reprocess image(s) from that buffer 4053 into a final capture when triggered by the user. In this mode, the camera device applies 4054 noise reduction to low-resolution streams (below maximum recording resolution) to maximize 4055 preview quality, but does not apply noise reduction to high-resolution streams, since 4056 those will be reprocessed later if necessary. 4057 4058 For YUV_REPROCESSING, these FAST/HIGH_QUALITY modes both mean that the camera device 4059 will apply FAST/HIGH_QUALITY YUV domain noise reduction, respectively. The camera device 4060 may adjust the noise reduction parameters for best image quality based on the 4061 android.reprocess.effectiveExposureFactor if it is set. 4062 </details> 4063 <hal_details> 4064 For YUV_REPROCESSING The HAL can use android.reprocess.effectiveExposureFactor to 4065 adjust the internal noise reduction parameters appropriately to get the best quality 4066 images. 4067 </hal_details> 4068 <tag id="V1" /> 4069 <tag id="REPROC" /> 4070 </entry> 4071 <entry name="strength" type="byte"> 4072 <description>Control the amount of noise reduction 4073 applied to the images</description> 4074 <units>1-10; 10 is max noise reduction</units> 4075 <range>1 - 10</range> 4076 <tag id="FUTURE" /> 4077 </entry> 4078 </controls> 4079 <static> 4080 <entry name="availableNoiseReductionModes" type="byte" visibility="public" 4081 type_notes="list of enums" container="array" typedef="enumList" hwlevel="limited"> 4082 <array> 4083 <size>n</size> 4084 </array> 4085 <description> 4086 List of noise reduction modes for android.noiseReduction.mode that are supported 4087 by this camera device. 4088 </description> 4089 <range>Any value listed in android.noiseReduction.mode</range> 4090 <details> 4091 Full-capability camera devices will always support OFF and FAST. 4092 4093 Camera devices that support YUV_REPROCESSING or PRIVATE_REPROCESSING will support 4094 ZERO_SHUTTER_LAG. 4095 4096 Legacy-capability camera devices will only support FAST mode. 4097 </details> 4098 <hal_details> 4099 HAL must support both FAST and HIGH_QUALITY if noise reduction control is available 4100 on the camera device, but the underlying implementation can be the same for both modes. 4101 That is, if the highest quality implementation on the camera device does not slow down 4102 capture rate, then FAST and HIGH_QUALITY will generate the same output. 4103 </hal_details> 4104 <tag id="V1" /> 4105 <tag id="REPROC" /> 4106 </entry> 4107 </static> 4108 <dynamic> 4109 <clone entry="android.noiseReduction.mode" kind="controls"> 4110 <tag id="V1" /> 4111 <tag id="REPROC" /> 4112 </clone> 4113 </dynamic> 4114 </section> 4115 <section name="quirks"> 4116 <static> 4117 <entry name="meteringCropRegion" type="byte" visibility="system" deprecated="true" optional="true"> 4118 <description>If set to 1, the camera service does not 4119 scale 'normalized' coordinates with respect to the crop 4120 region. This applies to metering input (a{e,f,wb}Region 4121 and output (face rectangles).</description> 4122 <details>Normalized coordinates refer to those in the 4123 (-1000,1000) range mentioned in the 4124 android.hardware.Camera API. 4125 4126 HAL implementations should instead always use and emit 4127 sensor array-relative coordinates for all region data. Does 4128 not need to be listed in static metadata. Support will be 4129 removed in future versions of camera service.</details> 4130 </entry> 4131 <entry name="triggerAfWithAuto" type="byte" visibility="system" deprecated="true" optional="true"> 4132 <description>If set to 1, then the camera service always 4133 switches to FOCUS_MODE_AUTO before issuing a AF 4134 trigger.</description> 4135 <details>HAL implementations should implement AF trigger 4136 modes for AUTO, MACRO, CONTINUOUS_FOCUS, and 4137 CONTINUOUS_PICTURE modes instead of using this flag. Does 4138 not need to be listed in static metadata. Support will be 4139 removed in future versions of camera service</details> 4140 </entry> 4141 <entry name="useZslFormat" type="byte" visibility="system" deprecated="true" optional="true"> 4142 <description>If set to 1, the camera service uses 4143 CAMERA2_PIXEL_FORMAT_ZSL instead of 4144 HAL_PIXEL_FORMAT_IMPLEMENTATION_DEFINED for the zero 4145 shutter lag stream</description> 4146 <details>HAL implementations should use gralloc usage flags 4147 to determine that a stream will be used for 4148 zero-shutter-lag, instead of relying on an explicit 4149 format setting. Does not need to be listed in static 4150 metadata. Support will be removed in future versions of 4151 camera service.</details> 4152 </entry> 4153 <entry name="usePartialResult" type="byte" visibility="hidden" deprecated="true" optional="true"> 4154 <description> 4155 If set to 1, the HAL will always split result 4156 metadata for a single capture into multiple buffers, 4157 returned using multiple process_capture_result calls. 4158 </description> 4159 <details> 4160 Does not need to be listed in static 4161 metadata. Support for partial results will be reworked in 4162 future versions of camera service. This quirk will stop 4163 working at that point; DO NOT USE without careful 4164 consideration of future support. 4165 </details> 4166 <hal_details> 4167 Refer to `camera3_capture_result::partial_result` 4168 for information on how to implement partial results. 4169 </hal_details> 4170 </entry> 4171 </static> 4172 <dynamic> 4173 <entry name="partialResult" type="byte" visibility="hidden" deprecated="true" optional="true" enum="true" typedef="boolean"> 4174 <enum> 4175 <value>FINAL 4176 <notes>The last or only metadata result buffer 4177 for this capture.</notes> 4178 </value> 4179 <value>PARTIAL 4180 <notes>A partial buffer of result metadata for this 4181 capture. More result buffers for this capture will be sent 4182 by the camera device, the last of which will be marked 4183 FINAL.</notes> 4184 </value> 4185 </enum> 4186 <description> 4187 Whether a result given to the framework is the 4188 final one for the capture, or only a partial that contains a 4189 subset of the full set of dynamic metadata 4190 values.</description> 4191 <range>Optional. Default value is FINAL.</range> 4192 <details> 4193 The entries in the result metadata buffers for a 4194 single capture may not overlap, except for this entry. The 4195 FINAL buffers must retain FIFO ordering relative to the 4196 requests that generate them, so the FINAL buffer for frame 3 must 4197 always be sent to the framework after the FINAL buffer for frame 2, and 4198 before the FINAL buffer for frame 4. PARTIAL buffers may be returned 4199 in any order relative to other frames, but all PARTIAL buffers for a given 4200 capture must arrive before the FINAL buffer for that capture. This entry may 4201 only be used by the camera device if quirks.usePartialResult is set to 1. 4202 </details> 4203 <hal_details> 4204 Refer to `camera3_capture_result::partial_result` 4205 for information on how to implement partial results. 4206 </hal_details> 4207 </entry> 4208 </dynamic> 4209 </section> 4210 <section name="request"> 4211 <controls> 4212 <entry name="frameCount" type="int32" visibility="system" deprecated="true"> 4213 <description>A frame counter set by the framework. Must 4214 be maintained unchanged in output frame. This value monotonically 4215 increases with every new result (that is, each new result has a unique 4216 frameCount value). 4217 </description> 4218 <units>incrementing integer</units> 4219 <range>Any int.</range> 4220 </entry> 4221 <entry name="id" type="int32" visibility="hidden"> 4222 <description>An application-specified ID for the current 4223 request. Must be maintained unchanged in output 4224 frame</description> 4225 <units>arbitrary integer assigned by application</units> 4226 <range>Any int</range> 4227 <tag id="V1" /> 4228 </entry> 4229 <entry name="inputStreams" type="int32" visibility="system" deprecated="true" 4230 container="array"> 4231 <array> 4232 <size>n</size> 4233 </array> 4234 <description>List which camera reprocess stream is used 4235 for the source of reprocessing data.</description> 4236 <units>List of camera reprocess stream IDs</units> 4237 <range> 4238 Typically, only one entry allowed, must be a valid reprocess stream ID. 4239 </range> 4240 <details>Only meaningful when android.request.type == 4241 REPROCESS. Ignored otherwise</details> 4242 <tag id="HAL2" /> 4243 </entry> 4244 <entry name="metadataMode" type="byte" visibility="system" 4245 enum="true"> 4246 <enum> 4247 <value>NONE 4248 <notes>No metadata should be produced on output, except 4249 for application-bound buffer data. If no 4250 application-bound streams exist, no frame should be 4251 placed in the output frame queue. If such streams 4252 exist, a frame should be placed on the output queue 4253 with null metadata but with the necessary output buffer 4254 information. Timestamp information should still be 4255 included with any output stream buffers</notes></value> 4256 <value>FULL 4257 <notes>All metadata should be produced. Statistics will 4258 only be produced if they are separately 4259 enabled</notes></value> 4260 </enum> 4261 <description>How much metadata to produce on 4262 output</description> 4263 <tag id="FUTURE" /> 4264 </entry> 4265 <entry name="outputStreams" type="int32" visibility="system" deprecated="true" 4266 container="array"> 4267 <array> 4268 <size>n</size> 4269 </array> 4270 <description>Lists which camera output streams image data 4271 from this capture must be sent to</description> 4272 <units>List of camera stream IDs</units> 4273 <range>List must only include streams that have been 4274 created</range> 4275 <details>If no output streams are listed, then the image 4276 data should simply be discarded. The image data must 4277 still be captured for metadata and statistics production, 4278 and the lens and flash must operate as requested.</details> 4279 <tag id="HAL2" /> 4280 </entry> 4281 <entry name="type" type="byte" visibility="system" deprecated="true" enum="true"> 4282 <enum> 4283 <value>CAPTURE 4284 <notes>Capture a new image from the imaging hardware, 4285 and process it according to the 4286 settings</notes></value> 4287 <value>REPROCESS 4288 <notes>Process previously captured data; the 4289 android.request.inputStreams parameter determines the 4290 source reprocessing stream. TODO: Mark dynamic metadata 4291 needed for reprocessing with [RP]</notes></value> 4292 </enum> 4293 <description>The type of the request; either CAPTURE or 4294 REPROCESS. For HAL3, this tag is redundant. 4295 </description> 4296 <tag id="HAL2" /> 4297 </entry> 4298 </controls> 4299 <static> 4300 <entry name="maxNumOutputStreams" type="int32" visibility="ndk_public" 4301 container="array" hwlevel="legacy"> 4302 <array> 4303 <size>3</size> 4304 </array> 4305 <description>The maximum numbers of different types of output streams 4306 that can be configured and used simultaneously by a camera device. 4307 </description> 4308 <range> 4309 For processed (and stalling) format streams, &gt;= 1. 4310 4311 For Raw format (either stalling or non-stalling) streams, &gt;= 0. 4312 4313 For processed (but not stalling) format streams, &gt;= 3 4314 for FULL mode devices (`android.info.supportedHardwareLevel == FULL`); 4315 &gt;= 2 for LIMITED mode devices (`android.info.supportedHardwareLevel == LIMITED`). 4316 </range> 4317 <details> 4318 This is a 3 element tuple that contains the max number of output simultaneous 4319 streams for raw sensor, processed (but not stalling), and processed (and stalling) 4320 formats respectively. For example, assuming that JPEG is typically a processed and 4321 stalling stream, if max raw sensor format output stream number is 1, max YUV streams 4322 number is 3, and max JPEG stream number is 2, then this tuple should be `(1, 3, 2)`. 4323 4324 This lists the upper bound of the number of output streams supported by 4325 the camera device. Using more streams simultaneously may require more hardware and 4326 CPU resources that will consume more power. The image format for an output stream can 4327 be any supported format provided by android.scaler.availableStreamConfigurations. 4328 The formats defined in android.scaler.availableStreamConfigurations can be catergorized 4329 into the 3 stream types as below: 4330 4331 * Processed (but stalling): any non-RAW format with a stallDurations &gt; 0. 4332 Typically {@link android.graphics.ImageFormat#JPEG JPEG format}. 4333 * Raw formats: {@link android.graphics.ImageFormat#RAW_SENSOR RAW_SENSOR}, {@link 4334 android.graphics.ImageFormat#RAW10 RAW10}, or {@link android.graphics.ImageFormat#RAW12 4335 RAW12}. 4336 * Processed (but not-stalling): any non-RAW format without a stall duration. 4337 Typically {@link android.graphics.ImageFormat#YUV_420_888 YUV_420_888}, 4338 {@link android.graphics.ImageFormat#NV21 NV21}, or 4339 {@link android.graphics.ImageFormat#YV12 YV12}. 4340 </details> 4341 <tag id="BC" /> 4342 </entry> 4343 <entry name="maxNumOutputRaw" type="int32" visibility="java_public" synthetic="true" 4344 hwlevel="legacy"> 4345 <description>The maximum numbers of different types of output streams 4346 that can be configured and used simultaneously by a camera device 4347 for any `RAW` formats. 4348 </description> 4349 <range> 4350 &gt;= 0 4351 </range> 4352 <details> 4353 This value contains the max number of output simultaneous 4354 streams from the raw sensor. 4355 4356 This lists the upper bound of the number of output streams supported by 4357 the camera device. Using more streams simultaneously may require more hardware and 4358 CPU resources that will consume more power. The image format for this kind of an output stream can 4359 be any `RAW` and supported format provided by android.scaler.streamConfigurationMap. 4360 4361 In particular, a `RAW` format is typically one of: 4362 4363 * {@link android.graphics.ImageFormat#RAW_SENSOR RAW_SENSOR} 4364 * {@link android.graphics.ImageFormat#RAW10 RAW10} 4365 * {@link android.graphics.ImageFormat#RAW12 RAW12} 4366 4367 LEGACY mode devices (android.info.supportedHardwareLevel `==` LEGACY) 4368 never support raw streams. 4369 </details> 4370 </entry> 4371 <entry name="maxNumOutputProc" type="int32" visibility="java_public" synthetic="true" 4372 hwlevel="legacy"> 4373 <description>The maximum numbers of different types of output streams 4374 that can be configured and used simultaneously by a camera device 4375 for any processed (but not-stalling) formats. 4376 </description> 4377 <range> 4378 &gt;= 3 4379 for FULL mode devices (`android.info.supportedHardwareLevel == FULL`); 4380 &gt;= 2 for LIMITED mode devices (`android.info.supportedHardwareLevel == LIMITED`). 4381 </range> 4382 <details> 4383 This value contains the max number of output simultaneous 4384 streams for any processed (but not-stalling) formats. 4385 4386 This lists the upper bound of the number of output streams supported by 4387 the camera device. Using more streams simultaneously may require more hardware and 4388 CPU resources that will consume more power. The image format for this kind of an output stream can 4389 be any non-`RAW` and supported format provided by android.scaler.streamConfigurationMap. 4390 4391 Processed (but not-stalling) is defined as any non-RAW format without a stall duration. 4392 Typically: 4393 4394 * {@link android.graphics.ImageFormat#YUV_420_888 YUV_420_888} 4395 * {@link android.graphics.ImageFormat#NV21 NV21} 4396 * {@link android.graphics.ImageFormat#YV12 YV12} 4397 * Implementation-defined formats, i.e. {@link 4398 android.hardware.camera2.params.StreamConfigurationMap#isOutputSupportedFor(Class)} 4399 4400 For full guarantees, query {@link 4401 android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration} with a 4402 processed format -- it will return 0 for a non-stalling stream. 4403 4404 LEGACY devices will support at least 2 processing/non-stalling streams. 4405 </details> 4406 </entry> 4407 <entry name="maxNumOutputProcStalling" type="int32" visibility="java_public" synthetic="true" 4408 hwlevel="legacy"> 4409 <description>The maximum numbers of different types of output streams 4410 that can be configured and used simultaneously by a camera device 4411 for any processed (and stalling) formats. 4412 </description> 4413 <range> 4414 &gt;= 1 4415 </range> 4416 <details> 4417 This value contains the max number of output simultaneous 4418 streams for any processed (but not-stalling) formats. 4419 4420 This lists the upper bound of the number of output streams supported by 4421 the camera device. Using more streams simultaneously may require more hardware and 4422 CPU resources that will consume more power. The image format for this kind of an output stream can 4423 be any non-`RAW` and supported format provided by android.scaler.streamConfigurationMap. 4424 4425 A processed and stalling format is defined as any non-RAW format with a stallDurations 4426 &gt; 0. Typically only the {@link android.graphics.ImageFormat#JPEG JPEG format} is a 4427 stalling format. 4428 4429 For full guarantees, query {@link 4430 android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration} with a 4431 processed format -- it will return a non-0 value for a stalling stream. 4432 4433 LEGACY devices will support up to 1 processing/stalling stream. 4434 </details> 4435 </entry> 4436 <entry name="maxNumReprocessStreams" type="int32" visibility="system" 4437 deprecated="true" container="array"> 4438 <array> 4439 <size>1</size> 4440 </array> 4441 <description>How many reprocessing streams of any type 4442 can be allocated at the same time.</description> 4443 <range>&gt;= 0</range> 4444 <details> 4445 Only used by HAL2.x. 4446 4447 When set to 0, it means no reprocess stream is supported. 4448 </details> 4449 <tag id="HAL2" /> 4450 </entry> 4451 <entry name="maxNumInputStreams" type="int32" visibility="public" hwlevel="full"> 4452 <description> 4453 The maximum numbers of any type of input streams 4454 that can be configured and used simultaneously by a camera device. 4455 </description> 4456 <range> 4457 0 or 1. 4458 </range> 4459 <details>When set to 0, it means no input stream is supported. 4460 4461 The image format for a input stream can be any supported format returned by {@link 4462 android.hardware.camera2.params.StreamConfigurationMap#getInputFormats}. When using an 4463 input stream, there must be at least one output stream configured to to receive the 4464 reprocessed images. 4465 4466 When an input stream and some output streams are used in a reprocessing request, 4467 only the input buffer will be used to produce these output stream buffers, and a 4468 new sensor image will not be captured. 4469 4470 For example, for Zero Shutter Lag (ZSL) still capture use case, the input 4471 stream image format will be PRIVATE, the associated output stream image format 4472 should be JPEG. 4473 </details> 4474 <hal_details> 4475 For the reprocessing flow and controls, see 4476 hardware/libhardware/include/hardware/camera3.h Section 10 for more details. 4477 </hal_details> 4478 <tag id="REPROC" /> 4479 </entry> 4480 </static> 4481 <dynamic> 4482 <entry name="frameCount" type="int32" visibility="hidden" deprecated="true"> 4483 <description>A frame counter set by the framework. This value monotonically 4484 increases with every new result (that is, each new result has a unique 4485 frameCount value).</description> 4486 <units>count of frames</units> 4487 <range>&gt; 0</range> 4488 <details>Reset on release()</details> 4489 </entry> 4490 <clone entry="android.request.id" kind="controls"></clone> 4491 <clone entry="android.request.metadataMode" 4492 kind="controls"></clone> 4493 <clone entry="android.request.outputStreams" 4494 kind="controls"></clone> 4495 <entry name="pipelineDepth" type="byte" visibility="public" hwlevel="legacy"> 4496 <description>Specifies the number of pipeline stages the frame went 4497 through from when it was exposed to when the final completed result 4498 was available to the framework.</description> 4499 <range>&lt;= android.request.pipelineMaxDepth</range> 4500 <details>Depending on what settings are used in the request, and 4501 what streams are configured, the data may undergo less processing, 4502 and some pipeline stages skipped. 4503 4504 See android.request.pipelineMaxDepth for more details. 4505 </details> 4506 <hal_details> 4507 This value must always represent the accurate count of how many 4508 pipeline stages were actually used. 4509 </hal_details> 4510 </entry> 4511 </dynamic> 4512 <static> 4513 <entry name="pipelineMaxDepth" type="byte" visibility="public" hwlevel="legacy"> 4514 <description>Specifies the number of maximum pipeline stages a frame 4515 has to go through from when it's exposed to when it's available 4516 to the framework.</description> 4517 <details>A typical minimum value for this is 2 (one stage to expose, 4518 one stage to readout) from the sensor. The ISP then usually adds 4519 its own stages to do custom HW processing. Further stages may be 4520 added by SW processing. 4521 4522 Depending on what settings are used (e.g. YUV, JPEG) and what 4523 processing is enabled (e.g. face detection), the actual pipeline 4524 depth (specified by android.request.pipelineDepth) may be less than 4525 the max pipeline depth. 4526 4527 A pipeline depth of X stages is equivalent to a pipeline latency of 4528 X frame intervals. 4529 4530 This value will normally be 8 or less, however, for high speed capture session, 4531 the max pipeline depth will be up to 8 x size of high speed capture request list. 4532 </details> 4533 <hal_details> 4534 This value should be 4 or less, expect for the high speed recording session, where the 4535 max batch sizes may be larger than 1. 4536 </hal_details> 4537 </entry> 4538 <entry name="partialResultCount" type="int32" visibility="public" optional="true"> 4539 <description>Defines how many sub-components 4540 a result will be composed of. 4541 </description> 4542 <range>&gt;= 1</range> 4543 <details>In order to combat the pipeline latency, partial results 4544 may be delivered to the application layer from the camera device as 4545 soon as they are available. 4546 4547 Optional; defaults to 1. A value of 1 means that partial 4548 results are not supported, and only the final TotalCaptureResult will 4549 be produced by the camera device. 4550 4551 A typical use case for this might be: after requesting an 4552 auto-focus (AF) lock the new AF state might be available 50% 4553 of the way through the pipeline. The camera device could 4554 then immediately dispatch this state via a partial result to 4555 the application, and the rest of the metadata via later 4556 partial results. 4557 </details> 4558 </entry> 4559 <entry name="availableCapabilities" type="byte" visibility="public" 4560 enum="true" container="array" hwlevel="legacy"> 4561 <array> 4562 <size>n</size> 4563 </array> 4564 <enum> 4565 <value>BACKWARD_COMPATIBLE 4566 <notes>The minimal set of capabilities that every camera 4567 device (regardless of android.info.supportedHardwareLevel) 4568 supports. 4569 4570 This capability is listed by all normal devices, and 4571 indicates that the camera device has a feature set 4572 that's comparable to the baseline requirements for the 4573 older android.hardware.Camera API. 4574 4575 Devices with the DEPTH_OUTPUT capability might not list this 4576 capability, indicating that they support only depth measurement, 4577 not standard color output. 4578 </notes> 4579 </value> 4580 <value optional="true">MANUAL_SENSOR 4581 <notes> 4582 The camera device can be manually controlled (3A algorithms such 4583 as auto-exposure, and auto-focus can be bypassed). 4584 The camera device supports basic manual control of the sensor image 4585 acquisition related stages. This means the following controls are 4586 guaranteed to be supported: 4587 4588 * Manual frame duration control 4589 * android.sensor.frameDuration 4590 * android.sensor.info.maxFrameDuration 4591 * Manual exposure control 4592 * android.sensor.exposureTime 4593 * android.sensor.info.exposureTimeRange 4594 * Manual sensitivity control 4595 * android.sensor.sensitivity 4596 * android.sensor.info.sensitivityRange 4597 * Manual lens control (if the lens is adjustable) 4598 * android.lens.* 4599 * Manual flash control (if a flash unit is present) 4600 * android.flash.* 4601 * Manual black level locking 4602 * android.blackLevel.lock 4603 * Auto exposure lock 4604 * android.control.aeLock 4605 4606 If any of the above 3A algorithms are enabled, then the camera 4607 device will accurately report the values applied by 3A in the 4608 result. 4609 4610 A given camera device may also support additional manual sensor controls, 4611 but this capability only covers the above list of controls. 4612 4613 If this is supported, android.scaler.streamConfigurationMap will 4614 additionally return a min frame duration that is greater than 4615 zero for each supported size-format combination. 4616 </notes> 4617 </value> 4618 <value optional="true">MANUAL_POST_PROCESSING 4619 <notes> 4620 The camera device post-processing stages can be manually controlled. 4621 The camera device supports basic manual control of the image post-processing 4622 stages. This means the following controls are guaranteed to be supported: 4623 4624 * Manual tonemap control 4625 * android.tonemap.curve 4626 * android.tonemap.mode 4627 * android.tonemap.maxCurvePoints 4628 * android.tonemap.gamma 4629 * android.tonemap.presetCurve 4630 4631 * Manual white balance control 4632 * android.colorCorrection.transform 4633 * android.colorCorrection.gains 4634 * Manual lens shading map control 4635 * android.shading.mode 4636 * android.statistics.lensShadingMapMode 4637 * android.statistics.lensShadingMap 4638 * android.lens.info.shadingMapSize 4639 * Manual aberration correction control (if aberration correction is supported) 4640 * android.colorCorrection.aberrationMode 4641 * android.colorCorrection.availableAberrationModes 4642 * Auto white balance lock 4643 * android.control.awbLock 4644 4645 If auto white balance is enabled, then the camera device 4646 will accurately report the values applied by AWB in the result. 4647 4648 A given camera device may also support additional post-processing 4649 controls, but this capability only covers the above list of controls. 4650 </notes> 4651 </value> 4652 <value optional="true">RAW 4653 <notes> 4654 The camera device supports outputting RAW buffers and 4655 metadata for interpreting them. 4656 4657 Devices supporting the RAW capability allow both for 4658 saving DNG files, and for direct application processing of 4659 raw sensor images. 4660 4661 * RAW_SENSOR is supported as an output format. 4662 * The maximum available resolution for RAW_SENSOR streams 4663 will match either the value in 4664 android.sensor.info.pixelArraySize or 4665 android.sensor.info.preCorrectionActiveArraySize. 4666 * All DNG-related optional metadata entries are provided 4667 by the camera device. 4668 </notes> 4669 </value> 4670 <value optional="true" ndk_hidden="true">PRIVATE_REPROCESSING 4671 <notes> 4672 The camera device supports the Zero Shutter Lag reprocessing use case. 4673 4674 * One input stream is supported, that is, `android.request.maxNumInputStreams == 1`. 4675 * {@link android.graphics.ImageFormat#PRIVATE} is supported as an output/input format, 4676 that is, {@link android.graphics.ImageFormat#PRIVATE} is included in the lists of 4677 formats returned by {@link 4678 android.hardware.camera2.params.StreamConfigurationMap#getInputFormats} and {@link 4679 android.hardware.camera2.params.StreamConfigurationMap#getOutputFormats}. 4680 * {@link android.hardware.camera2.params.StreamConfigurationMap#getValidOutputFormatsForInput} 4681 returns non empty int[] for each supported input format returned by {@link 4682 android.hardware.camera2.params.StreamConfigurationMap#getInputFormats}. 4683 * Each size returned by {@link 4684 android.hardware.camera2.params.StreamConfigurationMap#getInputSizes 4685 getInputSizes(ImageFormat.PRIVATE)} is also included in {@link 4686 android.hardware.camera2.params.StreamConfigurationMap#getOutputSizes 4687 getOutputSizes(ImageFormat.PRIVATE)} 4688 * Using {@link android.graphics.ImageFormat#PRIVATE} does not cause a frame rate drop 4689 relative to the sensor's maximum capture rate (at that resolution). 4690 * {@link android.graphics.ImageFormat#PRIVATE} will be reprocessable into both 4691 {@link android.graphics.ImageFormat#YUV_420_888} and 4692 {@link android.graphics.ImageFormat#JPEG} formats. 4693 * The maximum available resolution for PRIVATE streams 4694 (both input/output) will match the maximum available 4695 resolution of JPEG streams. 4696 * Static metadata android.reprocess.maxCaptureStall. 4697 * Only below controls are effective for reprocessing requests and 4698 will be present in capture results, other controls in reprocess 4699 requests will be ignored by the camera device. 4700 * android.jpeg.* 4701 * android.noiseReduction.mode 4702 * android.edge.mode 4703 * android.noiseReduction.availableNoiseReductionModes and 4704 android.edge.availableEdgeModes will both list ZERO_SHUTTER_LAG as a supported mode. 4705 </notes> 4706 </value> 4707 <value optional="true">READ_SENSOR_SETTINGS 4708 <notes> 4709 The camera device supports accurately reporting the sensor settings for many of 4710 the sensor controls while the built-in 3A algorithm is running. This allows 4711 reporting of sensor settings even when these settings cannot be manually changed. 4712 4713 The values reported for the following controls are guaranteed to be available 4714 in the CaptureResult, including when 3A is enabled: 4715 4716 * Exposure control 4717 * android.sensor.exposureTime 4718 * Sensitivity control 4719 * android.sensor.sensitivity 4720 * Lens controls (if the lens is adjustable) 4721 * android.lens.focusDistance 4722 * android.lens.aperture 4723 4724 This capability is a subset of the MANUAL_SENSOR control capability, and will 4725 always be included if the MANUAL_SENSOR capability is available. 4726 </notes> 4727 </value> 4728 <value optional="true">BURST_CAPTURE 4729 <notes> 4730 The camera device supports capturing high-resolution images at >= 20 frames per 4731 second, in at least the uncompressed YUV format, when post-processing settings are set 4732 to FAST. Additionally, maximum-resolution images can be captured at >= 10 frames 4733 per second. Here, 'high resolution' means at least 8 megapixels, or the maximum 4734 resolution of the device, whichever is smaller. 4735 4736 More specifically, this means that a size matching the camera device's active array 4737 size is listed as a supported size for the {@link 4738 android.graphics.ImageFormat#YUV_420_888} format in either {@link 4739 android.hardware.camera2.params.StreamConfigurationMap#getOutputSizes} or {@link 4740 android.hardware.camera2.params.StreamConfigurationMap#getHighResolutionOutputSizes}, 4741 with a minimum frame duration for that format and size of either <= 1/20 s, or 4742 <= 1/10 s, respectively; and the android.control.aeAvailableTargetFpsRanges entry 4743 lists at least one FPS range where the minimum FPS is >= 1 / minimumFrameDuration 4744 for the maximum-size YUV_420_888 format. If that maximum size is listed in {@link 4745 android.hardware.camera2.params.StreamConfigurationMap#getHighResolutionOutputSizes}, 4746 then the list of resolutions for YUV_420_888 from {@link 4747 android.hardware.camera2.params.StreamConfigurationMap#getOutputSizes} contains at 4748 least one resolution >= 8 megapixels, with a minimum frame duration of <= 1/20 4749 s. 4750 4751 If the device supports the {@link android.graphics.ImageFormat#RAW10}, {@link 4752 android.graphics.ImageFormat#RAW12}, then those can also be captured at the same rate 4753 as the maximum-size YUV_420_888 resolution is. 4754 4755 If the device supports the PRIVATE_REPROCESSING capability, then the same guarantees 4756 as for the YUV_420_888 format also apply to the {@link 4757 android.graphics.ImageFormat#PRIVATE} format. 4758 4759 In addition, the android.sync.maxLatency field is guaranted to have a value between 0 4760 and 4, inclusive. android.control.aeLockAvailable and android.control.awbLockAvailable 4761 are also guaranteed to be `true` so burst capture with these two locks ON yields 4762 consistent image output. 4763 </notes> 4764 </value> 4765 <value optional="true" ndk_hidden="true">YUV_REPROCESSING 4766 <notes> 4767 The camera device supports the YUV_420_888 reprocessing use case, similar as 4768 PRIVATE_REPROCESSING, This capability requires the camera device to support the 4769 following: 4770 4771 * One input stream is supported, that is, `android.request.maxNumInputStreams == 1`. 4772 * {@link android.graphics.ImageFormat#YUV_420_888} is supported as an output/input format, that is, 4773 YUV_420_888 is included in the lists of formats returned by 4774 {@link android.hardware.camera2.params.StreamConfigurationMap#getInputFormats} and 4775 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputFormats}. 4776 * {@link 4777 android.hardware.camera2.params.StreamConfigurationMap#getValidOutputFormatsForInput} 4778 returns non-empty int[] for each supported input format returned by {@link 4779 android.hardware.camera2.params.StreamConfigurationMap#getInputFormats}. 4780 * Each size returned by {@link 4781 android.hardware.camera2.params.StreamConfigurationMap#getInputSizes 4782 getInputSizes(YUV_420_888)} is also included in {@link 4783 android.hardware.camera2.params.StreamConfigurationMap#getOutputSizes 4784 getOutputSizes(YUV_420_888)} 4785 * Using {@link android.graphics.ImageFormat#YUV_420_888} does not cause a frame rate drop 4786 relative to the sensor's maximum capture rate (at that resolution). 4787 * {@link android.graphics.ImageFormat#YUV_420_888} will be reprocessable into both 4788 {@link android.graphics.ImageFormat#YUV_420_888} and {@link 4789 android.graphics.ImageFormat#JPEG} formats. 4790 * The maximum available resolution for {@link 4791 android.graphics.ImageFormat#YUV_420_888} streams (both input/output) will match the 4792 maximum available resolution of {@link android.graphics.ImageFormat#JPEG} streams. 4793 * Static metadata android.reprocess.maxCaptureStall. 4794 * Only the below controls are effective for reprocessing requests and will be present 4795 in capture results. The reprocess requests are from the original capture results that 4796 are associated with the intermediate {@link android.graphics.ImageFormat#YUV_420_888} 4797 output buffers. All other controls in the reprocess requests will be ignored by the 4798 camera device. 4799 * android.jpeg.* 4800 * android.noiseReduction.mode 4801 * android.edge.mode 4802 * android.reprocess.effectiveExposureFactor 4803 * android.noiseReduction.availableNoiseReductionModes and 4804 android.edge.availableEdgeModes will both list ZERO_SHUTTER_LAG as a supported mode. 4805 </notes> 4806 </value> 4807 <value optional="true">DEPTH_OUTPUT 4808 <notes> 4809 The camera device can produce depth measurements from its field of view. 4810 4811 This capability requires the camera device to support the following: 4812 4813 * {@link android.graphics.ImageFormat#DEPTH16} is supported as an output format. 4814 * {@link android.graphics.ImageFormat#DEPTH_POINT_CLOUD} is optionally supported as an 4815 output format. 4816 * This camera device, and all camera devices with the same android.lens.facing, 4817 will list the following calibration entries in both 4818 {@link android.hardware.camera2.CameraCharacteristics} and 4819 {@link android.hardware.camera2.CaptureResult}: 4820 - android.lens.poseTranslation 4821 - android.lens.poseRotation 4822 - android.lens.intrinsicCalibration 4823 - android.lens.radialDistortion 4824 * The android.depth.depthIsExclusive entry is listed by this device. 4825 * A LIMITED camera with only the DEPTH_OUTPUT capability does not have to support 4826 normal YUV_420_888, JPEG, and PRIV-format outputs. It only has to support the DEPTH16 4827 format. 4828 4829 Generally, depth output operates at a slower frame rate than standard color capture, 4830 so the DEPTH16 and DEPTH_POINT_CLOUD formats will commonly have a stall duration that 4831 should be accounted for (see 4832 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration}). 4833 On a device that supports both depth and color-based output, to enable smooth preview, 4834 using a repeating burst is recommended, where a depth-output target is only included 4835 once every N frames, where N is the ratio between preview output rate and depth output 4836 rate, including depth stall time. 4837 </notes> 4838 </value> 4839 <value optional="true" ndk_hidden="true">CONSTRAINED_HIGH_SPEED_VIDEO 4840 <notes> 4841 The device supports constrained high speed video recording (frame rate >=120fps) 4842 use case. The camera device will support high speed capture session created by 4843 {@link android.hardware.camera2.CameraDevice#createConstrainedHighSpeedCaptureSession}, which 4844 only accepts high speed request lists created by 4845 {@link android.hardware.camera2.CameraConstrainedHighSpeedCaptureSession#createHighSpeedRequestList}. 4846 4847 A camera device can still support high speed video streaming by advertising the high speed 4848 FPS ranges in android.control.aeAvailableTargetFpsRanges. For this case, all normal 4849 capture request per frame control and synchronization requirements will apply to 4850 the high speed fps ranges, the same as all other fps ranges. This capability describes 4851 the capability of a specialized operating mode with many limitations (see below), which 4852 is only targeted at high speed video recording. 4853 4854 The supported high speed video sizes and fps ranges are specified in 4855 {@link android.hardware.camera2.params.StreamConfigurationMap#getHighSpeedVideoFpsRanges}. 4856 To get desired output frame rates, the application is only allowed to select video size 4857 and FPS range combinations provided by 4858 {@link android.hardware.camera2.params.StreamConfigurationMap#getHighSpeedVideoSizes}. 4859 The fps range can be controlled via android.control.aeTargetFpsRange. 4860 4861 In this capability, the camera device will override aeMode, awbMode, and afMode to 4862 ON, AUTO, and CONTINUOUS_VIDEO, respectively. All post-processing block mode 4863 controls will be overridden to be FAST. Therefore, no manual control of capture 4864 and post-processing parameters is possible. All other controls operate the 4865 same as when android.control.mode == AUTO. This means that all other 4866 android.control.* fields continue to work, such as 4867 4868 * android.control.aeTargetFpsRange 4869 * android.control.aeExposureCompensation 4870 * android.control.aeLock 4871 * android.control.awbLock 4872 * android.control.effectMode 4873 * android.control.aeRegions 4874 * android.control.afRegions 4875 * android.control.awbRegions 4876 * android.control.afTrigger 4877 * android.control.aePrecaptureTrigger 4878 4879 Outside of android.control.*, the following controls will work: 4880 4881 * android.flash.mode (TORCH mode only, automatic flash for still capture will not 4882 work since aeMode is ON) 4883 * android.lens.opticalStabilizationMode (if it is supported) 4884 * android.scaler.cropRegion 4885 * android.statistics.faceDetectMode (if it is supported) 4886 4887 For high speed recording use case, the actual maximum supported frame rate may 4888 be lower than what camera can output, depending on the destination Surfaces for 4889 the image data. For example, if the destination surface is from video encoder, 4890 the application need check if the video encoder is capable of supporting the 4891 high frame rate for a given video size, or it will end up with lower recording 4892 frame rate. If the destination surface is from preview window, the actual preview frame 4893 rate will be bounded by the screen refresh rate. 4894 4895 The camera device will only support up to 2 high speed simultaneous output surfaces 4896 (preview and recording surfaces) 4897 in this mode. Above controls will be effective only if all of below conditions are true: 4898 4899 * The application creates a camera capture session with no more than 2 surfaces via 4900 {@link android.hardware.camera2.CameraDevice#createConstrainedHighSpeedCaptureSession}. The 4901 targeted surfaces must be preview surface (either from 4902 {@link android.view.SurfaceView} or {@link android.graphics.SurfaceTexture}) or 4903 recording surface(either from {@link android.media.MediaRecorder#getSurface} or 4904 {@link android.media.MediaCodec#createInputSurface}). 4905 * The stream sizes are selected from the sizes reported by 4906 {@link android.hardware.camera2.params.StreamConfigurationMap#getHighSpeedVideoSizes}. 4907 * The FPS ranges are selected from 4908 {@link android.hardware.camera2.params.StreamConfigurationMap#getHighSpeedVideoFpsRanges}. 4909 4910 When above conditions are NOT satistied, 4911 {@link android.hardware.camera2.CameraDevice#createConstrainedHighSpeedCaptureSession} 4912 will fail. 4913 4914 Switching to a FPS range that has different maximum FPS may trigger some camera device 4915 reconfigurations, which may introduce extra latency. It is recommended that 4916 the application avoids unnecessary maximum target FPS changes as much as possible 4917 during high speed streaming. 4918 </notes> 4919 </value> 4920 </enum> 4921 <description>List of capabilities that this camera device 4922 advertises as fully supporting.</description> 4923 <details> 4924 A capability is a contract that the camera device makes in order 4925 to be able to satisfy one or more use cases. 4926 4927 Listing a capability guarantees that the whole set of features 4928 required to support a common use will all be available. 4929 4930 Using a subset of the functionality provided by an unsupported 4931 capability may be possible on a specific camera device implementation; 4932 to do this query each of android.request.availableRequestKeys, 4933 android.request.availableResultKeys, 4934 android.request.availableCharacteristicsKeys. 4935 4936 The following capabilities are guaranteed to be available on 4937 android.info.supportedHardwareLevel `==` FULL devices: 4938 4939 * MANUAL_SENSOR 4940 * MANUAL_POST_PROCESSING 4941 4942 Other capabilities may be available on either FULL or LIMITED 4943 devices, but the application should query this key to be sure. 4944 </details> 4945 <hal_details> 4946 Additional constraint details per-capability will be available 4947 in the Compatibility Test Suite. 4948 4949 Minimum baseline requirements required for the 4950 BACKWARD_COMPATIBLE capability are not explicitly listed. 4951 Instead refer to "BC" tags and the camera CTS tests in the 4952 android.hardware.camera2.cts package. 4953 4954 Listed controls that can be either request or result (e.g. 4955 android.sensor.exposureTime) must be available both in the 4956 request and the result in order to be considered to be 4957 capability-compliant. 4958 4959 For example, if the HAL claims to support MANUAL control, 4960 then exposure time must be configurable via the request _and_ 4961 the actual exposure applied must be available via 4962 the result. 4963 4964 If MANUAL_SENSOR is omitted, the HAL may choose to omit the 4965 android.scaler.availableMinFrameDurations static property entirely. 4966 4967 For PRIVATE_REPROCESSING and YUV_REPROCESSING capabilities, see 4968 hardware/libhardware/include/hardware/camera3.h Section 10 for more information. 4969 4970 Devices that support the MANUAL_SENSOR capability must support the 4971 CAMERA3_TEMPLATE_MANUAL template defined in camera3.h. 4972 4973 Devices that support the PRIVATE_REPROCESSING capability or the 4974 YUV_REPROCESSING capability must support the 4975 CAMERA3_TEMPLATE_ZERO_SHUTTER_LAG template defined in camera3.h. 4976 4977 For DEPTH_OUTPUT, the depth-format keys 4978 android.depth.availableDepthStreamConfigurations, 4979 android.depth.availableDepthMinFrameDurations, 4980 android.depth.availableDepthStallDurations must be available, in 4981 addition to the other keys explicitly mentioned in the DEPTH_OUTPUT 4982 enum notes. The entry android.depth.maxDepthSamples must be available 4983 if the DEPTH_POINT_CLOUD format is supported (HAL pixel format BLOB, dataspace 4984 DEPTH). 4985 </hal_details> 4986 </entry> 4987 <entry name="availableRequestKeys" type="int32" visibility="ndk_public" 4988 container="array" hwlevel="legacy"> 4989 <array> 4990 <size>n</size> 4991 </array> 4992 <description>A list of all keys that the camera device has available 4993 to use with {@link android.hardware.camera2.CaptureRequest}.</description> 4994 4995 <details>Attempting to set a key into a CaptureRequest that is not 4996 listed here will result in an invalid request and will be rejected 4997 by the camera device. 4998 4999 This field can be used to query the feature set of a camera device 5000 at a more granular level than capabilities. This is especially 5001 important for optional keys that are not listed under any capability 5002 in android.request.availableCapabilities. 5003 </details> 5004 <hal_details> 5005 Vendor tags must not be listed here. Use the vendor tag metadata 5006 extensions C api instead (refer to camera3.h for more details). 5007 5008 Setting/getting vendor tags will be checked against the metadata 5009 vendor extensions API and not against this field. 5010 5011 The HAL must not consume any request tags that are not listed either 5012 here or in the vendor tag list. 5013 5014 The public camera2 API will always make the vendor tags visible 5015 via 5016 {@link android.hardware.camera2.CameraCharacteristics#getAvailableCaptureRequestKeys}. 5017 </hal_details> 5018 </entry> 5019 <entry name="availableResultKeys" type="int32" visibility="ndk_public" 5020 container="array" hwlevel="legacy"> 5021 <array> 5022 <size>n</size> 5023 </array> 5024 <description>A list of all keys that the camera device has available 5025 to use with {@link android.hardware.camera2.CaptureResult}.</description> 5026 5027 <details>Attempting to get a key from a CaptureResult that is not 5028 listed here will always return a `null` value. Getting a key from 5029 a CaptureResult that is listed here will generally never return a `null` 5030 value. 5031 5032 The following keys may return `null` unless they are enabled: 5033 5034 * android.statistics.lensShadingMap (non-null iff android.statistics.lensShadingMapMode == ON) 5035 5036 (Those sometimes-null keys will nevertheless be listed here 5037 if they are available.) 5038 5039 This field can be used to query the feature set of a camera device 5040 at a more granular level than capabilities. This is especially 5041 important for optional keys that are not listed under any capability 5042 in android.request.availableCapabilities. 5043 </details> 5044 <hal_details> 5045 Tags listed here must always have an entry in the result metadata, 5046 even if that size is 0 elements. Only array-type tags (e.g. lists, 5047 matrices, strings) are allowed to have 0 elements. 5048 5049 Vendor tags must not be listed here. Use the vendor tag metadata 5050 extensions C api instead (refer to camera3.h for more details). 5051 5052 Setting/getting vendor tags will be checked against the metadata 5053 vendor extensions API and not against this field. 5054 5055 The HAL must not produce any result tags that are not listed either 5056 here or in the vendor tag list. 5057 5058 The public camera2 API will always make the vendor tags visible via {@link 5059 android.hardware.camera2.CameraCharacteristics#getAvailableCaptureResultKeys}. 5060 </hal_details> 5061 </entry> 5062 <entry name="availableCharacteristicsKeys" type="int32" visibility="ndk_public" 5063 container="array" hwlevel="legacy"> 5064 <array> 5065 <size>n</size> 5066 </array> 5067 <description>A list of all keys that the camera device has available 5068 to use with {@link android.hardware.camera2.CameraCharacteristics}.</description> 5069 <details>This entry follows the same rules as 5070 android.request.availableResultKeys (except that it applies for 5071 CameraCharacteristics instead of CaptureResult). See above for more 5072 details. 5073 </details> 5074 <hal_details> 5075 Keys listed here must always have an entry in the static info metadata, 5076 even if that size is 0 elements. Only array-type tags (e.g. lists, 5077 matrices, strings) are allowed to have 0 elements. 5078 5079 Vendor tags must not be listed here. Use the vendor tag metadata 5080 extensions C api instead (refer to camera3.h for more details). 5081 5082 Setting/getting vendor tags will be checked against the metadata 5083 vendor extensions API and not against this field. 5084 5085 The HAL must not have any tags in its static info that are not listed 5086 either here or in the vendor tag list. 5087 5088 The public camera2 API will always make the vendor tags visible 5089 via {@link android.hardware.camera2.CameraCharacteristics#getKeys}. 5090 </hal_details> 5091 </entry> 5092 </static> 5093 </section> 5094 <section name="scaler"> 5095 <controls> 5096 <entry name="cropRegion" type="int32" visibility="public" 5097 container="array" typedef="rectangle" hwlevel="legacy"> 5098 <array> 5099 <size>4</size> 5100 </array> 5101 <description>The desired region of the sensor to read out for this capture.</description> 5102 <units>Pixel coordinates relative to 5103 android.sensor.info.activeArraySize</units> 5104 <details> 5105 This control can be used to implement digital zoom. 5106 5107 The crop region coordinate system is based off 5108 android.sensor.info.activeArraySize, with `(0, 0)` being the 5109 top-left corner of the sensor active array. 5110 5111 Output streams use this rectangle to produce their output, 5112 cropping to a smaller region if necessary to maintain the 5113 stream's aspect ratio, then scaling the sensor input to 5114 match the output's configured resolution. 5115 5116 The crop region is applied after the RAW to other color 5117 space (e.g. YUV) conversion. Since raw streams 5118 (e.g. RAW16) don't have the conversion stage, they are not 5119 croppable. The crop region will be ignored by raw streams. 5120 5121 For non-raw streams, any additional per-stream cropping will 5122 be done to maximize the final pixel area of the stream. 5123 5124 For example, if the crop region is set to a 4:3 aspect 5125 ratio, then 4:3 streams will use the exact crop 5126 region. 16:9 streams will further crop vertically 5127 (letterbox). 5128 5129 Conversely, if the crop region is set to a 16:9, then 4:3 5130 outputs will crop horizontally (pillarbox), and 16:9 5131 streams will match exactly. These additional crops will 5132 be centered within the crop region. 5133 5134 The width and height of the crop region cannot 5135 be set to be smaller than 5136 `floor( activeArraySize.width / android.scaler.availableMaxDigitalZoom )` and 5137 `floor( activeArraySize.height / android.scaler.availableMaxDigitalZoom )`, respectively. 5138 5139 The camera device may adjust the crop region to account 5140 for rounding and other hardware requirements; the final 5141 crop region used will be included in the output capture 5142 result. 5143 </details> 5144 <hal_details> 5145 The output streams must maintain square pixels at all 5146 times, no matter what the relative aspect ratios of the 5147 crop region and the stream are. Negative values for 5148 corner are allowed for raw output if full pixel array is 5149 larger than active pixel array. Width and height may be 5150 rounded to nearest larger supportable width, especially 5151 for raw output, where only a few fixed scales may be 5152 possible. 5153 5154 For a set of output streams configured, if the sensor output is cropped to a smaller 5155 size than active array size, the HAL need follow below cropping rules: 5156 5157 * The HAL need handle the cropRegion as if the sensor crop size is the effective active 5158 array size.More specifically, the HAL must transform the request cropRegion from 5159 android.sensor.info.activeArraySize to the sensor cropped pixel area size in this way: 5160 1. Translate the requested cropRegion w.r.t., the left top corner of the sensor 5161 cropped pixel area by (tx, ty), 5162 where `tx = sensorCrop.top * (sensorCrop.height / activeArraySize.height)` 5163 and `tx = sensorCrop.left * (sensorCrop.width / activeArraySize.width)`. The 5164 (sensorCrop.top, sensorCrop.left) is the coordinate based off the 5165 android.sensor.info.activeArraySize. 5166 2. Scale the width and height of requested cropRegion with scaling factor of 5167 sensorCrop.width/activeArraySize.width and sensorCrop.height/activeArraySize.height 5168 respectively. 5169 Once this new cropRegion is calculated, the HAL must use this region to crop the image 5170 with regard to the sensor crop size (effective active array size). The HAL still need 5171 follow the general cropping rule for this new cropRegion and effective active 5172 array size. 5173 5174 * The HAL must report the cropRegion with regard to android.sensor.info.activeArraySize. 5175 The HAL need convert the new cropRegion generated above w.r.t., full active array size. 5176 The reported cropRegion may be slightly different with the requested cropRegion since 5177 the HAL may adjust the crop region to account for rounding, conversion error, or other 5178 hardware limitations. 5179 5180 HAL2.x uses only (x, y, width) 5181 </hal_details> 5182 <tag id="BC" /> 5183 </entry> 5184 </controls> 5185 <static> 5186 <entry name="availableFormats" type="int32" 5187 visibility="hidden" deprecated="true" enum="true" 5188 container="array" typedef="imageFormat"> 5189 <array> 5190 <size>n</size> 5191 </array> 5192 <enum> 5193 <value optional="true" id="0x20">RAW16 5194 <notes> 5195 RAW16 is a standard, cross-platform format for raw image 5196 buffers with 16-bit pixels. 5197 5198 Buffers of this format are typically expected to have a 5199 Bayer Color Filter Array (CFA) layout, which is given in 5200 android.sensor.info.colorFilterArrangement. Sensors with 5201 CFAs that are not representable by a format in 5202 android.sensor.info.colorFilterArrangement should not 5203 use this format. 5204 5205 Buffers of this format will also follow the constraints given for 5206 RAW_OPAQUE buffers, but with relaxed performance constraints. 5207 5208 This format is intended to give users access to the full contents 5209 of the buffers coming directly from the image sensor prior to any 5210 cropping or scaling operations, and all coordinate systems for 5211 metadata used for this format are relative to the size of the 5212 active region of the image sensor before any geometric distortion 5213 correction has been applied (i.e. 5214 android.sensor.info.preCorrectionActiveArraySize). Supported 5215 dimensions for this format are limited to the full dimensions of 5216 the sensor (e.g. either android.sensor.info.pixelArraySize or 5217 android.sensor.info.preCorrectionActiveArraySize will be the 5218 only supported output size). 5219 5220 See android.scaler.availableInputOutputFormatsMap for 5221 the full set of performance guarantees. 5222 </notes> 5223 </value> 5224 <value optional="true" id="0x24">RAW_OPAQUE 5225 <notes> 5226 RAW_OPAQUE (or 5227 {@link android.graphics.ImageFormat#RAW_PRIVATE RAW_PRIVATE} 5228 as referred in public API) is a format for raw image buffers 5229 coming from an image sensor. 5230 5231 The actual structure of buffers of this format is 5232 platform-specific, but must follow several constraints: 5233 5234 1. No image post-processing operations may have been applied to 5235 buffers of this type. These buffers contain raw image data coming 5236 directly from the image sensor. 5237 1. If a buffer of this format is passed to the camera device for 5238 reprocessing, the resulting images will be identical to the images 5239 produced if the buffer had come directly from the sensor and was 5240 processed with the same settings. 5241 5242 The intended use for this format is to allow access to the native 5243 raw format buffers coming directly from the camera sensor without 5244 any additional conversions or decrease in framerate. 5245 5246 See android.scaler.availableInputOutputFormatsMap for the full set of 5247 performance guarantees. 5248 </notes> 5249 </value> 5250 <value optional="true" id="0x32315659">YV12 5251 <notes>YCrCb 4:2:0 Planar</notes> 5252 </value> 5253 <value optional="true" id="0x11">YCrCb_420_SP 5254 <notes>NV21</notes> 5255 </value> 5256 <value id="0x22">IMPLEMENTATION_DEFINED 5257 <notes>System internal format, not application-accessible</notes> 5258 </value> 5259 <value id="0x23">YCbCr_420_888 5260 <notes>Flexible YUV420 Format</notes> 5261 </value> 5262 <value id="0x21">BLOB 5263 <notes>JPEG format</notes> 5264 </value> 5265 </enum> 5266 <description>The list of image formats that are supported by this 5267 camera device for output streams.</description> 5268 <details> 5269 All camera devices will support JPEG and YUV_420_888 formats. 5270 5271 When set to YUV_420_888, application can access the YUV420 data directly. 5272 </details> 5273 <hal_details> 5274 These format values are from HAL_PIXEL_FORMAT_* in 5275 system/core/include/system/graphics.h. 5276 5277 When IMPLEMENTATION_DEFINED is used, the platform 5278 gralloc module will select a format based on the usage flags provided 5279 by the camera HAL device and the other endpoint of the stream. It is 5280 usually used by preview and recording streams, where the application doesn't 5281 need access the image data. 5282 5283 YCbCr_420_888 format must be supported by the HAL. When an image stream 5284 needs CPU/application direct access, this format will be used. 5285 5286 The BLOB format must be supported by the HAL. This is used for the JPEG stream. 5287 5288 A RAW_OPAQUE buffer should contain only pixel data. It is strongly 5289 recommended that any information used by the camera device when 5290 processing images is fully expressed by the result metadata 5291 for that image buffer. 5292 </hal_details> 5293 <tag id="BC" /> 5294 </entry> 5295 <entry name="availableJpegMinDurations" type="int64" visibility="hidden" deprecated="true" 5296 container="array"> 5297 <array> 5298 <size>n</size> 5299 </array> 5300 <description>The minimum frame duration that is supported 5301 for each resolution in android.scaler.availableJpegSizes. 5302 </description> 5303 <units>Nanoseconds</units> 5304 <range>TODO: Remove property.</range> 5305 <details> 5306 This corresponds to the minimum steady-state frame duration when only 5307 that JPEG stream is active and captured in a burst, with all 5308 processing (typically in android.*.mode) set to FAST. 5309 5310 When multiple streams are configured, the minimum 5311 frame duration will be &gt;= max(individual stream min 5312 durations)</details> 5313 <tag id="BC" /> 5314 </entry> 5315 <entry name="availableJpegSizes" type="int32" visibility="hidden" 5316 deprecated="true" container="array" typedef="size"> 5317 <array> 5318 <size>n</size> 5319 <size>2</size> 5320 </array> 5321 <description>The JPEG resolutions that are supported by this camera device.</description> 5322 <range>TODO: Remove property.</range> 5323 <details> 5324 The resolutions are listed as `(width, height)` pairs. All camera devices will support 5325 sensor maximum resolution (defined by android.sensor.info.activeArraySize). 5326 </details> 5327 <hal_details> 5328 The HAL must include sensor maximum resolution 5329 (defined by android.sensor.info.activeArraySize), 5330 and should include half/quarter of sensor maximum resolution. 5331 </hal_details> 5332 <tag id="BC" /> 5333 </entry> 5334 <entry name="availableMaxDigitalZoom" type="float" visibility="public" 5335 hwlevel="legacy"> 5336 <description>The maximum ratio between both active area width 5337 and crop region width, and active area height and 5338 crop region height, for android.scaler.cropRegion. 5339 </description> 5340 <units>Zoom scale factor</units> 5341 <range>&gt;=1</range> 5342 <details> 5343 This represents the maximum amount of zooming possible by 5344 the camera device, or equivalently, the minimum cropping 5345 window size. 5346 5347 Crop regions that have a width or height that is smaller 5348 than this ratio allows will be rounded up to the minimum 5349 allowed size by the camera device. 5350 </details> 5351 <tag id="BC" /> 5352 </entry> 5353 <entry name="availableProcessedMinDurations" type="int64" visibility="hidden" deprecated="true" 5354 container="array"> 5355 <array> 5356 <size>n</size> 5357 </array> 5358 <description>For each available processed output size (defined in 5359 android.scaler.availableProcessedSizes), this property lists the 5360 minimum supportable frame duration for that size. 5361 </description> 5362 <units>Nanoseconds</units> 5363 <details> 5364 This should correspond to the frame duration when only that processed 5365 stream is active, with all processing (typically in android.*.mode) 5366 set to FAST. 5367 5368 When multiple streams are configured, the minimum frame duration will 5369 be &gt;= max(individual stream min durations). 5370 </details> 5371 <tag id="BC" /> 5372 </entry> 5373 <entry name="availableProcessedSizes" type="int32" visibility="hidden" 5374 deprecated="true" container="array" typedef="size"> 5375 <array> 5376 <size>n</size> 5377 <size>2</size> 5378 </array> 5379 <description>The resolutions available for use with 5380 processed output streams, such as YV12, NV12, and 5381 platform opaque YUV/RGB streams to the GPU or video 5382 encoders.</description> 5383 <details> 5384 The resolutions are listed as `(width, height)` pairs. 5385 5386 For a given use case, the actual maximum supported resolution 5387 may be lower than what is listed here, depending on the destination 5388 Surface for the image data. For example, for recording video, 5389 the video encoder chosen may have a maximum size limit (e.g. 1080p) 5390 smaller than what the camera (e.g. maximum resolution is 3264x2448) 5391 can provide. 5392 5393 Please reference the documentation for the image data destination to 5394 check if it limits the maximum size for image data. 5395 </details> 5396 <hal_details> 5397 For FULL capability devices (`android.info.supportedHardwareLevel == FULL`), 5398 the HAL must include all JPEG sizes listed in android.scaler.availableJpegSizes 5399 and each below resolution if it is smaller than or equal to the sensor 5400 maximum resolution (if they are not listed in JPEG sizes already): 5401 5402 * 240p (320 x 240) 5403 * 480p (640 x 480) 5404 * 720p (1280 x 720) 5405 * 1080p (1920 x 1080) 5406 5407 For LIMITED capability devices (`android.info.supportedHardwareLevel == LIMITED`), 5408 the HAL only has to list up to the maximum video size supported by the devices. 5409 </hal_details> 5410 <tag id="BC" /> 5411 </entry> 5412 <entry name="availableRawMinDurations" type="int64" deprecated="true" 5413 container="array"> 5414 <array> 5415 <size>n</size> 5416 </array> 5417 <description> 5418 For each available raw output size (defined in 5419 android.scaler.availableRawSizes), this property lists the minimum 5420 supportable frame duration for that size. 5421 </description> 5422 <units>Nanoseconds</units> 5423 <details> 5424 Should correspond to the frame duration when only the raw stream is 5425 active. 5426 5427 When multiple streams are configured, the minimum 5428 frame duration will be &gt;= max(individual stream min 5429 durations)</details> 5430 <tag id="BC" /> 5431 </entry> 5432 <entry name="availableRawSizes" type="int32" deprecated="true" 5433 container="array" typedef="size"> 5434 <array> 5435 <size>n</size> 5436 <size>2</size> 5437 </array> 5438 <description>The resolutions available for use with raw 5439 sensor output streams, listed as width, 5440 height</description> 5441 </entry> 5442 </static> 5443 <dynamic> 5444 <clone entry="android.scaler.cropRegion" kind="controls"> 5445 </clone> 5446 </dynamic> 5447 <static> 5448 <entry name="availableInputOutputFormatsMap" type="int32" visibility="hidden" 5449 typedef="reprocessFormatsMap"> 5450 <description>The mapping of image formats that are supported by this 5451 camera device for input streams, to their corresponding output formats. 5452 </description> 5453 <details> 5454 All camera devices with at least 1 5455 android.request.maxNumInputStreams will have at least one 5456 available input format. 5457 5458 The camera device will support the following map of formats, 5459 if its dependent capability (android.request.availableCapabilities) is supported: 5460 5461 Input Format | Output Format | Capability 5462 :-------------------------------------------------|:--------------------------------------------------|:---------- 5463 {@link android.graphics.ImageFormat#PRIVATE} | {@link android.graphics.ImageFormat#JPEG} | PRIVATE_REPROCESSING 5464 {@link android.graphics.ImageFormat#PRIVATE} | {@link android.graphics.ImageFormat#YUV_420_888} | PRIVATE_REPROCESSING 5465 {@link android.graphics.ImageFormat#YUV_420_888} | {@link android.graphics.ImageFormat#JPEG} | YUV_REPROCESSING 5466 {@link android.graphics.ImageFormat#YUV_420_888} | {@link android.graphics.ImageFormat#YUV_420_888} | YUV_REPROCESSING 5467 5468 PRIVATE refers to a device-internal format that is not directly application-visible. A 5469 PRIVATE input surface can be acquired by {@link android.media.ImageReader#newInstance} 5470 with {@link android.graphics.ImageFormat#PRIVATE} as the format. 5471 5472 For a PRIVATE_REPROCESSING-capable camera device, using the PRIVATE format as either input 5473 or output will never hurt maximum frame rate (i.e. {@link 5474 android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration 5475 getOutputStallDuration(ImageFormat.PRIVATE, size)} is always 0), 5476 5477 Attempting to configure an input stream with output streams not 5478 listed as available in this map is not valid. 5479 </details> 5480 <hal_details> 5481 For the formats, see `system/core/include/system/graphics.h` for a definition 5482 of the image format enumerations. The PRIVATE format refers to the 5483 HAL_PIXEL_FORMAT_IMPLEMENTATION_DEFINED format. The HAL could determine 5484 the actual format by using the gralloc usage flags. 5485 For ZSL use case in particular, the HAL could choose appropriate format (partially 5486 processed YUV or RAW based format) by checking the format and GRALLOC_USAGE_HW_CAMERA_ZSL. 5487 See camera3.h for more details. 5488 5489 This value is encoded as a variable-size array-of-arrays. 5490 The inner array always contains `[format, length, ...]` where 5491 `...` has `length` elements. An inner array is followed by another 5492 inner array if the total metadata entry size hasn't yet been exceeded. 5493 5494 A code sample to read/write this encoding (with a device that 5495 supports reprocessing IMPLEMENTATION_DEFINED to YUV_420_888, and JPEG, 5496 and reprocessing YUV_420_888 to YUV_420_888 and JPEG): 5497 5498 // reading 5499 int32_t* contents = &entry.i32[0]; 5500 for (size_t i = 0; i < entry.count; ) { 5501 int32_t format = contents[i++]; 5502 int32_t length = contents[i++]; 5503 int32_t output_formats[length]; 5504 memcpy(&output_formats[0], &contents[i], 5505 length * sizeof(int32_t)); 5506 i += length; 5507 } 5508 5509 // writing (static example, PRIVATE_REPROCESSING + YUV_REPROCESSING) 5510 int32_t[] contents = { 5511 IMPLEMENTATION_DEFINED, 2, YUV_420_888, BLOB, 5512 YUV_420_888, 2, YUV_420_888, BLOB, 5513 }; 5514 update_camera_metadata_entry(metadata, index, &contents[0], 5515 sizeof(contents)/sizeof(contents[0]), &updated_entry); 5516 5517 If the HAL claims to support any of the capabilities listed in the 5518 above details, then it must also support all the input-output 5519 combinations listed for that capability. It can optionally support 5520 additional formats if it so chooses. 5521 </hal_details> 5522 <tag id="REPROC" /> 5523 </entry> 5524 <entry name="availableStreamConfigurations" type="int32" visibility="ndk_public" 5525 enum="true" container="array" typedef="streamConfiguration" hwlevel="legacy"> 5526 <array> 5527 <size>n</size> 5528 <size>4</size> 5529 </array> 5530 <enum> 5531 <value>OUTPUT</value> 5532 <value>INPUT</value> 5533 </enum> 5534 <description>The available stream configurations that this 5535 camera device supports 5536 (i.e. format, width, height, output/input stream). 5537 </description> 5538 <details> 5539 The configurations are listed as `(format, width, height, input?)` 5540 tuples. 5541 5542 For a given use case, the actual maximum supported resolution 5543 may be lower than what is listed here, depending on the destination 5544 Surface for the image data. For example, for recording video, 5545 the video encoder chosen may have a maximum size limit (e.g. 1080p) 5546 smaller than what the camera (e.g. maximum resolution is 3264x2448) 5547 can provide. 5548 5549 Please reference the documentation for the image data destination to 5550 check if it limits the maximum size for image data. 5551 5552 Not all output formats may be supported in a configuration with 5553 an input stream of a particular format. For more details, see 5554 android.scaler.availableInputOutputFormatsMap. 5555 5556 The following table describes the minimum required output stream 5557 configurations based on the hardware level 5558 (android.info.supportedHardwareLevel): 5559 5560 Format | Size | Hardware Level | Notes 5561 :-------------:|:--------------------------------------------:|:--------------:|:--------------: 5562 JPEG | android.sensor.info.activeArraySize | Any | 5563 JPEG | 1920x1080 (1080p) | Any | if 1080p <= activeArraySize 5564 JPEG | 1280x720 (720) | Any | if 720p <= activeArraySize 5565 JPEG | 640x480 (480p) | Any | if 480p <= activeArraySize 5566 JPEG | 320x240 (240p) | Any | if 240p <= activeArraySize 5567 YUV_420_888 | all output sizes available for JPEG | FULL | 5568 YUV_420_888 | all output sizes available for JPEG, up to the maximum video size | LIMITED | 5569 IMPLEMENTATION_DEFINED | same as YUV_420_888 | Any | 5570 5571 Refer to android.request.availableCapabilities for additional 5572 mandatory stream configurations on a per-capability basis. 5573 </details> 5574 <hal_details> 5575 It is recommended (but not mandatory) to also include half/quarter 5576 of sensor maximum resolution for JPEG formats (regardless of hardware 5577 level). 5578 5579 (The following is a rewording of the above required table): 5580 5581 For JPEG format, the sizes may be restricted by below conditions: 5582 5583 * The HAL may choose the aspect ratio of each Jpeg size to be one of well known ones 5584 (e.g. 4:3, 16:9, 3:2 etc.). If the sensor maximum resolution 5585 (defined by android.sensor.info.activeArraySize) has an aspect ratio other than these, 5586 it does not have to be included in the supported JPEG sizes. 5587 * Some hardware JPEG encoders may have pixel boundary alignment requirements, such as 5588 the dimensions being a multiple of 16. 5589 5590 Therefore, the maximum JPEG size may be smaller than sensor maximum resolution. 5591 However, the largest JPEG size must be as close as possible to the sensor maximum 5592 resolution given above constraints. It is required that after aspect ratio adjustments, 5593 additional size reduction due to other issues must be less than 3% in area. For example, 5594 if the sensor maximum resolution is 3280x2464, if the maximum JPEG size has aspect 5595 ratio 4:3, the JPEG encoder alignment requirement is 16, the maximum JPEG size will be 5596 3264x2448. 5597 5598 For FULL capability devices (`android.info.supportedHardwareLevel == FULL`), 5599 the HAL must include all YUV_420_888 sizes that have JPEG sizes listed 5600 here as output streams. 5601 5602 It must also include each below resolution if it is smaller than or 5603 equal to the sensor maximum resolution (for both YUV_420_888 and JPEG 5604 formats), as output streams: 5605 5606 * 240p (320 x 240) 5607 * 480p (640 x 480) 5608 * 720p (1280 x 720) 5609 * 1080p (1920 x 1080) 5610 5611 For LIMITED capability devices 5612 (`android.info.supportedHardwareLevel == LIMITED`), 5613 the HAL only has to list up to the maximum video size 5614 supported by the device. 5615 5616 Regardless of hardware level, every output resolution available for 5617 YUV_420_888 must also be available for IMPLEMENTATION_DEFINED. 5618 5619 This supercedes the following fields, which are now deprecated: 5620 5621 * availableFormats 5622 * available[Processed,Raw,Jpeg]Sizes 5623 </hal_details> 5624 </entry> 5625 <entry name="availableMinFrameDurations" type="int64" visibility="ndk_public" 5626 container="array" typedef="streamConfigurationDuration" hwlevel="legacy"> 5627 <array> 5628 <size>4</size> 5629 <size>n</size> 5630 </array> 5631 <description>This lists the minimum frame duration for each 5632 format/size combination. 5633 </description> 5634 <units>(format, width, height, ns) x n</units> 5635 <details> 5636 This should correspond to the frame duration when only that 5637 stream is active, with all processing (typically in android.*.mode) 5638 set to either OFF or FAST. 5639 5640 When multiple streams are used in a request, the minimum frame 5641 duration will be max(individual stream min durations). 5642 5643 The minimum frame duration of a stream (of a particular format, size) 5644 is the same regardless of whether the stream is input or output. 5645 5646 See android.sensor.frameDuration and 5647 android.scaler.availableStallDurations for more details about 5648 calculating the max frame rate. 5649 5650 (Keep in sync with 5651 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration}) 5652 </details> 5653 <tag id="V1" /> 5654 </entry> 5655 <entry name="availableStallDurations" type="int64" visibility="ndk_public" 5656 container="array" typedef="streamConfigurationDuration" hwlevel="legacy"> 5657 <array> 5658 <size>4</size> 5659 <size>n</size> 5660 </array> 5661 <description>This lists the maximum stall duration for each 5662 output format/size combination. 5663 </description> 5664 <units>(format, width, height, ns) x n</units> 5665 <details> 5666 A stall duration is how much extra time would get added 5667 to the normal minimum frame duration for a repeating request 5668 that has streams with non-zero stall. 5669 5670 For example, consider JPEG captures which have the following 5671 characteristics: 5672 5673 * JPEG streams act like processed YUV streams in requests for which 5674 they are not included; in requests in which they are directly 5675 referenced, they act as JPEG streams. This is because supporting a 5676 JPEG stream requires the underlying YUV data to always be ready for 5677 use by a JPEG encoder, but the encoder will only be used (and impact 5678 frame duration) on requests that actually reference a JPEG stream. 5679 * The JPEG processor can run concurrently to the rest of the camera 5680 pipeline, but cannot process more than 1 capture at a time. 5681 5682 In other words, using a repeating YUV request would result 5683 in a steady frame rate (let's say it's 30 FPS). If a single 5684 JPEG request is submitted periodically, the frame rate will stay 5685 at 30 FPS (as long as we wait for the previous JPEG to return each 5686 time). If we try to submit a repeating YUV + JPEG request, then 5687 the frame rate will drop from 30 FPS. 5688 5689 In general, submitting a new request with a non-0 stall time 5690 stream will _not_ cause a frame rate drop unless there are still 5691 outstanding buffers for that stream from previous requests. 5692 5693 Submitting a repeating request with streams (call this `S`) 5694 is the same as setting the minimum frame duration from 5695 the normal minimum frame duration corresponding to `S`, added with 5696 the maximum stall duration for `S`. 5697 5698 If interleaving requests with and without a stall duration, 5699 a request will stall by the maximum of the remaining times 5700 for each can-stall stream with outstanding buffers. 5701 5702 This means that a stalling request will not have an exposure start 5703 until the stall has completed. 5704 5705 This should correspond to the stall duration when only that stream is 5706 active, with all processing (typically in android.*.mode) set to FAST 5707 or OFF. Setting any of the processing modes to HIGH_QUALITY 5708 effectively results in an indeterminate stall duration for all 5709 streams in a request (the regular stall calculation rules are 5710 ignored). 5711 5712 The following formats may always have a stall duration: 5713 5714 * {@link android.graphics.ImageFormat#JPEG} 5715 * {@link android.graphics.ImageFormat#RAW_SENSOR} 5716 5717 The following formats will never have a stall duration: 5718 5719 * {@link android.graphics.ImageFormat#YUV_420_888} 5720 * {@link android.graphics.ImageFormat#RAW10} 5721 5722 All other formats may or may not have an allowed stall duration on 5723 a per-capability basis; refer to android.request.availableCapabilities 5724 for more details. 5725 5726 See android.sensor.frameDuration for more information about 5727 calculating the max frame rate (absent stalls). 5728 5729 (Keep up to date with 5730 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration} ) 5731 </details> 5732 <hal_details> 5733 If possible, it is recommended that all non-JPEG formats 5734 (such as RAW16) should not have a stall duration. RAW10, RAW12, RAW_OPAQUE 5735 and IMPLEMENTATION_DEFINED must not have stall durations. 5736 </hal_details> 5737 <tag id="V1" /> 5738 </entry> 5739 <entry name="streamConfigurationMap" type="int32" visibility="java_public" 5740 synthetic="true" typedef="streamConfigurationMap" 5741 hwlevel="legacy"> 5742 <description>The available stream configurations that this 5743 camera device supports; also includes the minimum frame durations 5744 and the stall durations for each format/size combination. 5745 </description> 5746 <details> 5747 All camera devices will support sensor maximum resolution (defined by 5748 android.sensor.info.activeArraySize) for the JPEG format. 5749 5750 For a given use case, the actual maximum supported resolution 5751 may be lower than what is listed here, depending on the destination 5752 Surface for the image data. For example, for recording video, 5753 the video encoder chosen may have a maximum size limit (e.g. 1080p) 5754 smaller than what the camera (e.g. maximum resolution is 3264x2448) 5755 can provide. 5756 5757 Please reference the documentation for the image data destination to 5758 check if it limits the maximum size for image data. 5759 5760 The following table describes the minimum required output stream 5761 configurations based on the hardware level 5762 (android.info.supportedHardwareLevel): 5763 5764 Format | Size | Hardware Level | Notes 5765 :-------------------------------------------------:|:--------------------------------------------:|:--------------:|:--------------: 5766 {@link android.graphics.ImageFormat#JPEG} | android.sensor.info.activeArraySize (*1) | Any | 5767 {@link android.graphics.ImageFormat#JPEG} | 1920x1080 (1080p) | Any | if 1080p <= activeArraySize 5768 {@link android.graphics.ImageFormat#JPEG} | 1280x720 (720p) | Any | if 720p <= activeArraySize 5769 {@link android.graphics.ImageFormat#JPEG} | 640x480 (480p) | Any | if 480p <= activeArraySize 5770 {@link android.graphics.ImageFormat#JPEG} | 320x240 (240p) | Any | if 240p <= activeArraySize 5771 {@link android.graphics.ImageFormat#YUV_420_888} | all output sizes available for JPEG | FULL | 5772 {@link android.graphics.ImageFormat#YUV_420_888} | all output sizes available for JPEG, up to the maximum video size | LIMITED | 5773 {@link android.graphics.ImageFormat#PRIVATE} | same as YUV_420_888 | Any | 5774 5775 Refer to android.request.availableCapabilities and {@link 5776 android.hardware.camera2.CameraDevice#createCaptureSession} for additional mandatory 5777 stream configurations on a per-capability basis. 5778 5779 *1: For JPEG format, the sizes may be restricted by below conditions: 5780 5781 * The HAL may choose the aspect ratio of each Jpeg size to be one of well known ones 5782 (e.g. 4:3, 16:9, 3:2 etc.). If the sensor maximum resolution 5783 (defined by android.sensor.info.activeArraySize) has an aspect ratio other than these, 5784 it does not have to be included in the supported JPEG sizes. 5785 * Some hardware JPEG encoders may have pixel boundary alignment requirements, such as 5786 the dimensions being a multiple of 16. 5787 Therefore, the maximum JPEG size may be smaller than sensor maximum resolution. 5788 However, the largest JPEG size will be as close as possible to the sensor maximum 5789 resolution given above constraints. It is required that after aspect ratio adjustments, 5790 additional size reduction due to other issues must be less than 3% in area. For example, 5791 if the sensor maximum resolution is 3280x2464, if the maximum JPEG size has aspect 5792 ratio 4:3, and the JPEG encoder alignment requirement is 16, the maximum JPEG size will be 5793 3264x2448. 5794 </details> 5795 <hal_details> 5796 Do not set this property directly 5797 (it is synthetic and will not be available at the HAL layer); 5798 set the android.scaler.availableStreamConfigurations instead. 5799 5800 Not all output formats may be supported in a configuration with 5801 an input stream of a particular format. For more details, see 5802 android.scaler.availableInputOutputFormatsMap. 5803 5804 It is recommended (but not mandatory) to also include half/quarter 5805 of sensor maximum resolution for JPEG formats (regardless of hardware 5806 level). 5807 5808 (The following is a rewording of the above required table): 5809 5810 The HAL must include sensor maximum resolution (defined by 5811 android.sensor.info.activeArraySize). 5812 5813 For FULL capability devices (`android.info.supportedHardwareLevel == FULL`), 5814 the HAL must include all YUV_420_888 sizes that have JPEG sizes listed 5815 here as output streams. 5816 5817 It must also include each below resolution if it is smaller than or 5818 equal to the sensor maximum resolution (for both YUV_420_888 and JPEG 5819 formats), as output streams: 5820 5821 * 240p (320 x 240) 5822 * 480p (640 x 480) 5823 * 720p (1280 x 720) 5824 * 1080p (1920 x 1080) 5825 5826 For LIMITED capability devices 5827 (`android.info.supportedHardwareLevel == LIMITED`), 5828 the HAL only has to list up to the maximum video size 5829 supported by the device. 5830 5831 Regardless of hardware level, every output resolution available for 5832 YUV_420_888 must also be available for IMPLEMENTATION_DEFINED. 5833 5834 This supercedes the following fields, which are now deprecated: 5835 5836 * availableFormats 5837 * available[Processed,Raw,Jpeg]Sizes 5838 </hal_details> 5839 </entry> 5840 <entry name="croppingType" type="byte" visibility="public" enum="true" 5841 hwlevel="legacy"> 5842 <enum> 5843 <value>CENTER_ONLY 5844 <notes> 5845 The camera device only supports centered crop regions. 5846 </notes> 5847 </value> 5848 <value>FREEFORM 5849 <notes> 5850 The camera device supports arbitrarily chosen crop regions. 5851 </notes> 5852 </value> 5853 </enum> 5854 <description>The crop type that this camera device supports.</description> 5855 <details> 5856 When passing a non-centered crop region (android.scaler.cropRegion) to a camera 5857 device that only supports CENTER_ONLY cropping, the camera device will move the 5858 crop region to the center of the sensor active array (android.sensor.info.activeArraySize) 5859 and keep the crop region width and height unchanged. The camera device will return the 5860 final used crop region in metadata result android.scaler.cropRegion. 5861 5862 Camera devices that support FREEFORM cropping will support any crop region that 5863 is inside of the active array. The camera device will apply the same crop region and 5864 return the final used crop region in capture result metadata android.scaler.cropRegion. 5865 5866 LEGACY capability devices will only support CENTER_ONLY cropping. 5867 </details> 5868 </entry> 5869 </static> 5870 </section> 5871 <section name="sensor"> 5872 <controls> 5873 <entry name="exposureTime" type="int64" visibility="public" hwlevel="full"> 5874 <description>Duration each pixel is exposed to 5875 light.</description> 5876 <units>Nanoseconds</units> 5877 <range>android.sensor.info.exposureTimeRange</range> 5878 <details>If the sensor can't expose this exact duration, it will shorten the 5879 duration exposed to the nearest possible value (rather than expose longer). 5880 The final exposure time used will be available in the output capture result. 5881 5882 This control is only effective if android.control.aeMode or android.control.mode is set to 5883 OFF; otherwise the auto-exposure algorithm will override this value. 5884 </details> 5885 <tag id="V1" /> 5886 </entry> 5887 <entry name="frameDuration" type="int64" visibility="public" hwlevel="full"> 5888 <description>Duration from start of frame exposure to 5889 start of next frame exposure.</description> 5890 <units>Nanoseconds</units> 5891 <range>See android.sensor.info.maxFrameDuration, 5892 android.scaler.streamConfigurationMap. The duration 5893 is capped to `max(duration, exposureTime + overhead)`.</range> 5894 <details> 5895 The maximum frame rate that can be supported by a camera subsystem is 5896 a function of many factors: 5897 5898 * Requested resolutions of output image streams 5899 * Availability of binning / skipping modes on the imager 5900 * The bandwidth of the imager interface 5901 * The bandwidth of the various ISP processing blocks 5902 5903 Since these factors can vary greatly between different ISPs and 5904 sensors, the camera abstraction tries to represent the bandwidth 5905 restrictions with as simple a model as possible. 5906 5907 The model presented has the following characteristics: 5908 5909 * The image sensor is always configured to output the smallest 5910 resolution possible given the application's requested output stream 5911 sizes. The smallest resolution is defined as being at least as large 5912 as the largest requested output stream size; the camera pipeline must 5913 never digitally upsample sensor data when the crop region covers the 5914 whole sensor. In general, this means that if only small output stream 5915 resolutions are configured, the sensor can provide a higher frame 5916 rate. 5917 * Since any request may use any or all the currently configured 5918 output streams, the sensor and ISP must be configured to support 5919 scaling a single capture to all the streams at the same time. This 5920 means the camera pipeline must be ready to produce the largest 5921 requested output size without any delay. Therefore, the overall 5922 frame rate of a given configured stream set is governed only by the 5923 largest requested stream resolution. 5924 * Using more than one output stream in a request does not affect the 5925 frame duration. 5926 * Certain format-streams may need to do additional background processing 5927 before data is consumed/produced by that stream. These processors 5928 can run concurrently to the rest of the camera pipeline, but 5929 cannot process more than 1 capture at a time. 5930 5931 The necessary information for the application, given the model above, 5932 is provided via the android.scaler.streamConfigurationMap field using 5933 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration}. 5934 These are used to determine the maximum frame rate / minimum frame 5935 duration that is possible for a given stream configuration. 5936 5937 Specifically, the application can use the following rules to 5938 determine the minimum frame duration it can request from the camera 5939 device: 5940 5941 1. Let the set of currently configured input/output streams 5942 be called `S`. 5943 1. Find the minimum frame durations for each stream in `S`, by looking 5944 it up in android.scaler.streamConfigurationMap using {@link 5945 android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration} 5946 (with its respective size/format). Let this set of frame durations be 5947 called `F`. 5948 1. For any given request `R`, the minimum frame duration allowed 5949 for `R` is the maximum out of all values in `F`. Let the streams 5950 used in `R` be called `S_r`. 5951 5952 If none of the streams in `S_r` have a stall time (listed in {@link 5953 android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration} 5954 using its respective size/format), then the frame duration in `F` 5955 determines the steady state frame rate that the application will get 5956 if it uses `R` as a repeating request. Let this special kind of 5957 request be called `Rsimple`. 5958 5959 A repeating request `Rsimple` can be _occasionally_ interleaved 5960 by a single capture of a new request `Rstall` (which has at least 5961 one in-use stream with a non-0 stall time) and if `Rstall` has the 5962 same minimum frame duration this will not cause a frame rate loss 5963 if all buffers from the previous `Rstall` have already been 5964 delivered. 5965 5966 For more details about stalling, see 5967 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputStallDuration}. 5968 5969 This control is only effective if android.control.aeMode or android.control.mode is set to 5970 OFF; otherwise the auto-exposure algorithm will override this value. 5971 </details> 5972 <hal_details> 5973 For more details about stalling, see 5974 android.scaler.availableStallDurations. 5975 </hal_details> 5976 <tag id="V1" /> 5977 </entry> 5978 <entry name="sensitivity" type="int32" visibility="public" hwlevel="full"> 5979 <description>The amount of gain applied to sensor data 5980 before processing.</description> 5981 <units>ISO arithmetic units</units> 5982 <range>android.sensor.info.sensitivityRange</range> 5983 <details> 5984 The sensitivity is the standard ISO sensitivity value, 5985 as defined in ISO 12232:2006. 5986 5987 The sensitivity must be within android.sensor.info.sensitivityRange, and 5988 if if it less than android.sensor.maxAnalogSensitivity, the camera device 5989 is guaranteed to use only analog amplification for applying the gain. 5990 5991 If the camera device cannot apply the exact sensitivity 5992 requested, it will reduce the gain to the nearest supported 5993 value. The final sensitivity used will be available in the 5994 output capture result. 5995 5996 This control is only effective if android.control.aeMode or android.control.mode is set to 5997 OFF; otherwise the auto-exposure algorithm will override this value. 5998 </details> 5999 <hal_details>ISO 12232:2006 REI method is acceptable.</hal_details> 6000 <tag id="V1" /> 6001 </entry> 6002 </controls> 6003 <static> 6004 <namespace name="info"> 6005 <entry name="activeArraySize" type="int32" visibility="public" 6006 type_notes="Four ints defining the active pixel rectangle" 6007 container="array" typedef="rectangle" hwlevel="legacy"> 6008 <array> 6009 <size>4</size> 6010 </array> 6011 <description> 6012 The area of the image sensor which corresponds to active pixels after any geometric 6013 distortion correction has been applied. 6014 </description> 6015 <units>Pixel coordinates on the image sensor</units> 6016 <details> 6017 This is the rectangle representing the size of the active region of the sensor (i.e. 6018 the region that actually receives light from the scene) after any geometric correction 6019 has been applied, and should be treated as the maximum size in pixels of any of the 6020 image output formats aside from the raw formats. 6021 6022 This rectangle is defined relative to the full pixel array; (0,0) is the top-left of 6023 the full pixel array, and the size of the full pixel array is given by 6024 android.sensor.info.pixelArraySize. 6025 6026 The coordinate system for most other keys that list pixel coordinates, including 6027 android.scaler.cropRegion, is defined relative to the active array rectangle given in 6028 this field, with `(0, 0)` being the top-left of this rectangle. 6029 6030 The active array may be smaller than the full pixel array, since the full array may 6031 include black calibration pixels or other inactive regions, and geometric correction 6032 resulting in scaling or cropping may have been applied. 6033 </details> 6034 <hal_details> 6035 This array contains `(xmin, ymin, width, height)`. The `(xmin, ymin)` must be 6036 &gt;= `(0,0)`. 6037 The `(width, height)` must be &lt;= `android.sensor.info.pixelArraySize`. 6038 </hal_details> 6039 <tag id="RAW" /> 6040 </entry> 6041 <entry name="sensitivityRange" type="int32" visibility="public" 6042 type_notes="Range of supported sensitivities" 6043 container="array" typedef="rangeInt" 6044 hwlevel="full"> 6045 <array> 6046 <size>2</size> 6047 </array> 6048 <description>Range of sensitivities for android.sensor.sensitivity supported by this 6049 camera device.</description> 6050 <range>Min <= 100, Max &gt;= 800</range> 6051 <details> 6052 The values are the standard ISO sensitivity values, 6053 as defined in ISO 12232:2006. 6054 </details> 6055 6056 <tag id="BC" /> 6057 <tag id="V1" /> 6058 </entry> 6059 <entry name="colorFilterArrangement" type="byte" visibility="public" enum="true" 6060 hwlevel="full"> 6061 <enum> 6062 <value>RGGB</value> 6063 <value>GRBG</value> 6064 <value>GBRG</value> 6065 <value>BGGR</value> 6066 <value>RGB 6067 <notes>Sensor is not Bayer; output has 3 16-bit 6068 values for each pixel, instead of just 1 16-bit value 6069 per pixel.</notes></value> 6070 </enum> 6071 <description>The arrangement of color filters on sensor; 6072 represents the colors in the top-left 2x2 section of 6073 the sensor, in reading order.</description> 6074 <tag id="RAW" /> 6075 </entry> 6076 <entry name="exposureTimeRange" type="int64" visibility="public" 6077 type_notes="nanoseconds" container="array" typedef="rangeLong" 6078 hwlevel="full"> 6079 <array> 6080 <size>2</size> 6081 </array> 6082 <description>The range of image exposure times for android.sensor.exposureTime supported 6083 by this camera device. 6084 </description> 6085 <units>Nanoseconds</units> 6086 <range>The minimum exposure time will be less than 100 us. For FULL 6087 capability devices (android.info.supportedHardwareLevel == FULL), 6088 the maximum exposure time will be greater than 100ms.</range> 6089 <hal_details>For FULL capability devices (android.info.supportedHardwareLevel == FULL), 6090 The maximum of the range SHOULD be at least 1 second (1e9), MUST be at least 6091 100ms. 6092 </hal_details> 6093 <tag id="V1" /> 6094 </entry> 6095 <entry name="maxFrameDuration" type="int64" visibility="public" 6096 hwlevel="full"> 6097 <description>The maximum possible frame duration (minimum frame rate) for 6098 android.sensor.frameDuration that is supported this camera device.</description> 6099 <units>Nanoseconds</units> 6100 <range>For FULL capability devices 6101 (android.info.supportedHardwareLevel == FULL), at least 100ms. 6102 </range> 6103 <details>Attempting to use frame durations beyond the maximum will result in the frame 6104 duration being clipped to the maximum. See that control for a full definition of frame 6105 durations. 6106 6107 Refer to {@link 6108 android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration} 6109 for the minimum frame duration values. 6110 </details> 6111 <hal_details> 6112 For FULL capability devices (android.info.supportedHardwareLevel == FULL), 6113 The maximum of the range SHOULD be at least 6114 1 second (1e9), MUST be at least 100ms (100e6). 6115 6116 android.sensor.info.maxFrameDuration must be greater or 6117 equal to the android.sensor.info.exposureTimeRange max 6118 value (since exposure time overrides frame duration). 6119 6120 Available minimum frame durations for JPEG must be no greater 6121 than that of the YUV_420_888/IMPLEMENTATION_DEFINED 6122 minimum frame durations (for that respective size). 6123 6124 Since JPEG processing is considered offline and can take longer than 6125 a single uncompressed capture, refer to 6126 android.scaler.availableStallDurations 6127 for details about encoding this scenario. 6128 </hal_details> 6129 <tag id="V1" /> 6130 </entry> 6131 <entry name="physicalSize" type="float" visibility="public" 6132 type_notes="width x height" 6133 container="array" typedef="sizeF" hwlevel="legacy"> 6134 <array> 6135 <size>2</size> 6136 </array> 6137 <description>The physical dimensions of the full pixel 6138 array.</description> 6139 <units>Millimeters</units> 6140 <details>This is the physical size of the sensor pixel 6141 array defined by android.sensor.info.pixelArraySize. 6142 </details> 6143 <hal_details>Needed for FOV calculation for old API</hal_details> 6144 <tag id="V1" /> 6145 <tag id="BC" /> 6146 </entry> 6147 <entry name="pixelArraySize" type="int32" visibility="public" 6148 container="array" typedef="size" hwlevel="legacy"> 6149 <array> 6150 <size>2</size> 6151 </array> 6152 <description>Dimensions of the full pixel array, possibly 6153 including black calibration pixels.</description> 6154 <units>Pixels</units> 6155 <details>The pixel count of the full pixel array of the image sensor, which covers 6156 android.sensor.info.physicalSize area. This represents the full pixel dimensions of 6157 the raw buffers produced by this sensor. 6158 6159 If a camera device supports raw sensor formats, either this or 6160 android.sensor.info.preCorrectionActiveArraySize is the maximum dimensions for the raw 6161 output formats listed in android.scaler.streamConfigurationMap (this depends on 6162 whether or not the image sensor returns buffers containing pixels that are not 6163 part of the active array region for blacklevel calibration or other purposes). 6164 6165 Some parts of the full pixel array may not receive light from the scene, 6166 or be otherwise inactive. The android.sensor.info.preCorrectionActiveArraySize key 6167 defines the rectangle of active pixels that will be included in processed image 6168 formats. 6169 </details> 6170 <tag id="RAW" /> 6171 <tag id="BC" /> 6172 </entry> 6173 <entry name="whiteLevel" type="int32" visibility="public"> 6174 <description> 6175 Maximum raw value output by sensor. 6176 </description> 6177 <range>&gt; 255 (8-bit output)</range> 6178 <details> 6179 This specifies the fully-saturated encoding level for the raw 6180 sample values from the sensor. This is typically caused by the 6181 sensor becoming highly non-linear or clipping. The minimum for 6182 each channel is specified by the offset in the 6183 android.sensor.blackLevelPattern key. 6184 6185 The white level is typically determined either by sensor bit depth 6186 (8-14 bits is expected), or by the point where the sensor response 6187 becomes too non-linear to be useful. The default value for this is 6188 maximum representable value for a 16-bit raw sample (2^16 - 1). 6189 6190 The white level values of captured images may vary for different 6191 capture settings (e.g., android.sensor.sensitivity). This key 6192 represents a coarse approximation for such case. It is recommended 6193 to use android.sensor.dynamicWhiteLevel for captures when supported 6194 by the camera device, which provides more accurate white level values. 6195 </details> 6196 <hal_details> 6197 The full bit depth of the sensor must be available in the raw data, 6198 so the value for linear sensors should not be significantly lower 6199 than maximum raw value supported, i.e. 2^(sensor bits per pixel). 6200 </hal_details> 6201 <tag id="RAW" /> 6202 </entry> 6203 <entry name="timestampSource" type="byte" visibility="public" 6204 enum="true" hwlevel="legacy"> 6205 <enum> 6206 <value>UNKNOWN 6207 <notes> 6208 Timestamps from android.sensor.timestamp are in nanoseconds and monotonic, 6209 but can not be compared to timestamps from other subsystems 6210 (e.g. accelerometer, gyro etc.), or other instances of the same or different 6211 camera devices in the same system. Timestamps between streams and results for 6212 a single camera instance are comparable, and the timestamps for all buffers 6213 and the result metadata generated by a single capture are identical. 6214 </notes> 6215 </value> 6216 <value>REALTIME 6217 <notes> 6218 Timestamps from android.sensor.timestamp are in the same timebase as 6219 {@link android.os.SystemClock#elapsedRealtimeNanos}, 6220 and they can be compared to other timestamps using that base. 6221 </notes> 6222 </value> 6223 </enum> 6224 <description>The time base source for sensor capture start timestamps.</description> 6225 <details> 6226 The timestamps provided for captures are always in nanoseconds and monotonic, but 6227 may not based on a time source that can be compared to other system time sources. 6228 6229 This characteristic defines the source for the timestamps, and therefore whether they 6230 can be compared against other system time sources/timestamps. 6231 </details> 6232 <hal_details> 6233 For camera devices implement UNKNOWN, the camera framework expects that the timestamp 6234 source to be SYSTEM_TIME_MONOTONIC. For camera devices implement REALTIME, the camera 6235 framework expects that the timestamp source to be SYSTEM_TIME_BOOTTIME. See 6236 system/core/include/utils/Timers.h for the definition of SYSTEM_TIME_MONOTONIC and 6237 SYSTEM_TIME_BOOTTIME. Note that HAL must follow above expectation; otherwise video 6238 recording might suffer unexpected behavior. 6239 6240 Also, camera devices implements REALTIME must pass the ITS sensor fusion test which 6241 tests the alignment between camera timestamps and gyro sensor timestamps. 6242 </hal_details> 6243 <tag id="V1" /> 6244 </entry> 6245 <entry name="lensShadingApplied" type="byte" visibility="public" enum="true" 6246 typedef="boolean"> 6247 <enum> 6248 <value>FALSE</value> 6249 <value>TRUE</value> 6250 </enum> 6251 <description>Whether the RAW images output from this camera device are subject to 6252 lens shading correction.</description> 6253 <details> 6254 If TRUE, all images produced by the camera device in the RAW image formats will 6255 have lens shading correction already applied to it. If FALSE, the images will 6256 not be adjusted for lens shading correction. 6257 See android.request.maxNumOutputRaw for a list of RAW image formats. 6258 6259 This key will be `null` for all devices do not report this information. 6260 Devices with RAW capability will always report this information in this key. 6261 </details> 6262 </entry> 6263 <entry name="preCorrectionActiveArraySize" type="int32" visibility="public" 6264 type_notes="Four ints defining the active pixel rectangle" container="array" 6265 typedef="rectangle" hwlevel="legacy"> 6266 <array> 6267 <size>4</size> 6268 </array> 6269 <description> 6270 The area of the image sensor which corresponds to active pixels prior to the 6271 application of any geometric distortion correction. 6272 </description> 6273 <units>Pixel coordinates on the image sensor</units> 6274 <details> 6275 This is the rectangle representing the size of the active region of the sensor (i.e. 6276 the region that actually receives light from the scene) before any geometric correction 6277 has been applied, and should be treated as the active region rectangle for any of the 6278 raw formats. All metadata associated with raw processing (e.g. the lens shading 6279 correction map, and radial distortion fields) treats the top, left of this rectangle as 6280 the origin, (0,0). 6281 6282 The size of this region determines the maximum field of view and the maximum number of 6283 pixels that an image from this sensor can contain, prior to the application of 6284 geometric distortion correction. The effective maximum pixel dimensions of a 6285 post-distortion-corrected image is given by the android.sensor.info.activeArraySize 6286 field, and the effective maximum field of view for a post-distortion-corrected image 6287 can be calculated by applying the geometric distortion correction fields to this 6288 rectangle, and cropping to the rectangle given in android.sensor.info.activeArraySize. 6289 6290 E.g. to calculate position of a pixel, (x,y), in a processed YUV output image with the 6291 dimensions in android.sensor.info.activeArraySize given the position of a pixel, 6292 (x', y'), in the raw pixel array with dimensions give in 6293 android.sensor.info.pixelArraySize: 6294 6295 1. Choose a pixel (x', y') within the active array region of the raw buffer given in 6296 android.sensor.info.preCorrectionActiveArraySize, otherwise this pixel is considered 6297 to be outside of the FOV, and will not be shown in the processed output image. 6298 1. Apply geometric distortion correction to get the post-distortion pixel coordinate, 6299 (x_i, y_i). When applying geometric correction metadata, note that metadata for raw 6300 buffers is defined relative to the top, left of the 6301 android.sensor.info.preCorrectionActiveArraySize rectangle. 6302 1. If the resulting corrected pixel coordinate is within the region given in 6303 android.sensor.info.activeArraySize, then the position of this pixel in the 6304 processed output image buffer is `(x_i - activeArray.left, y_i - activeArray.top)`, 6305 when the top, left coordinate of that buffer is treated as (0, 0). 6306 6307 Thus, for pixel x',y' = (25, 25) on a sensor where android.sensor.info.pixelArraySize 6308 is (100,100), android.sensor.info.preCorrectionActiveArraySize is (10, 10, 100, 100), 6309 android.sensor.info.activeArraySize is (20, 20, 80, 80), and the geometric distortion 6310 correction doesn't change the pixel coordinate, the resulting pixel selected in 6311 pixel coordinates would be x,y = (25, 25) relative to the top,left of the raw buffer 6312 with dimensions given in android.sensor.info.pixelArraySize, and would be (5, 5) 6313 relative to the top,left of post-processed YUV output buffer with dimensions given in 6314 android.sensor.info.activeArraySize. 6315 6316 The currently supported fields that correct for geometric distortion are: 6317 6318 1. android.lens.radialDistortion. 6319 6320 If all of the geometric distortion fields are no-ops, this rectangle will be the same 6321 as the post-distortion-corrected rectangle given in 6322 android.sensor.info.activeArraySize. 6323 6324 This rectangle is defined relative to the full pixel array; (0,0) is the top-left of 6325 the full pixel array, and the size of the full pixel array is given by 6326 android.sensor.info.pixelArraySize. 6327 6328 The pre-correction active array may be smaller than the full pixel array, since the 6329 full array may include black calibration pixels or other inactive regions. 6330 </details> 6331 <hal_details> 6332 This array contains `(xmin, ymin, width, height)`. The `(xmin, ymin)` must be 6333 &gt;= `(0,0)`. 6334 The `(width, height)` must be &lt;= `android.sensor.info.pixelArraySize`. 6335 6336 If omitted by the HAL implementation, the camera framework will assume that this is 6337 the same as the post-correction active array region given in 6338 android.sensor.info.activeArraySize. 6339 </hal_details> 6340 <tag id="RAW" /> 6341 </entry> 6342 </namespace> 6343 <entry name="referenceIlluminant1" type="byte" visibility="public" 6344 enum="true"> 6345 <enum> 6346 <value id="1">DAYLIGHT</value> 6347 <value id="2">FLUORESCENT</value> 6348 <value id="3">TUNGSTEN 6349 <notes>Incandescent light</notes> 6350 </value> 6351 <value id="4">FLASH</value> 6352 <value id="9">FINE_WEATHER</value> 6353 <value id="10">CLOUDY_WEATHER</value> 6354 <value id="11">SHADE</value> 6355 <value id="12">DAYLIGHT_FLUORESCENT 6356 <notes>D 5700 - 7100K</notes> 6357 </value> 6358 <value id="13">DAY_WHITE_FLUORESCENT 6359 <notes>N 4600 - 5400K</notes> 6360 </value> 6361 <value id="14">COOL_WHITE_FLUORESCENT 6362 <notes>W 3900 - 4500K</notes> 6363 </value> 6364 <value id="15">WHITE_FLUORESCENT 6365 <notes>WW 3200 - 3700K</notes> 6366 </value> 6367 <value id="17">STANDARD_A</value> 6368 <value id="18">STANDARD_B</value> 6369 <value id="19">STANDARD_C</value> 6370 <value id="20">D55</value> 6371 <value id="21">D65</value> 6372 <value id="22">D75</value> 6373 <value id="23">D50</value> 6374 <value id="24">ISO_STUDIO_TUNGSTEN</value> 6375 </enum> 6376 <description> 6377 The standard reference illuminant used as the scene light source when 6378 calculating the android.sensor.colorTransform1, 6379 android.sensor.calibrationTransform1, and 6380 android.sensor.forwardMatrix1 matrices. 6381 </description> 6382 <details> 6383 The values in this key correspond to the values defined for the 6384 EXIF LightSource tag. These illuminants are standard light sources 6385 that are often used calibrating camera devices. 6386 6387 If this key is present, then android.sensor.colorTransform1, 6388 android.sensor.calibrationTransform1, and 6389 android.sensor.forwardMatrix1 will also be present. 6390 6391 Some devices may choose to provide a second set of calibration 6392 information for improved quality, including 6393 android.sensor.referenceIlluminant2 and its corresponding matrices. 6394 </details> 6395 <hal_details> 6396 The first reference illuminant (android.sensor.referenceIlluminant1) 6397 and corresponding matrices must be present to support the RAW capability 6398 and DNG output. 6399 6400 When producing raw images with a color profile that has only been 6401 calibrated against a single light source, it is valid to omit 6402 android.sensor.referenceIlluminant2 along with the 6403 android.sensor.colorTransform2, android.sensor.calibrationTransform2, 6404 and android.sensor.forwardMatrix2 matrices. 6405 6406 If only android.sensor.referenceIlluminant1 is included, it should be 6407 chosen so that it is representative of typical scene lighting. In 6408 general, D50 or DAYLIGHT will be chosen for this case. 6409 6410 If both android.sensor.referenceIlluminant1 and 6411 android.sensor.referenceIlluminant2 are included, they should be 6412 chosen to represent the typical range of scene lighting conditions. 6413 In general, low color temperature illuminant such as Standard-A will 6414 be chosen for the first reference illuminant and a higher color 6415 temperature illuminant such as D65 will be chosen for the second 6416 reference illuminant. 6417 </hal_details> 6418 <tag id="RAW" /> 6419 </entry> 6420 <entry name="referenceIlluminant2" type="byte" visibility="public"> 6421 <description> 6422 The standard reference illuminant used as the scene light source when 6423 calculating the android.sensor.colorTransform2, 6424 android.sensor.calibrationTransform2, and 6425 android.sensor.forwardMatrix2 matrices. 6426 </description> 6427 <range>Any value listed in android.sensor.referenceIlluminant1</range> 6428 <details> 6429 See android.sensor.referenceIlluminant1 for more details. 6430 6431 If this key is present, then android.sensor.colorTransform2, 6432 android.sensor.calibrationTransform2, and 6433 android.sensor.forwardMatrix2 will also be present. 6434 </details> 6435 <tag id="RAW" /> 6436 </entry> 6437 <entry name="calibrationTransform1" type="rational" 6438 visibility="public" optional="true" 6439 type_notes="3x3 matrix in row-major-order" container="array" 6440 typedef="colorSpaceTransform"> 6441 <array> 6442 <size>3</size> 6443 <size>3</size> 6444 </array> 6445 <description> 6446 A per-device calibration transform matrix that maps from the 6447 reference sensor colorspace to the actual device sensor colorspace. 6448 </description> 6449 <details> 6450 This matrix is used to correct for per-device variations in the 6451 sensor colorspace, and is used for processing raw buffer data. 6452 6453 The matrix is expressed as a 3x3 matrix in row-major-order, and 6454 contains a per-device calibration transform that maps colors 6455 from reference sensor color space (i.e. the "golden module" 6456 colorspace) into this camera device's native sensor color 6457 space under the first reference illuminant 6458 (android.sensor.referenceIlluminant1). 6459 </details> 6460 <tag id="RAW" /> 6461 </entry> 6462 <entry name="calibrationTransform2" type="rational" 6463 visibility="public" optional="true" 6464 type_notes="3x3 matrix in row-major-order" container="array" 6465 typedef="colorSpaceTransform"> 6466 <array> 6467 <size>3</size> 6468 <size>3</size> 6469 </array> 6470 <description> 6471 A per-device calibration transform matrix that maps from the 6472 reference sensor colorspace to the actual device sensor colorspace 6473 (this is the colorspace of the raw buffer data). 6474 </description> 6475 <details> 6476 This matrix is used to correct for per-device variations in the 6477 sensor colorspace, and is used for processing raw buffer data. 6478 6479 The matrix is expressed as a 3x3 matrix in row-major-order, and 6480 contains a per-device calibration transform that maps colors 6481 from reference sensor color space (i.e. the "golden module" 6482 colorspace) into this camera device's native sensor color 6483 space under the second reference illuminant 6484 (android.sensor.referenceIlluminant2). 6485 6486 This matrix will only be present if the second reference 6487 illuminant is present. 6488 </details> 6489 <tag id="RAW" /> 6490 </entry> 6491 <entry name="colorTransform1" type="rational" 6492 visibility="public" optional="true" 6493 type_notes="3x3 matrix in row-major-order" container="array" 6494 typedef="colorSpaceTransform"> 6495 <array> 6496 <size>3</size> 6497 <size>3</size> 6498 </array> 6499 <description> 6500 A matrix that transforms color values from CIE XYZ color space to 6501 reference sensor color space. 6502 </description> 6503 <details> 6504 This matrix is used to convert from the standard CIE XYZ color 6505 space to the reference sensor colorspace, and is used when processing 6506 raw buffer data. 6507 6508 The matrix is expressed as a 3x3 matrix in row-major-order, and 6509 contains a color transform matrix that maps colors from the CIE 6510 XYZ color space to the reference sensor color space (i.e. the 6511 "golden module" colorspace) under the first reference illuminant 6512 (android.sensor.referenceIlluminant1). 6513 6514 The white points chosen in both the reference sensor color space 6515 and the CIE XYZ colorspace when calculating this transform will 6516 match the standard white point for the first reference illuminant 6517 (i.e. no chromatic adaptation will be applied by this transform). 6518 </details> 6519 <tag id="RAW" /> 6520 </entry> 6521 <entry name="colorTransform2" type="rational" 6522 visibility="public" optional="true" 6523 type_notes="3x3 matrix in row-major-order" container="array" 6524 typedef="colorSpaceTransform"> 6525 <array> 6526 <size>3</size> 6527 <size>3</size> 6528 </array> 6529 <description> 6530 A matrix that transforms color values from CIE XYZ color space to 6531 reference sensor color space. 6532 </description> 6533 <details> 6534 This matrix is used to convert from the standard CIE XYZ color 6535 space to the reference sensor colorspace, and is used when processing 6536 raw buffer data. 6537 6538 The matrix is expressed as a 3x3 matrix in row-major-order, and 6539 contains a color transform matrix that maps colors from the CIE 6540 XYZ color space to the reference sensor color space (i.e. the 6541 "golden module" colorspace) under the second reference illuminant 6542 (android.sensor.referenceIlluminant2). 6543 6544 The white points chosen in both the reference sensor color space 6545 and the CIE XYZ colorspace when calculating this transform will 6546 match the standard white point for the second reference illuminant 6547 (i.e. no chromatic adaptation will be applied by this transform). 6548 6549 This matrix will only be present if the second reference 6550 illuminant is present. 6551 </details> 6552 <tag id="RAW" /> 6553 </entry> 6554 <entry name="forwardMatrix1" type="rational" 6555 visibility="public" optional="true" 6556 type_notes="3x3 matrix in row-major-order" container="array" 6557 typedef="colorSpaceTransform"> 6558 <array> 6559 <size>3</size> 6560 <size>3</size> 6561 </array> 6562 <description> 6563 A matrix that transforms white balanced camera colors from the reference 6564 sensor colorspace to the CIE XYZ colorspace with a D50 whitepoint. 6565 </description> 6566 <details> 6567 This matrix is used to convert to the standard CIE XYZ colorspace, and 6568 is used when processing raw buffer data. 6569 6570 This matrix is expressed as a 3x3 matrix in row-major-order, and contains 6571 a color transform matrix that maps white balanced colors from the 6572 reference sensor color space to the CIE XYZ color space with a D50 white 6573 point. 6574 6575 Under the first reference illuminant (android.sensor.referenceIlluminant1) 6576 this matrix is chosen so that the standard white point for this reference 6577 illuminant in the reference sensor colorspace is mapped to D50 in the 6578 CIE XYZ colorspace. 6579 </details> 6580 <tag id="RAW" /> 6581 </entry> 6582 <entry name="forwardMatrix2" type="rational" 6583 visibility="public" optional="true" 6584 type_notes="3x3 matrix in row-major-order" container="array" 6585 typedef="colorSpaceTransform"> 6586 <array> 6587 <size>3</size> 6588 <size>3</size> 6589 </array> 6590 <description> 6591 A matrix that transforms white balanced camera colors from the reference 6592 sensor colorspace to the CIE XYZ colorspace with a D50 whitepoint. 6593 </description> 6594 <details> 6595 This matrix is used to convert to the standard CIE XYZ colorspace, and 6596 is used when processing raw buffer data. 6597 6598 This matrix is expressed as a 3x3 matrix in row-major-order, and contains 6599 a color transform matrix that maps white balanced colors from the 6600 reference sensor color space to the CIE XYZ color space with a D50 white 6601 point. 6602 6603 Under the second reference illuminant (android.sensor.referenceIlluminant2) 6604 this matrix is chosen so that the standard white point for this reference 6605 illuminant in the reference sensor colorspace is mapped to D50 in the 6606 CIE XYZ colorspace. 6607 6608 This matrix will only be present if the second reference 6609 illuminant is present. 6610 </details> 6611 <tag id="RAW" /> 6612 </entry> 6613 <entry name="baseGainFactor" type="rational" 6614 optional="true"> 6615 <description>Gain factor from electrons to raw units when 6616 ISO=100</description> 6617 <tag id="FUTURE" /> 6618 </entry> 6619 <entry name="blackLevelPattern" type="int32" visibility="public" 6620 optional="true" type_notes="2x2 raw count block" container="array" 6621 typedef="blackLevelPattern"> 6622 <array> 6623 <size>4</size> 6624 </array> 6625 <description> 6626 A fixed black level offset for each of the color filter arrangement 6627 (CFA) mosaic channels. 6628 </description> 6629 <range>&gt;= 0 for each.</range> 6630 <details> 6631 This key specifies the zero light value for each of the CFA mosaic 6632 channels in the camera sensor. The maximal value output by the 6633 sensor is represented by the value in android.sensor.info.whiteLevel. 6634 6635 The values are given in the same order as channels listed for the CFA 6636 layout key (see android.sensor.info.colorFilterArrangement), i.e. the 6637 nth value given corresponds to the black level offset for the nth 6638 color channel listed in the CFA. 6639 6640 The black level values of captured images may vary for different 6641 capture settings (e.g., android.sensor.sensitivity). This key 6642 represents a coarse approximation for such case. It is recommended to 6643 use android.sensor.dynamicBlackLevel or use pixels from 6644 android.sensor.opticalBlackRegions directly for captures when 6645 supported by the camera device, which provides more accurate black 6646 level values. For raw capture in particular, it is recommended to use 6647 pixels from android.sensor.opticalBlackRegions to calculate black 6648 level values for each frame. 6649 </details> 6650 <hal_details> 6651 The values are given in row-column scan order, with the first value 6652 corresponding to the element of the CFA in row=0, column=0. 6653 </hal_details> 6654 <tag id="RAW" /> 6655 </entry> 6656 <entry name="maxAnalogSensitivity" type="int32" visibility="public" 6657 optional="true" hwlevel="full"> 6658 <description>Maximum sensitivity that is implemented 6659 purely through analog gain.</description> 6660 <details>For android.sensor.sensitivity values less than or 6661 equal to this, all applied gain must be analog. For 6662 values above this, the gain applied can be a mix of analog and 6663 digital.</details> 6664 <tag id="V1" /> 6665 <tag id="FULL" /> 6666 </entry> 6667 <entry name="orientation" type="int32" visibility="public" 6668 hwlevel="legacy"> 6669 <description>Clockwise angle through which the output image needs to be rotated to be 6670 upright on the device screen in its native orientation. 6671 </description> 6672 <units>Degrees of clockwise rotation; always a multiple of 6673 90</units> 6674 <range>0, 90, 180, 270</range> 6675 <details> 6676 Also defines the direction of rolling shutter readout, which is from top to bottom in 6677 the sensor's coordinate system. 6678 </details> 6679 <tag id="BC" /> 6680 </entry> 6681 <entry name="profileHueSatMapDimensions" type="int32" 6682 visibility="system" optional="true" 6683 type_notes="Number of samples for hue, saturation, and value" 6684 container="array"> 6685 <array> 6686 <size>3</size> 6687 </array> 6688 <description> 6689 The number of input samples for each dimension of 6690 android.sensor.profileHueSatMap. 6691 </description> 6692 <range> 6693 Hue &gt;= 1, 6694 Saturation &gt;= 2, 6695 Value &gt;= 1 6696 </range> 6697 <details> 6698 The number of input samples for the hue, saturation, and value 6699 dimension of android.sensor.profileHueSatMap. The order of the 6700 dimensions given is hue, saturation, value; where hue is the 0th 6701 element. 6702 </details> 6703 <tag id="RAW" /> 6704 </entry> 6705 </static> 6706 <dynamic> 6707 <clone entry="android.sensor.exposureTime" kind="controls"> 6708 </clone> 6709 <clone entry="android.sensor.frameDuration" 6710 kind="controls"></clone> 6711 <clone entry="android.sensor.sensitivity" kind="controls"> 6712 </clone> 6713 <entry name="timestamp" type="int64" visibility="public" 6714 hwlevel="legacy"> 6715 <description>Time at start of exposure of first 6716 row of the image sensor active array, in nanoseconds.</description> 6717 <units>Nanoseconds</units> 6718 <range>&gt; 0</range> 6719 <details>The timestamps are also included in all image 6720 buffers produced for the same capture, and will be identical 6721 on all the outputs. 6722 6723 When android.sensor.info.timestampSource `==` UNKNOWN, 6724 the timestamps measure time since an unspecified starting point, 6725 and are monotonically increasing. They can be compared with the 6726 timestamps for other captures from the same camera device, but are 6727 not guaranteed to be comparable to any other time source. 6728 6729 When android.sensor.info.timestampSource `==` REALTIME, the 6730 timestamps measure time in the same timebase as {@link 6731 android.os.SystemClock#elapsedRealtimeNanos}, and they can 6732 be compared to other timestamps from other subsystems that 6733 are using that base. 6734 6735 For reprocessing, the timestamp will match the start of exposure of 6736 the input image, i.e. {@link CaptureResult#SENSOR_TIMESTAMP the 6737 timestamp} in the TotalCaptureResult that was used to create the 6738 reprocess capture request. 6739 </details> 6740 <hal_details> 6741 All timestamps must be in reference to the kernel's 6742 CLOCK_BOOTTIME monotonic clock, which properly accounts for 6743 time spent asleep. This allows for synchronization with 6744 sensors that continue to operate while the system is 6745 otherwise asleep. 6746 6747 If android.sensor.info.timestampSource `==` REALTIME, 6748 The timestamp must be synchronized with the timestamps from other 6749 sensor subsystems that are using the same timebase. 6750 6751 For reprocessing, the input image's start of exposure can be looked up 6752 with android.sensor.timestamp from the metadata included in the 6753 capture request. 6754 </hal_details> 6755 <tag id="BC" /> 6756 </entry> 6757 <entry name="temperature" type="float" 6758 optional="true"> 6759 <description>The temperature of the sensor, sampled at the time 6760 exposure began for this frame. 6761 6762 The thermal diode being queried should be inside the sensor PCB, or 6763 somewhere close to it. 6764 </description> 6765 6766 <units>Celsius</units> 6767 <range>Optional. This value is missing if no temperature is available.</range> 6768 <tag id="FUTURE" /> 6769 </entry> 6770 <entry name="neutralColorPoint" type="rational" visibility="public" 6771 optional="true" container="array"> 6772 <array> 6773 <size>3</size> 6774 </array> 6775 <description> 6776 The estimated camera neutral color in the native sensor colorspace at 6777 the time of capture. 6778 </description> 6779 <details> 6780 This value gives the neutral color point encoded as an RGB value in the 6781 native sensor color space. The neutral color point indicates the 6782 currently estimated white point of the scene illumination. It can be 6783 used to interpolate between the provided color transforms when 6784 processing raw sensor data. 6785 6786 The order of the values is R, G, B; where R is in the lowest index. 6787 </details> 6788 <tag id="RAW" /> 6789 </entry> 6790 <entry name="noiseProfile" type="double" visibility="public" 6791 optional="true" type_notes="Pairs of noise model coefficients" 6792 container="array" typedef="pairDoubleDouble"> 6793 <array> 6794 <size>2</size> 6795 <size>CFA Channels</size> 6796 </array> 6797 <description> 6798 Noise model coefficients for each CFA mosaic channel. 6799 </description> 6800 <details> 6801 This key contains two noise model coefficients for each CFA channel 6802 corresponding to the sensor amplification (S) and sensor readout 6803 noise (O). These are given as pairs of coefficients for each channel 6804 in the same order as channels listed for the CFA layout key 6805 (see android.sensor.info.colorFilterArrangement). This is 6806 represented as an array of Pair&lt;Double, Double&gt;, where 6807 the first member of the Pair at index n is the S coefficient and the 6808 second member is the O coefficient for the nth color channel in the CFA. 6809 6810 These coefficients are used in a two parameter noise model to describe 6811 the amount of noise present in the image for each CFA channel. The 6812 noise model used here is: 6813 6814 N(x) = sqrt(Sx + O) 6815 6816 Where x represents the recorded signal of a CFA channel normalized to 6817 the range [0, 1], and S and O are the noise model coeffiecients for 6818 that channel. 6819 6820 A more detailed description of the noise model can be found in the 6821 Adobe DNG specification for the NoiseProfile tag. 6822 </details> 6823 <hal_details> 6824 For a CFA layout of RGGB, the list of coefficients would be given as 6825 an array of doubles S0,O0,S1,O1,..., where S0 and O0 are the coefficients 6826 for the red channel, S1 and O1 are the coefficients for the first green 6827 channel, etc. 6828 </hal_details> 6829 <tag id="RAW" /> 6830 </entry> 6831 <entry name="profileHueSatMap" type="float" 6832 visibility="system" optional="true" 6833 type_notes="Mapping for hue, saturation, and value" 6834 container="array"> 6835 <array> 6836 <size>hue_samples</size> 6837 <size>saturation_samples</size> 6838 <size>value_samples</size> 6839 <size>3</size> 6840 </array> 6841 <description> 6842 A mapping containing a hue shift, saturation scale, and value scale 6843 for each pixel. 6844 </description> 6845 <units> 6846 The hue shift is given in degrees; saturation and value scale factors are 6847 unitless and are between 0 and 1 inclusive 6848 </units> 6849 <details> 6850 hue_samples, saturation_samples, and value_samples are given in 6851 android.sensor.profileHueSatMapDimensions. 6852 6853 Each entry of this map contains three floats corresponding to the 6854 hue shift, saturation scale, and value scale, respectively; where the 6855 hue shift has the lowest index. The map entries are stored in the key 6856 in nested loop order, with the value divisions in the outer loop, the 6857 hue divisions in the middle loop, and the saturation divisions in the 6858 inner loop. All zero input saturation entries are required to have a 6859 value scale factor of 1.0. 6860 </details> 6861 <tag id="RAW" /> 6862 </entry> 6863 <entry name="profileToneCurve" type="float" 6864 visibility="system" optional="true" 6865 type_notes="Samples defining a spline for a tone-mapping curve" 6866 container="array"> 6867 <array> 6868 <size>samples</size> 6869 <size>2</size> 6870 </array> 6871 <description> 6872 A list of x,y samples defining a tone-mapping curve for gamma adjustment. 6873 </description> 6874 <range> 6875 Each sample has an input range of `[0, 1]` and an output range of 6876 `[0, 1]`. The first sample is required to be `(0, 0)`, and the last 6877 sample is required to be `(1, 1)`. 6878 </range> 6879 <details> 6880 This key contains a default tone curve that can be applied while 6881 processing the image as a starting point for user adjustments. 6882 The curve is specified as a list of value pairs in linear gamma. 6883 The curve is interpolated using a cubic spline. 6884 </details> 6885 <tag id="RAW" /> 6886 </entry> 6887 <entry name="greenSplit" type="float" visibility="public" optional="true"> 6888 <description> 6889 The worst-case divergence between Bayer green channels. 6890 </description> 6891 <range> 6892 &gt;= 0 6893 </range> 6894 <details> 6895 This value is an estimate of the worst case split between the 6896 Bayer green channels in the red and blue rows in the sensor color 6897 filter array. 6898 6899 The green split is calculated as follows: 6900 6901 1. A 5x5 pixel (or larger) window W within the active sensor array is 6902 chosen. The term 'pixel' here is taken to mean a group of 4 Bayer 6903 mosaic channels (R, Gr, Gb, B). The location and size of the window 6904 chosen is implementation defined, and should be chosen to provide a 6905 green split estimate that is both representative of the entire image 6906 for this camera sensor, and can be calculated quickly. 6907 1. The arithmetic mean of the green channels from the red 6908 rows (mean_Gr) within W is computed. 6909 1. The arithmetic mean of the green channels from the blue 6910 rows (mean_Gb) within W is computed. 6911 1. The maximum ratio R of the two means is computed as follows: 6912 `R = max((mean_Gr + 1)/(mean_Gb + 1), (mean_Gb + 1)/(mean_Gr + 1))` 6913 6914 The ratio R is the green split divergence reported for this property, 6915 which represents how much the green channels differ in the mosaic 6916 pattern. This value is typically used to determine the treatment of 6917 the green mosaic channels when demosaicing. 6918 6919 The green split value can be roughly interpreted as follows: 6920 6921 * R &lt; 1.03 is a negligible split (&lt;3% divergence). 6922 * 1.20 &lt;= R &gt;= 1.03 will require some software 6923 correction to avoid demosaic errors (3-20% divergence). 6924 * R &gt; 1.20 will require strong software correction to produce 6925 a usuable image (&gt;20% divergence). 6926 </details> 6927 <hal_details> 6928 The green split given may be a static value based on prior 6929 characterization of the camera sensor using the green split 6930 calculation method given here over a large, representative, sample 6931 set of images. Other methods of calculation that produce equivalent 6932 results, and can be interpreted in the same manner, may be used. 6933 </hal_details> 6934 <tag id="RAW" /> 6935 </entry> 6936 </dynamic> 6937 <controls> 6938 <entry name="testPatternData" type="int32" visibility="public" optional="true" container="array"> 6939 <array> 6940 <size>4</size> 6941 </array> 6942 <description> 6943 A pixel `[R, G_even, G_odd, B]` that supplies the test pattern 6944 when android.sensor.testPatternMode is SOLID_COLOR. 6945 </description> 6946 <details> 6947 Each color channel is treated as an unsigned 32-bit integer. 6948 The camera device then uses the most significant X bits 6949 that correspond to how many bits are in its Bayer raw sensor 6950 output. 6951 6952 For example, a sensor with RAW10 Bayer output would use the 6953 10 most significant bits from each color channel. 6954 </details> 6955 <hal_details> 6956 </hal_details> 6957 </entry> 6958 <entry name="testPatternMode" type="int32" visibility="public" optional="true" 6959 enum="true"> 6960 <enum> 6961 <value>OFF 6962 <notes>No test pattern mode is used, and the camera 6963 device returns captures from the image sensor. 6964 6965 This is the default if the key is not set.</notes> 6966 </value> 6967 <value>SOLID_COLOR 6968 <notes> 6969 Each pixel in `[R, G_even, G_odd, B]` is replaced by its 6970 respective color channel provided in 6971 android.sensor.testPatternData. 6972 6973 For example: 6974 6975 android.testPatternData = [0, 0xFFFFFFFF, 0xFFFFFFFF, 0] 6976 6977 All green pixels are 100% green. All red/blue pixels are black. 6978 6979 android.testPatternData = [0xFFFFFFFF, 0, 0xFFFFFFFF, 0] 6980 6981 All red pixels are 100% red. Only the odd green pixels 6982 are 100% green. All blue pixels are 100% black. 6983 </notes> 6984 </value> 6985 <value>COLOR_BARS 6986 <notes> 6987 All pixel data is replaced with an 8-bar color pattern. 6988 6989 The vertical bars (left-to-right) are as follows: 6990 6991 * 100% white 6992 * yellow 6993 * cyan 6994 * green 6995 * magenta 6996 * red 6997 * blue 6998 * black 6999 7000 In general the image would look like the following: 7001 7002 W Y C G M R B K 7003 W Y C G M R B K 7004 W Y C G M R B K 7005 W Y C G M R B K 7006 W Y C G M R B K 7007 . . . . . . . . 7008 . . . . . . . . 7009 . . . . . . . . 7010 7011 (B = Blue, K = Black) 7012 7013 Each bar should take up 1/8 of the sensor pixel array width. 7014 When this is not possible, the bar size should be rounded 7015 down to the nearest integer and the pattern can repeat 7016 on the right side. 7017 7018 Each bar's height must always take up the full sensor 7019 pixel array height. 7020 7021 Each pixel in this test pattern must be set to either 7022 0% intensity or 100% intensity. 7023 </notes> 7024 </value> 7025 <value>COLOR_BARS_FADE_TO_GRAY 7026 <notes> 7027 The test pattern is similar to COLOR_BARS, except that 7028 each bar should start at its specified color at the top, 7029 and fade to gray at the bottom. 7030 7031 Furthermore each bar is further subdivided into a left and 7032 right half. The left half should have a smooth gradient, 7033 and the right half should have a quantized gradient. 7034 7035 In particular, the right half's should consist of blocks of the 7036 same color for 1/16th active sensor pixel array width. 7037 7038 The least significant bits in the quantized gradient should 7039 be copied from the most significant bits of the smooth gradient. 7040 7041 The height of each bar should always be a multiple of 128. 7042 When this is not the case, the pattern should repeat at the bottom 7043 of the image. 7044 </notes> 7045 </value> 7046 <value>PN9 7047 <notes> 7048 All pixel data is replaced by a pseudo-random sequence 7049 generated from a PN9 512-bit sequence (typically implemented 7050 in hardware with a linear feedback shift register). 7051 7052 The generator should be reset at the beginning of each frame, 7053 and thus each subsequent raw frame with this test pattern should 7054 be exactly the same as the last. 7055 </notes> 7056 </value> 7057 <value id="256">CUSTOM1 7058 <notes>The first custom test pattern. All custom patterns that are 7059 available only on this camera device are at least this numeric 7060 value. 7061 7062 All of the custom test patterns will be static 7063 (that is the raw image must not vary from frame to frame). 7064 </notes> 7065 </value> 7066 </enum> 7067 <description>When enabled, the sensor sends a test pattern instead of 7068 doing a real exposure from the camera. 7069 </description> 7070 <range>android.sensor.availableTestPatternModes</range> 7071 <details> 7072 When a test pattern is enabled, all manual sensor controls specified 7073 by android.sensor.* will be ignored. All other controls should 7074 work as normal. 7075 7076 For example, if manual flash is enabled, flash firing should still 7077 occur (and that the test pattern remain unmodified, since the flash 7078 would not actually affect it). 7079 7080 Defaults to OFF. 7081 </details> 7082 <hal_details> 7083 All test patterns are specified in the Bayer domain. 7084 7085 The HAL may choose to substitute test patterns from the sensor 7086 with test patterns from on-device memory. In that case, it should be 7087 indistinguishable to the ISP whether the data came from the 7088 sensor interconnect bus (such as CSI2) or memory. 7089 </hal_details> 7090 </entry> 7091 </controls> 7092 <dynamic> 7093 <clone entry="android.sensor.testPatternData" kind="controls"> 7094 </clone> 7095 <clone entry="android.sensor.testPatternMode" kind="controls"> 7096 </clone> 7097 </dynamic> 7098 <static> 7099 <entry name="availableTestPatternModes" type="int32" visibility="public" optional="true" 7100 type_notes="list of enums" container="array"> 7101 <array> 7102 <size>n</size> 7103 </array> 7104 <description>List of sensor test pattern modes for android.sensor.testPatternMode 7105 supported by this camera device. 7106 </description> 7107 <range>Any value listed in android.sensor.testPatternMode</range> 7108 <details> 7109 Defaults to OFF, and always includes OFF if defined. 7110 </details> 7111 <hal_details> 7112 All custom modes must be >= CUSTOM1. 7113 </hal_details> 7114 </entry> 7115 </static> 7116 <dynamic> 7117 <entry name="rollingShutterSkew" type="int64" visibility="public" hwlevel="limited"> 7118 <description>Duration between the start of first row exposure 7119 and the start of last row exposure.</description> 7120 <units>Nanoseconds</units> 7121 <range> &gt;= 0 and &lt; 7122 {@link android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration}.</range> 7123 <details> 7124 This is the exposure time skew between the first and last 7125 row exposure start times. The first row and the last row are 7126 the first and last rows inside of the 7127 android.sensor.info.activeArraySize. 7128 7129 For typical camera sensors that use rolling shutters, this is also equivalent 7130 to the frame readout time. 7131 </details> 7132 <hal_details> 7133 The HAL must report `0` if the sensor is using global shutter, where all pixels begin 7134 exposure at the same time. 7135 </hal_details> 7136 <tag id="V1" /> 7137 </entry> 7138 </dynamic> 7139 <static> 7140 <entry name="opticalBlackRegions" type="int32" visibility="public" optional="true" 7141 container="array" typedef="rectangle"> 7142 <array> 7143 <size>4</size> 7144 <size>num_regions</size> 7145 </array> 7146 <description>List of disjoint rectangles indicating the sensor 7147 optically shielded black pixel regions. 7148 </description> 7149 <details> 7150 In most camera sensors, the active array is surrounded by some 7151 optically shielded pixel areas. By blocking light, these pixels 7152 provides a reliable black reference for black level compensation 7153 in active array region. 7154 7155 This key provides a list of disjoint rectangles specifying the 7156 regions of optically shielded (with metal shield) black pixel 7157 regions if the camera device is capable of reading out these black 7158 pixels in the output raw images. In comparison to the fixed black 7159 level values reported by android.sensor.blackLevelPattern, this key 7160 may provide a more accurate way for the application to calculate 7161 black level of each captured raw images. 7162 7163 When this key is reported, the android.sensor.dynamicBlackLevel and 7164 android.sensor.dynamicWhiteLevel will also be reported. 7165 </details> 7166 <hal_details> 7167 This array contains (xmin, ymin, width, height). The (xmin, ymin) 7168 must be &gt;= (0,0) and &lt;= 7169 android.sensor.info.pixelArraySize. The (width, height) must be 7170 &lt;= android.sensor.info.pixelArraySize. Each region must be 7171 outside the region reported by 7172 android.sensor.info.preCorrectionActiveArraySize. 7173 7174 The HAL must report minimal number of disjoint regions for the 7175 optically shielded back pixel regions. For example, if a region can 7176 be covered by one rectangle, the HAL must not split this region into 7177 multiple rectangles. 7178 </hal_details> 7179 </entry> 7180 </static> 7181 <dynamic> 7182 <entry name="dynamicBlackLevel" type="float" visibility="public" 7183 optional="true" type_notes="2x2 raw count block" container="array"> 7184 <array> 7185 <size>4</size> 7186 </array> 7187 <description> 7188 A per-frame dynamic black level offset for each of the color filter 7189 arrangement (CFA) mosaic channels. 7190 </description> 7191 <range>&gt;= 0 for each.</range> 7192 <details> 7193 Camera sensor black levels may vary dramatically for different 7194 capture settings (e.g. android.sensor.sensitivity). The fixed black 7195 level reported by android.sensor.blackLevelPattern may be too 7196 inaccurate to represent the actual value on a per-frame basis. The 7197 camera device internal pipeline relies on reliable black level values 7198 to process the raw images appropriately. To get the best image 7199 quality, the camera device may choose to estimate the per frame black 7200 level values either based on optically shielded black regions 7201 (android.sensor.opticalBlackRegions) or its internal model. 7202 7203 This key reports the camera device estimated per-frame zero light 7204 value for each of the CFA mosaic channels in the camera sensor. The 7205 android.sensor.blackLevelPattern may only represent a coarse 7206 approximation of the actual black level values. This value is the 7207 black level used in camera device internal image processing pipeline 7208 and generally more accurate than the fixed black level values. 7209 However, since they are estimated values by the camera device, they 7210 may not be as accurate as the black level values calculated from the 7211 optical black pixels reported by android.sensor.opticalBlackRegions. 7212 7213 The values are given in the same order as channels listed for the CFA 7214 layout key (see android.sensor.info.colorFilterArrangement), i.e. the 7215 nth value given corresponds to the black level offset for the nth 7216 color channel listed in the CFA. 7217 7218 This key will be available if android.sensor.opticalBlackRegions is 7219 available or the camera device advertises this key via 7220 {@link android.hardware.camera2.CameraCharacteristics#getAvailableCaptureResultKeys}. 7221 </details> 7222 <hal_details> 7223 The values are given in row-column scan order, with the first value 7224 corresponding to the element of the CFA in row=0, column=0. 7225 </hal_details> 7226 <tag id="RAW" /> 7227 </entry> 7228 <entry name="dynamicWhiteLevel" type="int32" visibility="public" 7229 optional="true" > 7230 <description> 7231 Maximum raw value output by sensor for this frame. 7232 </description> 7233 <range> &gt;= 0</range> 7234 <details> 7235 Since the android.sensor.blackLevelPattern may change for different 7236 capture settings (e.g., android.sensor.sensitivity), the white 7237 level will change accordingly. This key is similar to 7238 android.sensor.info.whiteLevel, but specifies the camera device 7239 estimated white level for each frame. 7240 7241 This key will be available if android.sensor.opticalBlackRegions is 7242 available or the camera device advertises this key via 7243 {@link android.hardware.camera2.CameraCharacteristics#getAvailableCaptureRequestKeys}. 7244 </details> 7245 <hal_details> 7246 The full bit depth of the sensor must be available in the raw data, 7247 so the value for linear sensors should not be significantly lower 7248 than maximum raw value supported, i.e. 2^(sensor bits per pixel). 7249 </hal_details> 7250 <tag id="RAW" /> 7251 </entry> 7252 </dynamic> 7253 <static> 7254 <entry name="opaqueRawSize" type="int32" visibility="system" container="array"> 7255 <array> 7256 <size>n</size> 7257 <size>3</size> 7258 </array> 7259 <description>Size in bytes for all the listed opaque RAW buffer sizes</description> 7260 <range>Must be large enough to fit the opaque RAW of corresponding size produced by 7261 the camera</range> 7262 <details> 7263 This configurations are listed as `(width, height, size_in_bytes)` tuples. 7264 This is used for sizing the gralloc buffers for opaque RAW buffers. 7265 All RAW_OPAQUE output stream configuration listed in 7266 android.scaler.availableStreamConfigurations will have a corresponding tuple in 7267 this key. 7268 </details> 7269 <hal_details> 7270 This key is added in HAL3.4. 7271 For HAL3.4 or above: devices advertising RAW_OPAQUE format output must list this key. 7272 For HAL3.3 or earlier devices: if RAW_OPAQUE ouput is advertised, camera framework 7273 will derive this key by assuming each pixel takes two bytes and no padding bytes 7274 between rows. 7275 </hal_details> 7276 </entry> 7277 </static> 7278 </section> 7279 <section name="shading"> 7280 <controls> 7281 <entry name="mode" type="byte" visibility="public" enum="true" hwlevel="full"> 7282 <enum> 7283 <value>OFF 7284 <notes>No lens shading correction is applied.</notes></value> 7285 <value>FAST 7286 <notes>Apply lens shading corrections, without slowing 7287 frame rate relative to sensor raw output</notes></value> 7288 <value>HIGH_QUALITY 7289 <notes>Apply high-quality lens shading correction, at the 7290 cost of possibly reduced frame rate.</notes></value> 7291 </enum> 7292 <description>Quality of lens shading correction applied 7293 to the image data.</description> 7294 <range>android.shading.availableModes</range> 7295 <details> 7296 When set to OFF mode, no lens shading correction will be applied by the 7297 camera device, and an identity lens shading map data will be provided 7298 if `android.statistics.lensShadingMapMode == ON`. For example, for lens 7299 shading map with size of `[ 4, 3 ]`, 7300 the output android.statistics.lensShadingCorrectionMap for this case will be an identity 7301 map shown below: 7302 7303 [ 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7304 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7305 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7306 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7307 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 7308 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0 ] 7309 7310 When set to other modes, lens shading correction will be applied by the camera 7311 device. Applications can request lens shading map data by setting 7312 android.statistics.lensShadingMapMode to ON, and then the camera device will provide lens 7313 shading map data in android.statistics.lensShadingCorrectionMap; the returned shading map 7314 data will be the one applied by the camera device for this capture request. 7315 7316 The shading map data may depend on the auto-exposure (AE) and AWB statistics, therefore 7317 the reliability of the map data may be affected by the AE and AWB algorithms. When AE and 7318 AWB are in AUTO modes(android.control.aeMode `!=` OFF and android.control.awbMode `!=` 7319 OFF), to get best results, it is recommended that the applications wait for the AE and AWB 7320 to be converged before using the returned shading map data. 7321 </details> 7322 </entry> 7323 <entry name="strength" type="byte"> 7324 <description>Control the amount of shading correction 7325 applied to the images</description> 7326 <units>unitless: 1-10; 10 is full shading 7327 compensation</units> 7328 <tag id="FUTURE" /> 7329 </entry> 7330 </controls> 7331 <dynamic> 7332 <clone entry="android.shading.mode" kind="controls"> 7333 </clone> 7334 </dynamic> 7335 <static> 7336 <entry name="availableModes" type="byte" visibility="public" 7337 type_notes="List of enums (android.shading.mode)." container="array" 7338 typedef="enumList" hwlevel="legacy"> 7339 <array> 7340 <size>n</size> 7341 </array> 7342 <description> 7343 List of lens shading modes for android.shading.mode that are supported by this camera device. 7344 </description> 7345 <range>Any value listed in android.shading.mode</range> 7346 <details> 7347 This list contains lens shading modes that can be set for the camera device. 7348 Camera devices that support the MANUAL_POST_PROCESSING capability will always 7349 list OFF and FAST mode. This includes all FULL level devices. 7350 LEGACY devices will always only support FAST mode. 7351 </details> 7352 <hal_details> 7353 HAL must support both FAST and HIGH_QUALITY if lens shading correction control is 7354 available on the camera device, but the underlying implementation can be the same for 7355 both modes. That is, if the highest quality implementation on the camera device does not 7356 slow down capture rate, then FAST and HIGH_QUALITY will generate the same output. 7357 </hal_details> 7358 </entry> 7359 </static> 7360 </section> 7361 <section name="statistics"> 7362 <controls> 7363 <entry name="faceDetectMode" type="byte" visibility="public" enum="true" 7364 hwlevel="legacy"> 7365 <enum> 7366 <value>OFF 7367 <notes>Do not include face detection statistics in capture 7368 results.</notes></value> 7369 <value optional="true">SIMPLE 7370 <notes>Return face rectangle and confidence values only. 7371 </notes></value> 7372 <value optional="true">FULL 7373 <notes>Return all face 7374 metadata. 7375 7376 In this mode, face rectangles, scores, landmarks, and face IDs are all valid. 7377 </notes></value> 7378 </enum> 7379 <description>Operating mode for the face detector 7380 unit.</description> 7381 <range>android.statistics.info.availableFaceDetectModes</range> 7382 <details>Whether face detection is enabled, and whether it 7383 should output just the basic fields or the full set of 7384 fields.</details> 7385 <hal_details> 7386 SIMPLE mode must fill in android.statistics.faceRectangles and 7387 android.statistics.faceScores. 7388 FULL mode must also fill in android.statistics.faceIds, and 7389 android.statistics.faceLandmarks. 7390 </hal_details> 7391 <tag id="BC" /> 7392 </entry> 7393 <entry name="histogramMode" type="byte" enum="true" typedef="boolean"> 7394 <enum> 7395 <value>OFF</value> 7396 <value>ON</value> 7397 </enum> 7398 <description>Operating mode for histogram 7399 generation</description> 7400 <tag id="FUTURE" /> 7401 </entry> 7402 <entry name="sharpnessMapMode" type="byte" enum="true" typedef="boolean"> 7403 <enum> 7404 <value>OFF</value> 7405 <value>ON</value> 7406 </enum> 7407 <description>Operating mode for sharpness map 7408 generation</description> 7409 <tag id="FUTURE" /> 7410 </entry> 7411 <entry name="hotPixelMapMode" type="byte" visibility="public" enum="true" 7412 typedef="boolean"> 7413 <enum> 7414 <value>OFF 7415 <notes>Hot pixel map production is disabled. 7416 </notes></value> 7417 <value>ON 7418 <notes>Hot pixel map production is enabled. 7419 </notes></value> 7420 </enum> 7421 <description> 7422 Operating mode for hot pixel map generation. 7423 </description> 7424 <range>android.statistics.info.availableHotPixelMapModes</range> 7425 <details> 7426 If set to `true`, a hot pixel map is returned in android.statistics.hotPixelMap. 7427 If set to `false`, no hot pixel map will be returned. 7428 </details> 7429 <tag id="V1" /> 7430 <tag id="RAW" /> 7431 </entry> 7432 </controls> 7433 <static> 7434 <namespace name="info"> 7435 <entry name="availableFaceDetectModes" type="byte" 7436 visibility="public" 7437 type_notes="List of enums from android.statistics.faceDetectMode" 7438 container="array" 7439 typedef="enumList" 7440 hwlevel="legacy"> 7441 <array> 7442 <size>n</size> 7443 </array> 7444 <description>List of face detection modes for android.statistics.faceDetectMode that are 7445 supported by this camera device. 7446 </description> 7447 <range>Any value listed in android.statistics.faceDetectMode</range> 7448 <details>OFF is always supported. 7449 </details> 7450 </entry> 7451 <entry name="histogramBucketCount" type="int32"> 7452 <description>Number of histogram buckets 7453 supported</description> 7454 <range>&gt;= 64</range> 7455 <tag id="FUTURE" /> 7456 </entry> 7457 <entry name="maxFaceCount" type="int32" visibility="public" hwlevel="legacy"> 7458 <description>The maximum number of simultaneously detectable 7459 faces.</description> 7460 <range>0 for cameras without available face detection; otherwise: 7461 `>=4` for LIMITED or FULL hwlevel devices or 7462 `>0` for LEGACY devices.</range> 7463 <tag id="BC" /> 7464 </entry> 7465 <entry name="maxHistogramCount" type="int32"> 7466 <description>Maximum value possible for a histogram 7467 bucket</description> 7468 <tag id="FUTURE" /> 7469 </entry> 7470 <entry name="maxSharpnessMapValue" type="int32"> 7471 <description>Maximum value possible for a sharpness map 7472 region.</description> 7473 <tag id="FUTURE" /> 7474 </entry> 7475 <entry name="sharpnessMapSize" type="int32" 7476 type_notes="width x height" container="array" typedef="size"> 7477 <array> 7478 <size>2</size> 7479 </array> 7480 <description>Dimensions of the sharpness 7481 map</description> 7482 <range>Must be at least 32 x 32</range> 7483 <tag id="FUTURE" /> 7484 </entry> 7485 <entry name="availableHotPixelMapModes" type="byte" visibility="public" 7486 type_notes="list of enums" container="array" typedef="boolean"> 7487 <array> 7488 <size>n</size> 7489 </array> 7490 <description> 7491 List of hot pixel map output modes for android.statistics.hotPixelMapMode that are 7492 supported by this camera device. 7493 </description> 7494 <range>Any value listed in android.statistics.hotPixelMapMode</range> 7495 <details> 7496 If no hotpixel map output is available for this camera device, this will contain only 7497 `false`. 7498 7499 ON is always supported on devices with the RAW capability. 7500 </details> 7501 <tag id="V1" /> 7502 <tag id="RAW" /> 7503 </entry> 7504 <entry name="availableLensShadingMapModes" type="byte" visibility="public" 7505 type_notes="list of enums" container="array" typedef="enumList"> 7506 <array> 7507 <size>n</size> 7508 </array> 7509 <description> 7510 List of lens shading map output modes for android.statistics.lensShadingMapMode that 7511 are supported by this camera device. 7512 </description> 7513 <range>Any value listed in android.statistics.lensShadingMapMode</range> 7514 <details> 7515 If no lens shading map output is available for this camera device, this key will 7516 contain only OFF. 7517 7518 ON is always supported on devices with the RAW capability. 7519 LEGACY mode devices will always only support OFF. 7520 </details> 7521 </entry> 7522 </namespace> 7523 </static> 7524 <dynamic> 7525 <clone entry="android.statistics.faceDetectMode" 7526 kind="controls"></clone> 7527 <entry name="faceIds" type="int32" visibility="ndk_public" 7528 container="array" hwlevel="legacy"> 7529 <array> 7530 <size>n</size> 7531 </array> 7532 <description>List of unique IDs for detected faces.</description> 7533 <details> 7534 Each detected face is given a unique ID that is valid for as long as the face is visible 7535 to the camera device. A face that leaves the field of view and later returns may be 7536 assigned a new ID. 7537 7538 Only available if android.statistics.faceDetectMode == FULL</details> 7539 <tag id="BC" /> 7540 </entry> 7541 <entry name="faceLandmarks" type="int32" visibility="ndk_public" 7542 type_notes="(leftEyeX, leftEyeY, rightEyeX, rightEyeY, mouthX, mouthY)" 7543 container="array" hwlevel="legacy"> 7544 <array> 7545 <size>n</size> 7546 <size>6</size> 7547 </array> 7548 <description>List of landmarks for detected 7549 faces.</description> 7550 <details> 7551 The coordinate system is that of android.sensor.info.activeArraySize, with 7552 `(0, 0)` being the top-left pixel of the active array. 7553 7554 Only available if android.statistics.faceDetectMode == FULL</details> 7555 <tag id="BC" /> 7556 </entry> 7557 <entry name="faceRectangles" type="int32" visibility="ndk_public" 7558 type_notes="(xmin, ymin, xmax, ymax). (0,0) is top-left of active pixel area" 7559 container="array" typedef="rectangle" hwlevel="legacy"> 7560 <array> 7561 <size>n</size> 7562 <size>4</size> 7563 </array> 7564 <description>List of the bounding rectangles for detected 7565 faces.</description> 7566 <details> 7567 The coordinate system is that of android.sensor.info.activeArraySize, with 7568 `(0, 0)` being the top-left pixel of the active array. 7569 7570 Only available if android.statistics.faceDetectMode != OFF</details> 7571 <tag id="BC" /> 7572 </entry> 7573 <entry name="faceScores" type="byte" visibility="ndk_public" 7574 container="array" hwlevel="legacy"> 7575 <array> 7576 <size>n</size> 7577 </array> 7578 <description>List of the face confidence scores for 7579 detected faces</description> 7580 <range>1-100</range> 7581 <details>Only available if android.statistics.faceDetectMode != OFF. 7582 </details> 7583 <hal_details> 7584 The value should be meaningful (for example, setting 100 at 7585 all times is illegal).</hal_details> 7586 <tag id="BC" /> 7587 </entry> 7588 <entry name="faces" type="int32" visibility="java_public" synthetic="true" 7589 container="array" typedef="face" hwlevel="legacy"> 7590 <array> 7591 <size>n</size> 7592 </array> 7593 <description>List of the faces detected through camera face detection 7594 in this capture.</description> 7595 <details> 7596 Only available if android.statistics.faceDetectMode `!=` OFF. 7597 </details> 7598 </entry> 7599 <entry name="histogram" type="int32" 7600 type_notes="count of pixels for each color channel that fall into each histogram bucket, scaled to be between 0 and maxHistogramCount" 7601 container="array"> 7602 <array> 7603 <size>n</size> 7604 <size>3</size> 7605 </array> 7606 <description>A 3-channel histogram based on the raw 7607 sensor data</description> 7608 <details>The k'th bucket (0-based) covers the input range 7609 (with w = android.sensor.info.whiteLevel) of [ k * w/N, 7610 (k + 1) * w / N ). If only a monochrome sharpness map is 7611 supported, all channels should have the same data</details> 7612 <tag id="FUTURE" /> 7613 </entry> 7614 <clone entry="android.statistics.histogramMode" 7615 kind="controls"></clone> 7616 <entry name="sharpnessMap" type="int32" 7617 type_notes="estimated sharpness for each region of the input image. Normalized to be between 0 and maxSharpnessMapValue. Higher values mean sharper (better focused)" 7618 container="array"> 7619 <array> 7620 <size>n</size> 7621 <size>m</size> 7622 <size>3</size> 7623 </array> 7624 <description>A 3-channel sharpness map, based on the raw 7625 sensor data</description> 7626 <details>If only a monochrome sharpness map is supported, 7627 all channels should have the same data</details> 7628 <tag id="FUTURE" /> 7629 </entry> 7630 <clone entry="android.statistics.sharpnessMapMode" 7631 kind="controls"></clone> 7632 <entry name="lensShadingCorrectionMap" type="byte" visibility="java_public" 7633 typedef="lensShadingMap" hwlevel="full"> 7634 <description>The shading map is a low-resolution floating-point map 7635 that lists the coefficients used to correct for vignetting, for each 7636 Bayer color channel.</description> 7637 <range>Each gain factor is &gt;= 1</range> 7638 <details> 7639 The map provided here is the same map that is used by the camera device to 7640 correct both color shading and vignetting for output non-RAW images. 7641 7642 When there is no lens shading correction applied to RAW 7643 output images (android.sensor.info.lensShadingApplied `==` 7644 false), this map is the complete lens shading correction 7645 map; when there is some lens shading correction applied to 7646 the RAW output image (android.sensor.info.lensShadingApplied 7647 `==` true), this map reports the remaining lens shading 7648 correction map that needs to be applied to get shading 7649 corrected images that match the camera device's output for 7650 non-RAW formats. 7651 7652 For a complete shading correction map, the least shaded 7653 section of the image will have a gain factor of 1; all 7654 other sections will have gains above 1. 7655 7656 When android.colorCorrection.mode = TRANSFORM_MATRIX, the map 7657 will take into account the colorCorrection settings. 7658 7659 The shading map is for the entire active pixel array, and is not 7660 affected by the crop region specified in the request. Each shading map 7661 entry is the value of the shading compensation map over a specific 7662 pixel on the sensor. Specifically, with a (N x M) resolution shading 7663 map, and an active pixel array size (W x H), shading map entry 7664 (x,y) (0 ... N-1, 0 ... M-1) is the value of the shading map at 7665 pixel ( ((W-1)/(N-1)) * x, ((H-1)/(M-1)) * y) for the four color channels. 7666 The map is assumed to be bilinearly interpolated between the sample points. 7667 7668 The channel order is [R, Geven, Godd, B], where Geven is the green 7669 channel for the even rows of a Bayer pattern, and Godd is the odd rows. 7670 The shading map is stored in a fully interleaved format. 7671 7672 The shading map will generally have on the order of 30-40 rows and columns, 7673 and will be smaller than 64x64. 7674 7675 As an example, given a very small map defined as: 7676 7677 width,height = [ 4, 3 ] 7678 values = 7679 [ 1.3, 1.2, 1.15, 1.2, 1.2, 1.2, 1.15, 1.2, 7680 1.1, 1.2, 1.2, 1.2, 1.3, 1.2, 1.3, 1.3, 7681 1.2, 1.2, 1.25, 1.1, 1.1, 1.1, 1.1, 1.0, 7682 1.0, 1.0, 1.0, 1.0, 1.2, 1.3, 1.25, 1.2, 7683 1.3, 1.2, 1.2, 1.3, 1.2, 1.15, 1.1, 1.2, 7684 1.2, 1.1, 1.0, 1.2, 1.3, 1.15, 1.2, 1.3 ] 7685 7686 The low-resolution scaling map images for each channel are 7687 (displayed using nearest-neighbor interpolation): 7688 7689  7690  7691  7692  7693 7694 As a visualization only, inverting the full-color map to recover an 7695 image of a gray wall (using bicubic interpolation for visual quality) as captured by the sensor gives: 7696 7697  7698 </details> 7699 </entry> 7700 <entry name="lensShadingMap" type="float" visibility="ndk_public" 7701 type_notes="2D array of float gain factors per channel to correct lens shading" 7702 container="array" hwlevel="full"> 7703 <array> 7704 <size>4</size> 7705 <size>n</size> 7706 <size>m</size> 7707 </array> 7708 <description>The shading map is a low-resolution floating-point map 7709 that lists the coefficients used to correct for vignetting and color shading, 7710 for each Bayer color channel of RAW image data.</description> 7711 <range>Each gain factor is &gt;= 1</range> 7712 <details> 7713 The map provided here is the same map that is used by the camera device to 7714 correct both color shading and vignetting for output non-RAW images. 7715 7716 When there is no lens shading correction applied to RAW 7717 output images (android.sensor.info.lensShadingApplied `==` 7718 false), this map is the complete lens shading correction 7719 map; when there is some lens shading correction applied to 7720 the RAW output image (android.sensor.info.lensShadingApplied 7721 `==` true), this map reports the remaining lens shading 7722 correction map that needs to be applied to get shading 7723 corrected images that match the camera device's output for 7724 non-RAW formats. 7725 7726 For a complete shading correction map, the least shaded 7727 section of the image will have a gain factor of 1; all 7728 other sections will have gains above 1. 7729 7730 When android.colorCorrection.mode = TRANSFORM_MATRIX, the map 7731 will take into account the colorCorrection settings. 7732 7733 The shading map is for the entire active pixel array, and is not 7734 affected by the crop region specified in the request. Each shading map 7735 entry is the value of the shading compensation map over a specific 7736 pixel on the sensor. Specifically, with a (N x M) resolution shading 7737 map, and an active pixel array size (W x H), shading map entry 7738 (x,y) (0 ... N-1, 0 ... M-1) is the value of the shading map at 7739 pixel ( ((W-1)/(N-1)) * x, ((H-1)/(M-1)) * y) for the four color channels. 7740 The map is assumed to be bilinearly interpolated between the sample points. 7741 7742 The channel order is [R, Geven, Godd, B], where Geven is the green 7743 channel for the even rows of a Bayer pattern, and Godd is the odd rows. 7744 The shading map is stored in a fully interleaved format, and its size 7745 is provided in the camera static metadata by android.lens.info.shadingMapSize. 7746 7747 The shading map will generally have on the order of 30-40 rows and columns, 7748 and will be smaller than 64x64. 7749 7750 As an example, given a very small map defined as: 7751 7752 android.lens.info.shadingMapSize = [ 4, 3 ] 7753 android.statistics.lensShadingMap = 7754 [ 1.3, 1.2, 1.15, 1.2, 1.2, 1.2, 1.15, 1.2, 7755 1.1, 1.2, 1.2, 1.2, 1.3, 1.2, 1.3, 1.3, 7756 1.2, 1.2, 1.25, 1.1, 1.1, 1.1, 1.1, 1.0, 7757 1.0, 1.0, 1.0, 1.0, 1.2, 1.3, 1.25, 1.2, 7758 1.3, 1.2, 1.2, 1.3, 1.2, 1.15, 1.1, 1.2, 7759 1.2, 1.1, 1.0, 1.2, 1.3, 1.15, 1.2, 1.3 ] 7760 7761 The low-resolution scaling map images for each channel are 7762 (displayed using nearest-neighbor interpolation): 7763 7764  7765  7766  7767  7768 7769 As a visualization only, inverting the full-color map to recover an 7770 image of a gray wall (using bicubic interpolation for visual quality) 7771 as captured by the sensor gives: 7772 7773  7774 7775 Note that the RAW image data might be subject to lens shading 7776 correction not reported on this map. Query 7777 android.sensor.info.lensShadingApplied to see if RAW image data has subject 7778 to lens shading correction. If android.sensor.info.lensShadingApplied 7779 is TRUE, the RAW image data is subject to partial or full lens shading 7780 correction. In the case full lens shading correction is applied to RAW 7781 images, the gain factor map reported in this key will contain all 1.0 gains. 7782 In other words, the map reported in this key is the remaining lens shading 7783 that needs to be applied on the RAW image to get images without lens shading 7784 artifacts. See android.request.maxNumOutputRaw for a list of RAW image 7785 formats. 7786 </details> 7787 <hal_details> 7788 The lens shading map calculation may depend on exposure and white balance statistics. 7789 When AE and AWB are in AUTO modes 7790 (android.control.aeMode `!=` OFF and android.control.awbMode `!=` OFF), the HAL 7791 may have all the information it need to generate most accurate lens shading map. When 7792 AE or AWB are in manual mode 7793 (android.control.aeMode `==` OFF or android.control.awbMode `==` OFF), the shading map 7794 may be adversely impacted by manual exposure or white balance parameters. To avoid 7795 generating unreliable shading map data, the HAL may choose to lock the shading map with 7796 the latest known good map generated when the AE and AWB are in AUTO modes. 7797 </hal_details> 7798 </entry> 7799 <entry name="predictedColorGains" type="float" 7800 visibility="hidden" 7801 deprecated="true" 7802 optional="true" 7803 type_notes="A 1D array of floats for 4 color channel gains" 7804 container="array"> 7805 <array> 7806 <size>4</size> 7807 </array> 7808 <description>The best-fit color channel gains calculated 7809 by the camera device's statistics units for the current output frame. 7810 </description> 7811 <details> 7812 This may be different than the gains used for this frame, 7813 since statistics processing on data from a new frame 7814 typically completes after the transform has already been 7815 applied to that frame. 7816 7817 The 4 channel gains are defined in Bayer domain, 7818 see android.colorCorrection.gains for details. 7819 7820 This value should always be calculated by the auto-white balance (AWB) block, 7821 regardless of the android.control.* current values. 7822 </details> 7823 </entry> 7824 <entry name="predictedColorTransform" type="rational" 7825 visibility="hidden" 7826 deprecated="true" 7827 optional="true" 7828 type_notes="3x3 rational matrix in row-major order" 7829 container="array"> 7830 <array> 7831 <size>3</size> 7832 <size>3</size> 7833 </array> 7834 <description>The best-fit color transform matrix estimate 7835 calculated by the camera device's statistics units for the current 7836 output frame.</description> 7837 <details>The camera device will provide the estimate from its 7838 statistics unit on the white balance transforms to use 7839 for the next frame. These are the values the camera device believes 7840 are the best fit for the current output frame. This may 7841 be different than the transform used for this frame, since 7842 statistics processing on data from a new frame typically 7843 completes after the transform has already been applied to 7844 that frame. 7845 7846 These estimates must be provided for all frames, even if 7847 capture settings and color transforms are set by the application. 7848 7849 This value should always be calculated by the auto-white balance (AWB) block, 7850 regardless of the android.control.* current values. 7851 </details> 7852 </entry> 7853 <entry name="sceneFlicker" type="byte" visibility="public" enum="true" 7854 hwlevel="full"> 7855 <enum> 7856 <value>NONE 7857 <notes>The camera device does not detect any flickering illumination 7858 in the current scene.</notes></value> 7859 <value>50HZ 7860 <notes>The camera device detects illumination flickering at 50Hz 7861 in the current scene.</notes></value> 7862 <value>60HZ 7863 <notes>The camera device detects illumination flickering at 60Hz 7864 in the current scene.</notes></value> 7865 </enum> 7866 <description>The camera device estimated scene illumination lighting 7867 frequency.</description> 7868 <details> 7869 Many light sources, such as most fluorescent lights, flicker at a rate 7870 that depends on the local utility power standards. This flicker must be 7871 accounted for by auto-exposure routines to avoid artifacts in captured images. 7872 The camera device uses this entry to tell the application what the scene 7873 illuminant frequency is. 7874 7875 When manual exposure control is enabled 7876 (`android.control.aeMode == OFF` or `android.control.mode == 7877 OFF`), the android.control.aeAntibandingMode doesn't perform 7878 antibanding, and the application can ensure it selects 7879 exposure times that do not cause banding issues by looking 7880 into this metadata field. See 7881 android.control.aeAntibandingMode for more details. 7882 7883 Reports NONE if there doesn't appear to be flickering illumination. 7884 </details> 7885 </entry> 7886 <clone entry="android.statistics.hotPixelMapMode" kind="controls"> 7887 </clone> 7888 <entry name="hotPixelMap" type="int32" visibility="public" 7889 type_notes="list of coordinates based on android.sensor.pixelArraySize" 7890 container="array" typedef="point"> 7891 <array> 7892 <size>2</size> 7893 <size>n</size> 7894 </array> 7895 <description> 7896 List of `(x, y)` coordinates of hot/defective pixels on the sensor. 7897 </description> 7898 <range> 7899 n <= number of pixels on the sensor. 7900 The `(x, y)` coordinates must be bounded by 7901 android.sensor.info.pixelArraySize. 7902 </range> 7903 <details> 7904 A coordinate `(x, y)` must lie between `(0, 0)`, and 7905 `(width - 1, height - 1)` (inclusive), which are the top-left and 7906 bottom-right of the pixel array, respectively. The width and 7907 height dimensions are given in android.sensor.info.pixelArraySize. 7908 This may include hot pixels that lie outside of the active array 7909 bounds given by android.sensor.info.activeArraySize. 7910 </details> 7911 <hal_details> 7912 A hotpixel map contains the coordinates of pixels on the camera 7913 sensor that do report valid values (usually due to defects in 7914 the camera sensor). This includes pixels that are stuck at certain 7915 values, or have a response that does not accuractly encode the 7916 incoming light from the scene. 7917 7918 To avoid performance issues, there should be significantly fewer hot 7919 pixels than actual pixels on the camera sensor. 7920 </hal_details> 7921 <tag id="V1" /> 7922 <tag id="RAW" /> 7923 </entry> 7924 </dynamic> 7925 <controls> 7926 <entry name="lensShadingMapMode" type="byte" visibility="public" enum="true" hwlevel="full"> 7927 <enum> 7928 <value>OFF 7929 <notes>Do not include a lens shading map in the capture result.</notes></value> 7930 <value>ON 7931 <notes>Include a lens shading map in the capture result.</notes></value> 7932 </enum> 7933 <description>Whether the camera device will output the lens 7934 shading map in output result metadata.</description> 7935 <range>android.statistics.info.availableLensShadingMapModes</range> 7936 <details>When set to ON, 7937 android.statistics.lensShadingMap will be provided in 7938 the output result metadata. 7939 7940 ON is always supported on devices with the RAW capability. 7941 </details> 7942 <tag id="RAW" /> 7943 </entry> 7944 </controls> 7945 <dynamic> 7946 <clone entry="android.statistics.lensShadingMapMode" kind="controls"> 7947 </clone> 7948 </dynamic> 7949 </section> 7950 <section name="tonemap"> 7951 <controls> 7952 <entry name="curveBlue" type="float" visibility="ndk_public" 7953 type_notes="1D array of float pairs (P_IN, P_OUT). The maximum number of pairs is specified by android.tonemap.maxCurvePoints." 7954 container="array" hwlevel="full"> 7955 <array> 7956 <size>n</size> 7957 <size>2</size> 7958 </array> 7959 <description>Tonemapping / contrast / gamma curve for the blue 7960 channel, to use when android.tonemap.mode is 7961 CONTRAST_CURVE.</description> 7962 <details>See android.tonemap.curveRed for more details.</details> 7963 </entry> 7964 <entry name="curveGreen" type="float" visibility="ndk_public" 7965 type_notes="1D array of float pairs (P_IN, P_OUT). The maximum number of pairs is specified by android.tonemap.maxCurvePoints." 7966 container="array" hwlevel="full"> 7967 <array> 7968 <size>n</size> 7969 <size>2</size> 7970 </array> 7971 <description>Tonemapping / contrast / gamma curve for the green 7972 channel, to use when android.tonemap.mode is 7973 CONTRAST_CURVE.</description> 7974 <details>See android.tonemap.curveRed for more details.</details> 7975 </entry> 7976 <entry name="curveRed" type="float" visibility="ndk_public" 7977 type_notes="1D array of float pairs (P_IN, P_OUT). The maximum number of pairs is specified by android.tonemap.maxCurvePoints." 7978 container="array" hwlevel="full"> 7979 <array> 7980 <size>n</size> 7981 <size>2</size> 7982 </array> 7983 <description>Tonemapping / contrast / gamma curve for the red 7984 channel, to use when android.tonemap.mode is 7985 CONTRAST_CURVE.</description> 7986 <range>0-1 on both input and output coordinates, normalized 7987 as a floating-point value such that 0 == black and 1 == white. 7988 </range> 7989 <details> 7990 Each channel's curve is defined by an array of control points: 7991 7992 android.tonemap.curveRed = 7993 [ P0in, P0out, P1in, P1out, P2in, P2out, P3in, P3out, ..., PNin, PNout ] 7994 2 <= N <= android.tonemap.maxCurvePoints 7995 7996 These are sorted in order of increasing `Pin`; it is 7997 required that input values 0.0 and 1.0 are included in the list to 7998 define a complete mapping. For input values between control points, 7999 the camera device must linearly interpolate between the control 8000 points. 8001 8002 Each curve can have an independent number of points, and the number 8003 of points can be less than max (that is, the request doesn't have to 8004 always provide a curve with number of points equivalent to 8005 android.tonemap.maxCurvePoints). 8006 8007 A few examples, and their corresponding graphical mappings; these 8008 only specify the red channel and the precision is limited to 4 8009 digits, for conciseness. 8010 8011 Linear mapping: 8012 8013 android.tonemap.curveRed = [ 0, 0, 1.0, 1.0 ] 8014 8015  8016 8017 Invert mapping: 8018 8019 android.tonemap.curveRed = [ 0, 1.0, 1.0, 0 ] 8020 8021  8022 8023 Gamma 1/2.2 mapping, with 16 control points: 8024 8025 android.tonemap.curveRed = [ 8026 0.0000, 0.0000, 0.0667, 0.2920, 0.1333, 0.4002, 0.2000, 0.4812, 8027 0.2667, 0.5484, 0.3333, 0.6069, 0.4000, 0.6594, 0.4667, 0.7072, 8028 0.5333, 0.7515, 0.6000, 0.7928, 0.6667, 0.8317, 0.7333, 0.8685, 8029 0.8000, 0.9035, 0.8667, 0.9370, 0.9333, 0.9691, 1.0000, 1.0000 ] 8030 8031  8032 8033 Standard sRGB gamma mapping, per IEC 61966-2-1:1999, with 16 control points: 8034 8035 android.tonemap.curveRed = [ 8036 0.0000, 0.0000, 0.0667, 0.2864, 0.1333, 0.4007, 0.2000, 0.4845, 8037 0.2667, 0.5532, 0.3333, 0.6125, 0.4000, 0.6652, 0.4667, 0.7130, 8038 0.5333, 0.7569, 0.6000, 0.7977, 0.6667, 0.8360, 0.7333, 0.8721, 8039 0.8000, 0.9063, 0.8667, 0.9389, 0.9333, 0.9701, 1.0000, 1.0000 ] 8040 8041  8042 </details> 8043 <hal_details> 8044 For good quality of mapping, at least 128 control points are 8045 preferred. 8046 8047 A typical use case of this would be a gamma-1/2.2 curve, with as many 8048 control points used as are available. 8049 </hal_details> 8050 </entry> 8051 <entry name="curve" type="float" visibility="java_public" synthetic="true" 8052 typedef="tonemapCurve" 8053 hwlevel="full"> 8054 <description>Tonemapping / contrast / gamma curve to use when android.tonemap.mode 8055 is CONTRAST_CURVE.</description> 8056 <details> 8057 The tonemapCurve consist of three curves for each of red, green, and blue 8058 channels respectively. The following example uses the red channel as an 8059 example. The same logic applies to green and blue channel. 8060 Each channel's curve is defined by an array of control points: 8061 8062 curveRed = 8063 [ P0(in, out), P1(in, out), P2(in, out), P3(in, out), ..., PN(in, out) ] 8064 2 <= N <= android.tonemap.maxCurvePoints 8065 8066 These are sorted in order of increasing `Pin`; it is always 8067 guaranteed that input values 0.0 and 1.0 are included in the list to 8068 define a complete mapping. For input values between control points, 8069 the camera device must linearly interpolate between the control 8070 points. 8071 8072 Each curve can have an independent number of points, and the number 8073 of points can be less than max (that is, the request doesn't have to 8074 always provide a curve with number of points equivalent to 8075 android.tonemap.maxCurvePoints). 8076 8077 A few examples, and their corresponding graphical mappings; these 8078 only specify the red channel and the precision is limited to 4 8079 digits, for conciseness. 8080 8081 Linear mapping: 8082 8083 curveRed = [ (0, 0), (1.0, 1.0) ] 8084 8085  8086 8087 Invert mapping: 8088 8089 curveRed = [ (0, 1.0), (1.0, 0) ] 8090 8091  8092 8093 Gamma 1/2.2 mapping, with 16 control points: 8094 8095 curveRed = [ 8096 (0.0000, 0.0000), (0.0667, 0.2920), (0.1333, 0.4002), (0.2000, 0.4812), 8097 (0.2667, 0.5484), (0.3333, 0.6069), (0.4000, 0.6594), (0.4667, 0.7072), 8098 (0.5333, 0.7515), (0.6000, 0.7928), (0.6667, 0.8317), (0.7333, 0.8685), 8099 (0.8000, 0.9035), (0.8667, 0.9370), (0.9333, 0.9691), (1.0000, 1.0000) ] 8100 8101  8102 8103 Standard sRGB gamma mapping, per IEC 61966-2-1:1999, with 16 control points: 8104 8105 curveRed = [ 8106 (0.0000, 0.0000), (0.0667, 0.2864), (0.1333, 0.4007), (0.2000, 0.4845), 8107 (0.2667, 0.5532), (0.3333, 0.6125), (0.4000, 0.6652), (0.4667, 0.7130), 8108 (0.5333, 0.7569), (0.6000, 0.7977), (0.6667, 0.8360), (0.7333, 0.8721), 8109 (0.8000, 0.9063), (0.8667, 0.9389), (0.9333, 0.9701), (1.0000, 1.0000) ] 8110 8111  8112 </details> 8113 <hal_details> 8114 This entry is created by the framework from the curveRed, curveGreen and 8115 curveBlue entries. 8116 </hal_details> 8117 </entry> 8118 <entry name="mode" type="byte" visibility="public" enum="true" 8119 hwlevel="full"> 8120 <enum> 8121 <value>CONTRAST_CURVE 8122 <notes>Use the tone mapping curve specified in 8123 the android.tonemap.curve* entries. 8124 8125 All color enhancement and tonemapping must be disabled, except 8126 for applying the tonemapping curve specified by 8127 android.tonemap.curve. 8128 8129 Must not slow down frame rate relative to raw 8130 sensor output. 8131 </notes> 8132 </value> 8133 <value>FAST 8134 <notes> 8135 Advanced gamma mapping and color enhancement may be applied, without 8136 reducing frame rate compared to raw sensor output. 8137 </notes> 8138 </value> 8139 <value>HIGH_QUALITY 8140 <notes> 8141 High-quality gamma mapping and color enhancement will be applied, at 8142 the cost of possibly reduced frame rate compared to raw sensor output. 8143 </notes> 8144 </value> 8145 <value>GAMMA_VALUE 8146 <notes> 8147 Use the gamma value specified in android.tonemap.gamma to peform 8148 tonemapping. 8149 8150 All color enhancement and tonemapping must be disabled, except 8151 for applying the tonemapping curve specified by android.tonemap.gamma. 8152 8153 Must not slow down frame rate relative to raw sensor output. 8154 </notes> 8155 </value> 8156 <value>PRESET_CURVE 8157 <notes> 8158 Use the preset tonemapping curve specified in 8159 android.tonemap.presetCurve to peform tonemapping. 8160 8161 All color enhancement and tonemapping must be disabled, except 8162 for applying the tonemapping curve specified by 8163 android.tonemap.presetCurve. 8164 8165 Must not slow down frame rate relative to raw sensor output. 8166 </notes> 8167 </value> 8168 </enum> 8169 <description>High-level global contrast/gamma/tonemapping control. 8170 </description> 8171 <range>android.tonemap.availableToneMapModes</range> 8172 <details> 8173 When switching to an application-defined contrast curve by setting 8174 android.tonemap.mode to CONTRAST_CURVE, the curve is defined 8175 per-channel with a set of `(in, out)` points that specify the 8176 mapping from input high-bit-depth pixel value to the output 8177 low-bit-depth value. Since the actual pixel ranges of both input 8178 and output may change depending on the camera pipeline, the values 8179 are specified by normalized floating-point numbers. 8180 8181 More-complex color mapping operations such as 3D color look-up 8182 tables, selective chroma enhancement, or other non-linear color 8183 transforms will be disabled when android.tonemap.mode is 8184 CONTRAST_CURVE. 8185 8186 When using either FAST or HIGH_QUALITY, the camera device will 8187 emit its own tonemap curve in android.tonemap.curve. 8188 These values are always available, and as close as possible to the 8189 actually used nonlinear/nonglobal transforms. 8190 8191 If a request is sent with CONTRAST_CURVE with the camera device's 8192 provided curve in FAST or HIGH_QUALITY, the image's tonemap will be 8193 roughly the same.</details> 8194 </entry> 8195 </controls> 8196 <static> 8197 <entry name="maxCurvePoints" type="int32" visibility="public" 8198 hwlevel="full"> 8199 <description>Maximum number of supported points in the 8200 tonemap curve that can be used for android.tonemap.curve. 8201 </description> 8202 <details> 8203 If the actual number of points provided by the application (in android.tonemap.curve*) is 8204 less than this maximum, the camera device will resample the curve to its internal 8205 representation, using linear interpolation. 8206 8207 The output curves in the result metadata may have a different number 8208 of points than the input curves, and will represent the actual 8209 hardware curves used as closely as possible when linearly interpolated. 8210 </details> 8211 <hal_details> 8212 This value must be at least 64. This should be at least 128. 8213 </hal_details> 8214 </entry> 8215 <entry name="availableToneMapModes" type="byte" visibility="public" 8216 type_notes="list of enums" container="array" typedef="enumList" hwlevel="full"> 8217 <array> 8218 <size>n</size> 8219 </array> 8220 <description> 8221 List of tonemapping modes for android.tonemap.mode that are supported by this camera 8222 device. 8223 </description> 8224 <range>Any value listed in android.tonemap.mode</range> 8225 <details> 8226 Camera devices that support the MANUAL_POST_PROCESSING capability will always contain 8227 at least one of below mode combinations: 8228 8229 * CONTRAST_CURVE, FAST and HIGH_QUALITY 8230 * GAMMA_VALUE, PRESET_CURVE, FAST and HIGH_QUALITY 8231 8232 This includes all FULL level devices. 8233 </details> 8234 <hal_details> 8235 HAL must support both FAST and HIGH_QUALITY if automatic tonemap control is available 8236 on the camera device, but the underlying implementation can be the same for both modes. 8237 That is, if the highest quality implementation on the camera device does not slow down 8238 capture rate, then FAST and HIGH_QUALITY will generate the same output. 8239 </hal_details> 8240 </entry> 8241 </static> 8242 <dynamic> 8243 <clone entry="android.tonemap.curveBlue" kind="controls"> 8244 </clone> 8245 <clone entry="android.tonemap.curveGreen" kind="controls"> 8246 </clone> 8247 <clone entry="android.tonemap.curveRed" kind="controls"> 8248 </clone> 8249 <clone entry="android.tonemap.curve" kind="controls"> 8250 </clone> 8251 <clone entry="android.tonemap.mode" kind="controls"> 8252 </clone> 8253 </dynamic> 8254 <controls> 8255 <entry name="gamma" type="float" visibility="public"> 8256 <description> Tonemapping curve to use when android.tonemap.mode is 8257 GAMMA_VALUE 8258 </description> 8259 <details> 8260 The tonemap curve will be defined the following formula: 8261 * OUT = pow(IN, 1.0 / gamma) 8262 where IN and OUT is the input pixel value scaled to range [0.0, 1.0], 8263 pow is the power function and gamma is the gamma value specified by this 8264 key. 8265 8266 The same curve will be applied to all color channels. The camera device 8267 may clip the input gamma value to its supported range. The actual applied 8268 value will be returned in capture result. 8269 8270 The valid range of gamma value varies on different devices, but values 8271 within [1.0, 5.0] are guaranteed not to be clipped. 8272 </details> 8273 </entry> 8274 <entry name="presetCurve" type="byte" visibility="public" enum="true"> 8275 <enum> 8276 <value>SRGB 8277 <notes>Tonemapping curve is defined by sRGB</notes> 8278 </value> 8279 <value>REC709 8280 <notes>Tonemapping curve is defined by ITU-R BT.709</notes> 8281 </value> 8282 </enum> 8283 <description> Tonemapping curve to use when android.tonemap.mode is 8284 PRESET_CURVE 8285 </description> 8286 <details> 8287 The tonemap curve will be defined by specified standard. 8288 8289 sRGB (approximated by 16 control points): 8290 8291  8292 8293 Rec. 709 (approximated by 16 control points): 8294 8295  8296 8297 Note that above figures show a 16 control points approximation of preset 8298 curves. Camera devices may apply a different approximation to the curve. 8299 </details> 8300 </entry> 8301 </controls> 8302 <dynamic> 8303 <clone entry="android.tonemap.gamma" kind="controls"> 8304 </clone> 8305 <clone entry="android.tonemap.presetCurve" kind="controls"> 8306 </clone> 8307 </dynamic> 8308 </section> 8309 <section name="led"> 8310 <controls> 8311 <entry name="transmit" type="byte" visibility="hidden" optional="true" 8312 enum="true" typedef="boolean"> 8313 <enum> 8314 <value>OFF</value> 8315 <value>ON</value> 8316 </enum> 8317 <description>This LED is nominally used to indicate to the user 8318 that the camera is powered on and may be streaming images back to the 8319 Application Processor. In certain rare circumstances, the OS may 8320 disable this when video is processed locally and not transmitted to 8321 any untrusted applications. 8322 8323 In particular, the LED *must* always be on when the data could be 8324 transmitted off the device. The LED *should* always be on whenever 8325 data is stored locally on the device. 8326 8327 The LED *may* be off if a trusted application is using the data that 8328 doesn't violate the above rules. 8329 </description> 8330 </entry> 8331 </controls> 8332 <dynamic> 8333 <clone entry="android.led.transmit" kind="controls"></clone> 8334 </dynamic> 8335 <static> 8336 <entry name="availableLeds" type="byte" visibility="hidden" optional="true" 8337 enum="true" 8338 container="array"> 8339 <array> 8340 <size>n</size> 8341 </array> 8342 <enum> 8343 <value>TRANSMIT 8344 <notes>android.led.transmit control is used.</notes> 8345 </value> 8346 </enum> 8347 <description>A list of camera LEDs that are available on this system. 8348 </description> 8349 </entry> 8350 </static> 8351 </section> 8352 <section name="info"> 8353 <static> 8354 <entry name="supportedHardwareLevel" type="byte" visibility="public" 8355 enum="true" hwlevel="legacy"> 8356 <enum> 8357 <value> 8358 LIMITED 8359 <notes> 8360 This camera device does not have enough capabilities to qualify as a `FULL` device or 8361 better. 8362 8363 Only the stream configurations listed in the `LEGACY` and `LIMITED` tables in the 8364 {@link android.hardware.camera2.CameraDevice#createCaptureSession 8365 createCaptureSession} documentation are guaranteed to be supported. 8366 8367 All `LIMITED` devices support the `BACKWARDS_COMPATIBLE` capability, indicating basic 8368 support for color image capture. The only exception is that the device may 8369 alternatively support only the `DEPTH_OUTPUT` capability, if it can only output depth 8370 measurements and not color images. 8371 8372 `LIMITED` devices and above require the use of android.control.aePrecaptureTrigger 8373 to lock exposure metering (and calculate flash power, for cameras with flash) before 8374 capturing a high-quality still image. 8375 8376 A `LIMITED` device that only lists the `BACKWARDS_COMPATIBLE` capability is only 8377 required to support full-automatic operation and post-processing (`OFF` is not 8378 supported for android.control.aeMode, android.control.afMode, or 8379 android.control.awbMode) 8380 8381 Additional capabilities may optionally be supported by a `LIMITED`-level device, and 8382 can be checked for in android.request.availableCapabilities. 8383 </notes> 8384 </value> 8385 <value> 8386 FULL 8387 <notes> 8388 This camera device is capable of supporting advanced imaging applications. 8389 8390 The stream configurations listed in the `FULL`, `LEGACY` and `LIMITED` tables in the 8391 {@link android.hardware.camera2.CameraDevice#createCaptureSession 8392 createCaptureSession} documentation are guaranteed to be supported. 8393 8394 A `FULL` device will support below capabilities: 8395 8396 * `BURST_CAPTURE` capability (android.request.availableCapabilities contains 8397 `BURST_CAPTURE`) 8398 * Per frame control (android.sync.maxLatency `==` PER_FRAME_CONTROL) 8399 * Manual sensor control (android.request.availableCapabilities contains `MANUAL_SENSOR`) 8400 * Manual post-processing control (android.request.availableCapabilities contains 8401 `MANUAL_POST_PROCESSING`) 8402 * The required exposure time range defined in android.sensor.info.exposureTimeRange 8403 * The required maxFrameDuration defined in android.sensor.info.maxFrameDuration 8404 8405 Note: 8406 Pre-API level 23, FULL devices also supported arbitrary cropping region 8407 (android.scaler.croppingType `== FREEFORM`); this requirement was relaxed in API level 8408 23, and `FULL` devices may only support `CENTERED` cropping. 8409 </notes> 8410 </value> 8411 <value> 8412 LEGACY 8413 <notes> 8414 This camera device is running in backward compatibility mode. 8415 8416 Only the stream configurations listed in the `LEGACY` table in the {@link 8417 android.hardware.camera2.CameraDevice#createCaptureSession createCaptureSession} 8418 documentation are supported. 8419 8420 A `LEGACY` device does not support per-frame control, manual sensor control, manual 8421 post-processing, arbitrary cropping regions, and has relaxed performance constraints. 8422 No additional capabilities beyond `BACKWARD_COMPATIBLE` will ever be listed by a 8423 `LEGACY` device in android.request.availableCapabilities. 8424 8425 In addition, the android.control.aePrecaptureTrigger is not functional on `LEGACY` 8426 devices. Instead, every request that includes a JPEG-format output target is treated 8427 as triggering a still capture, internally executing a precapture trigger. This may 8428 fire the flash for flash power metering during precapture, and then fire the flash 8429 for the final capture, if a flash is available on the device and the AE mode is set to 8430 enable the flash. 8431 </notes> 8432 </value> 8433 <value> 8434 3 8435 <notes> 8436 This camera device is capable of YUV reprocessing and RAW data capture, in addition to 8437 FULL-level capabilities. 8438 8439 The stream configurations listed in the `LEVEL_3`, `RAW`, `FULL`, `LEGACY` and 8440 `LIMITED` tables in the {@link 8441 android.hardware.camera2.CameraDevice#createCaptureSession createCaptureSession} 8442 documentation are guaranteed to be supported. 8443 8444 The following additional capabilities are guaranteed to be supported: 8445 8446 * `YUV_REPROCESSING` capability (android.request.availableCapabilities contains 8447 `YUV_REPROCESSING`) 8448 * `RAW` capability (android.request.availableCapabilities contains 8449 `RAW`) 8450 </notes> 8451 </value> 8452 </enum> 8453 <description> 8454 Generally classifies the overall set of the camera device functionality. 8455 </description> 8456 <details> 8457 The supported hardware level is a high-level description of the camera device's 8458 capabilities, summarizing several capabilities into one field. Each level adds additional 8459 features to the previous one, and is always a strict superset of the previous level. 8460 The ordering is `LEGACY < LIMITED < FULL < LEVEL_3`. 8461 8462 Starting from `LEVEL_3`, the level enumerations are guaranteed to be in increasing 8463 numerical value as well. To check if a given device is at least at a given hardware level, 8464 the following code snippet can be used: 8465 8466 // Returns true if the device supports the required hardware level, or better. 8467 boolean isHardwareLevelSupported(CameraCharacteristics c, int requiredLevel) { 8468 int deviceLevel = c.get(CameraCharacteristics.INFO_SUPPORTED_HARDWARE_LEVEL); 8469 if (deviceLevel == CameraCharacteristics.INFO_SUPPORTED_HARDWARE_LEVEL_LEGACY) { 8470 return requiredLevel == deviceLevel; 8471 } 8472 // deviceLevel is not LEGACY, can use numerical sort 8473 return requiredLevel <= deviceLevel; 8474 } 8475 8476 At a high level, the levels are: 8477 8478 * `LEGACY` devices operate in a backwards-compatibility mode for older 8479 Android devices, and have very limited capabilities. 8480 * `LIMITED` devices represent the 8481 baseline feature set, and may also include additional capabilities that are 8482 subsets of `FULL`. 8483 * `FULL` devices additionally support per-frame manual control of sensor, flash, lens and 8484 post-processing settings, and image capture at a high rate. 8485 * `LEVEL_3` devices additionally support YUV reprocessing and RAW image capture, along 8486 with additional output stream configurations. 8487 8488 See the individual level enums for full descriptions of the supported capabilities. The 8489 android.request.availableCapabilities entry describes the device's capabilities at a 8490 finer-grain level, if needed. In addition, many controls have their available settings or 8491 ranges defined in individual {@link android.hardware.camera2.CameraCharacteristics} entries. 8492 8493 Some features are not part of any particular hardware level or capability and must be 8494 queried separately. These include: 8495 8496 * Calibrated timestamps (android.sensor.info.timestampSource `==` REALTIME) 8497 * Precision lens control (android.lens.info.focusDistanceCalibration `==` CALIBRATED) 8498 * Face detection (android.statistics.info.availableFaceDetectModes) 8499 * Optical or electrical image stabilization 8500 (android.lens.info.availableOpticalStabilization, 8501 android.control.availableVideoStabilizationModes) 8502 8503 </details> 8504 <hal_details> 8505 The camera 3 HAL device can implement one of three possible operational modes; LIMITED, 8506 FULL, and LEVEL_3. 8507 8508 FULL support or better is expected from new higher-end devices. Limited 8509 mode has hardware requirements roughly in line with those for a camera HAL device v1 8510 implementation, and is expected from older or inexpensive devices. Each level is a strict 8511 superset of the previous level, and they share the same essential operational flow. 8512 8513 For full details refer to "S3. Operational Modes" in camera3.h 8514 8515 Camera HAL3+ must not implement LEGACY mode. It is there for backwards compatibility in 8516 the `android.hardware.camera2` user-facing API only on HALv1 devices, and is implemented 8517 by the camera framework code. 8518 </hal_details> 8519 </entry> 8520 </static> 8521 </section> 8522 <section name="blackLevel"> 8523 <controls> 8524 <entry name="lock" type="byte" visibility="public" enum="true" 8525 typedef="boolean" hwlevel="full"> 8526 <enum> 8527 <value>OFF</value> 8528 <value>ON</value> 8529 </enum> 8530 <description> Whether black-level compensation is locked 8531 to its current values, or is free to vary.</description> 8532 <details>When set to `true` (ON), the values used for black-level 8533 compensation will not change until the lock is set to 8534 `false` (OFF). 8535 8536 Since changes to certain capture parameters (such as 8537 exposure time) may require resetting of black level 8538 compensation, the camera device must report whether setting 8539 the black level lock was successful in the output result 8540 metadata. 8541 8542 For example, if a sequence of requests is as follows: 8543 8544 * Request 1: Exposure = 10ms, Black level lock = OFF 8545 * Request 2: Exposure = 10ms, Black level lock = ON 8546 * Request 3: Exposure = 10ms, Black level lock = ON 8547 * Request 4: Exposure = 20ms, Black level lock = ON 8548 * Request 5: Exposure = 20ms, Black level lock = ON 8549 * Request 6: Exposure = 20ms, Black level lock = ON 8550 8551 And the exposure change in Request 4 requires the camera 8552 device to reset the black level offsets, then the output 8553 result metadata is expected to be: 8554 8555 * Result 1: Exposure = 10ms, Black level lock = OFF 8556 * Result 2: Exposure = 10ms, Black level lock = ON 8557 * Result 3: Exposure = 10ms, Black level lock = ON 8558 * Result 4: Exposure = 20ms, Black level lock = OFF 8559 * Result 5: Exposure = 20ms, Black level lock = ON 8560 * Result 6: Exposure = 20ms, Black level lock = ON 8561 8562 This indicates to the application that on frame 4, black 8563 levels were reset due to exposure value changes, and pixel 8564 values may not be consistent across captures. 8565 8566 The camera device will maintain the lock to the extent 8567 possible, only overriding the lock to OFF when changes to 8568 other request parameters require a black level recalculation 8569 or reset. 8570 </details> 8571 <hal_details> 8572 If for some reason black level locking is no longer possible 8573 (for example, the analog gain has changed, which forces 8574 black level offsets to be recalculated), then the HAL must 8575 override this request (and it must report 'OFF' when this 8576 does happen) until the next capture for which locking is 8577 possible again.</hal_details> 8578 <tag id="HAL2" /> 8579 </entry> 8580 </controls> 8581 <dynamic> 8582 <clone entry="android.blackLevel.lock" 8583 kind="controls"> 8584 <details> 8585 Whether the black level offset was locked for this frame. Should be 8586 ON if android.blackLevel.lock was ON in the capture request, unless 8587 a change in other capture settings forced the camera device to 8588 perform a black level reset. 8589 </details> 8590 </clone> 8591 </dynamic> 8592 </section> 8593 <section name="sync"> 8594 <dynamic> 8595 <entry name="frameNumber" type="int64" visibility="ndk_public" 8596 enum="true" hwlevel="legacy"> 8597 <enum> 8598 <value id="-1">CONVERGING 8599 <notes> 8600 The current result is not yet fully synchronized to any request. 8601 8602 Synchronization is in progress, and reading metadata from this 8603 result may include a mix of data that have taken effect since the 8604 last synchronization time. 8605 8606 In some future result, within android.sync.maxLatency frames, 8607 this value will update to the actual frame number frame number 8608 the result is guaranteed to be synchronized to (as long as the 8609 request settings remain constant). 8610 </notes> 8611 </value> 8612 <value id="-2">UNKNOWN 8613 <notes> 8614 The current result's synchronization status is unknown. 8615 8616 The result may have already converged, or it may be in 8617 progress. Reading from this result may include some mix 8618 of settings from past requests. 8619 8620 After a settings change, the new settings will eventually all 8621 take effect for the output buffers and results. However, this 8622 value will not change when that happens. Altering settings 8623 rapidly may provide outcomes using mixes of settings from recent 8624 requests. 8625 8626 This value is intended primarily for backwards compatibility with 8627 the older camera implementations (for android.hardware.Camera). 8628 </notes> 8629 </value> 8630 </enum> 8631 <description>The frame number corresponding to the last request 8632 with which the output result (metadata + buffers) has been fully 8633 synchronized.</description> 8634 <range>Either a non-negative value corresponding to a 8635 `frame_number`, or one of the two enums (CONVERGING / UNKNOWN). 8636 </range> 8637 <details> 8638 When a request is submitted to the camera device, there is usually a 8639 delay of several frames before the controls get applied. A camera 8640 device may either choose to account for this delay by implementing a 8641 pipeline and carefully submit well-timed atomic control updates, or 8642 it may start streaming control changes that span over several frame 8643 boundaries. 8644 8645 In the latter case, whenever a request's settings change relative to 8646 the previous submitted request, the full set of changes may take 8647 multiple frame durations to fully take effect. Some settings may 8648 take effect sooner (in less frame durations) than others. 8649 8650 While a set of control changes are being propagated, this value 8651 will be CONVERGING. 8652 8653 Once it is fully known that a set of control changes have been 8654 finished propagating, and the resulting updated control settings 8655 have been read back by the camera device, this value will be set 8656 to a non-negative frame number (corresponding to the request to 8657 which the results have synchronized to). 8658 8659 Older camera device implementations may not have a way to detect 8660 when all camera controls have been applied, and will always set this 8661 value to UNKNOWN. 8662 8663 FULL capability devices will always have this value set to the 8664 frame number of the request corresponding to this result. 8665 8666 _Further details_: 8667 8668 * Whenever a request differs from the last request, any future 8669 results not yet returned may have this value set to CONVERGING (this 8670 could include any in-progress captures not yet returned by the camera 8671 device, for more details see pipeline considerations below). 8672 * Submitting a series of multiple requests that differ from the 8673 previous request (e.g. r1, r2, r3 s.t. r1 != r2 != r3) 8674 moves the new synchronization frame to the last non-repeating 8675 request (using the smallest frame number from the contiguous list of 8676 repeating requests). 8677 * Submitting the same request repeatedly will not change this value 8678 to CONVERGING, if it was already a non-negative value. 8679 * When this value changes to non-negative, that means that all of the 8680 metadata controls from the request have been applied, all of the 8681 metadata controls from the camera device have been read to the 8682 updated values (into the result), and all of the graphics buffers 8683 corresponding to this result are also synchronized to the request. 8684 8685 _Pipeline considerations_: 8686 8687 Submitting a request with updated controls relative to the previously 8688 submitted requests may also invalidate the synchronization state 8689 of all the results corresponding to currently in-flight requests. 8690 8691 In other words, results for this current request and up to 8692 android.request.pipelineMaxDepth prior requests may have their 8693 android.sync.frameNumber change to CONVERGING. 8694 </details> 8695 <hal_details> 8696 Using UNKNOWN here is illegal unless android.sync.maxLatency 8697 is also UNKNOWN. 8698 8699 FULL capability devices should simply set this value to the 8700 `frame_number` of the request this result corresponds to. 8701 </hal_details> 8702 <tag id="V1" /> 8703 </entry> 8704 </dynamic> 8705 <static> 8706 <entry name="maxLatency" type="int32" visibility="public" enum="true" 8707 hwlevel="legacy"> 8708 <enum> 8709 <value id="0">PER_FRAME_CONTROL 8710 <notes> 8711 Every frame has the requests immediately applied. 8712 8713 Changing controls over multiple requests one after another will 8714 produce results that have those controls applied atomically 8715 each frame. 8716 8717 All FULL capability devices will have this as their maxLatency. 8718 </notes> 8719 </value> 8720 <value id="-1">UNKNOWN 8721 <notes> 8722 Each new frame has some subset (potentially the entire set) 8723 of the past requests applied to the camera settings. 8724 8725 By submitting a series of identical requests, the camera device 8726 will eventually have the camera settings applied, but it is 8727 unknown when that exact point will be. 8728 8729 All LEGACY capability devices will have this as their maxLatency. 8730 </notes> 8731 </value> 8732 </enum> 8733 <description> 8734 The maximum number of frames that can occur after a request 8735 (different than the previous) has been submitted, and before the 8736 result's state becomes synchronized. 8737 </description> 8738 <units>Frame counts</units> 8739 <range>A positive value, PER_FRAME_CONTROL, or UNKNOWN.</range> 8740 <details> 8741 This defines the maximum distance (in number of metadata results), 8742 between the frame number of the request that has new controls to apply 8743 and the frame number of the result that has all the controls applied. 8744 8745 In other words this acts as an upper boundary for how many frames 8746 must occur before the camera device knows for a fact that the new 8747 submitted camera settings have been applied in outgoing frames. 8748 </details> 8749 <hal_details> 8750 For example if maxLatency was 2, 8751 8752 initial request = X (repeating) 8753 request1 = X 8754 request2 = Y 8755 request3 = Y 8756 request4 = Y 8757 8758 where requestN has frameNumber N, and the first of the repeating 8759 initial request's has frameNumber F (and F < 1). 8760 8761 initial result = X' + { android.sync.frameNumber == F } 8762 result1 = X' + { android.sync.frameNumber == F } 8763 result2 = X' + { android.sync.frameNumber == CONVERGING } 8764 result3 = X' + { android.sync.frameNumber == CONVERGING } 8765 result4 = X' + { android.sync.frameNumber == 2 } 8766 8767 where resultN has frameNumber N. 8768 8769 Since `result4` has a `frameNumber == 4` and 8770 `android.sync.frameNumber == 2`, the distance is clearly 8771 `4 - 2 = 2`. 8772 8773 Use `frame_count` from camera3_request_t instead of 8774 android.request.frameCount or 8775 `{@link android.hardware.camera2.CaptureResult#getFrameNumber}`. 8776 8777 LIMITED devices are strongly encouraged to use a non-negative 8778 value. If UNKNOWN is used here then app developers do not have a way 8779 to know when sensor settings have been applied. 8780 </hal_details> 8781 <tag id="V1" /> 8782 </entry> 8783 </static> 8784 </section> 8785 <section name="reprocess"> 8786 <controls> 8787 <entry name="effectiveExposureFactor" type="float" visibility="java_public" hwlevel="limited"> 8788 <description> 8789 The amount of exposure time increase factor applied to the original output 8790 frame by the application processing before sending for reprocessing. 8791 </description> 8792 <units>Relative exposure time increase factor.</units> 8793 <range> &gt;= 1.0</range> 8794 <details> 8795 This is optional, and will be supported if the camera device supports YUV_REPROCESSING 8796 capability (android.request.availableCapabilities contains YUV_REPROCESSING). 8797 8798 For some YUV reprocessing use cases, the application may choose to filter the original 8799 output frames to effectively reduce the noise to the same level as a frame that was 8800 captured with longer exposure time. To be more specific, assuming the original captured 8801 images were captured with a sensitivity of S and an exposure time of T, the model in 8802 the camera device is that the amount of noise in the image would be approximately what 8803 would be expected if the original capture parameters had been a sensitivity of 8804 S/effectiveExposureFactor and an exposure time of T*effectiveExposureFactor, rather 8805 than S and T respectively. If the captured images were processed by the application 8806 before being sent for reprocessing, then the application may have used image processing 8807 algorithms and/or multi-frame image fusion to reduce the noise in the 8808 application-processed images (input images). By using the effectiveExposureFactor 8809 control, the application can communicate to the camera device the actual noise level 8810 improvement in the application-processed image. With this information, the camera 8811 device can select appropriate noise reduction and edge enhancement parameters to avoid 8812 excessive noise reduction (android.noiseReduction.mode) and insufficient edge 8813 enhancement (android.edge.mode) being applied to the reprocessed frames. 8814 8815 For example, for multi-frame image fusion use case, the application may fuse 8816 multiple output frames together to a final frame for reprocessing. When N image are 8817 fused into 1 image for reprocessing, the exposure time increase factor could be up to 8818 square root of N (based on a simple photon shot noise model). The camera device will 8819 adjust the reprocessing noise reduction and edge enhancement parameters accordingly to 8820 produce the best quality images. 8821 8822 This is relative factor, 1.0 indicates the application hasn't processed the input 8823 buffer in a way that affects its effective exposure time. 8824 8825 This control is only effective for YUV reprocessing capture request. For noise 8826 reduction reprocessing, it is only effective when `android.noiseReduction.mode != OFF`. 8827 Similarly, for edge enhancement reprocessing, it is only effective when 8828 `android.edge.mode != OFF`. 8829 </details> 8830 <tag id="REPROC" /> 8831 </entry> 8832 </controls> 8833 <dynamic> 8834 <clone entry="android.reprocess.effectiveExposureFactor" kind="controls"> 8835 </clone> 8836 </dynamic> 8837 <static> 8838 <entry name="maxCaptureStall" type="int32" visibility="java_public" hwlevel="limited"> 8839 <description> 8840 The maximal camera capture pipeline stall (in unit of frame count) introduced by a 8841 reprocess capture request. 8842 </description> 8843 <units>Number of frames.</units> 8844 <range> &lt;= 4</range> 8845 <details> 8846 The key describes the maximal interference that one reprocess (input) request 8847 can introduce to the camera simultaneous streaming of regular (output) capture 8848 requests, including repeating requests. 8849 8850 When a reprocessing capture request is submitted while a camera output repeating request 8851 (e.g. preview) is being served by the camera device, it may preempt the camera capture 8852 pipeline for at least one frame duration so that the camera device is unable to process 8853 the following capture request in time for the next sensor start of exposure boundary. 8854 When this happens, the application may observe a capture time gap (longer than one frame 8855 duration) between adjacent capture output frames, which usually exhibits as preview 8856 glitch if the repeating request output targets include a preview surface. This key gives 8857 the worst-case number of frame stall introduced by one reprocess request with any kind of 8858 formats/sizes combination. 8859 8860 If this key reports 0, it means a reprocess request doesn't introduce any glitch to the 8861 ongoing camera repeating request outputs, as if this reprocess request is never issued. 8862 8863 This key is supported if the camera device supports PRIVATE or YUV reprocessing ( 8864 i.e. android.request.availableCapabilities contains PRIVATE_REPROCESSING or 8865 YUV_REPROCESSING). 8866 </details> 8867 <tag id="REPROC" /> 8868 </entry> 8869 </static> 8870 </section> 8871 <section name="depth"> 8872 <static> 8873 <entry name="maxDepthSamples" type="int32" visibility="system" hwlevel="limited"> 8874 <description>Maximum number of points that a depth point cloud may contain. 8875 </description> 8876 <details> 8877 If a camera device supports outputting depth range data in the form of a depth point 8878 cloud ({@link android.graphics.ImageFormat#DEPTH_POINT_CLOUD}), this is the maximum 8879 number of points an output buffer may contain. 8880 8881 Any given buffer may contain between 0 and maxDepthSamples points, inclusive. 8882 If output in the depth point cloud format is not supported, this entry will 8883 not be defined. 8884 </details> 8885 <tag id="DEPTH" /> 8886 </entry> 8887 <entry name="availableDepthStreamConfigurations" type="int32" visibility="ndk_public" 8888 enum="true" container="array" typedef="streamConfiguration" hwlevel="limited"> 8889 <array> 8890 <size>n</size> 8891 <size>4</size> 8892 </array> 8893 <enum> 8894 <value>OUTPUT</value> 8895 <value>INPUT</value> 8896 </enum> 8897 <description>The available depth dataspace stream 8898 configurations that this camera device supports 8899 (i.e. format, width, height, output/input stream). 8900 </description> 8901 <details> 8902 These are output stream configurations for use with 8903 dataSpace HAL_DATASPACE_DEPTH. The configurations are 8904 listed as `(format, width, height, input?)` tuples. 8905 8906 Only devices that support depth output for at least 8907 the HAL_PIXEL_FORMAT_Y16 dense depth map may include 8908 this entry. 8909 8910 A device that also supports the HAL_PIXEL_FORMAT_BLOB 8911 sparse depth point cloud must report a single entry for 8912 the format in this list as `(HAL_PIXEL_FORMAT_BLOB, 8913 android.depth.maxDepthSamples, 1, OUTPUT)` in addition to 8914 the entries for HAL_PIXEL_FORMAT_Y16. 8915 </details> 8916 <tag id="DEPTH" /> 8917 </entry> 8918 <entry name="availableDepthMinFrameDurations" type="int64" visibility="ndk_public" 8919 container="array" typedef="streamConfigurationDuration" hwlevel="limited"> 8920 <array> 8921 <size>4</size> 8922 <size>n</size> 8923 </array> 8924 <description>This lists the minimum frame duration for each 8925 format/size combination for depth output formats. 8926 </description> 8927 <units>(format, width, height, ns) x n</units> 8928 <details> 8929 This should correspond to the frame duration when only that 8930 stream is active, with all processing (typically in android.*.mode) 8931 set to either OFF or FAST. 8932 8933 When multiple streams are used in a request, the minimum frame 8934 duration will be max(individual stream min durations). 8935 8936 The minimum frame duration of a stream (of a particular format, size) 8937 is the same regardless of whether the stream is input or output. 8938 8939 See android.sensor.frameDuration and 8940 android.scaler.availableStallDurations for more details about 8941 calculating the max frame rate. 8942 8943 (Keep in sync with {@link 8944 android.hardware.camera2.params.StreamConfigurationMap#getOutputMinFrameDuration}) 8945 </details> 8946 <tag id="DEPTH" /> 8947 </entry> 8948 <entry name="availableDepthStallDurations" type="int64" visibility="ndk_public" 8949 container="array" typedef="streamConfigurationDuration" hwlevel="limited"> 8950 <array> 8951 <size>4</size> 8952 <size>n</size> 8953 </array> 8954 <description>This lists the maximum stall duration for each 8955 output format/size combination for depth streams. 8956 </description> 8957 <units>(format, width, height, ns) x n</units> 8958 <details> 8959 A stall duration is how much extra time would get added 8960 to the normal minimum frame duration for a repeating request 8961 that has streams with non-zero stall. 8962 8963 This functions similarly to 8964 android.scaler.availableStallDurations for depth 8965 streams. 8966 8967 All depth output stream formats may have a nonzero stall 8968 duration. 8969 </details> 8970 <tag id="DEPTH" /> 8971 </entry> 8972 <entry name="depthIsExclusive" type="byte" visibility="public" 8973 enum="true" typedef="boolean" hwlevel="limited"> 8974 <enum> 8975 <value>FALSE</value> 8976 <value>TRUE</value> 8977 </enum> 8978 <description>Indicates whether a capture request may target both a 8979 DEPTH16 / DEPTH_POINT_CLOUD output, and normal color outputs (such as 8980 YUV_420_888, JPEG, or RAW) simultaneously. 8981 </description> 8982 <details> 8983 If TRUE, including both depth and color outputs in a single 8984 capture request is not supported. An application must interleave color 8985 and depth requests. If FALSE, a single request can target both types 8986 of output. 8987 8988 Typically, this restriction exists on camera devices that 8989 need to emit a specific pattern or wavelength of light to 8990 measure depth values, which causes the color image to be 8991 corrupted during depth measurement. 8992 </details> 8993 </entry> 8994 </static> 8995 </section> 8996 </namespace> 8997 </metadata> 8998
