1 /* Copyright (c) 2007-2008 CSIRO 2 Copyright (c) 2007-2009 Xiph.Org Foundation 3 Written by Jean-Marc Valin */ 4 /* 5 Redistribution and use in source and binary forms, with or without 6 modification, are permitted provided that the following conditions 7 are met: 8 9 - Redistributions of source code must retain the above copyright 10 notice, this list of conditions and the following disclaimer. 11 12 - Redistributions in binary form must reproduce the above copyright 13 notice, this list of conditions and the following disclaimer in the 14 documentation and/or other materials provided with the distribution. 15 16 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 17 ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 18 LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 19 A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER 20 OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, 21 EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 22 PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR 23 PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF 24 LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING 25 NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 26 SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 27 */ 28 29 #ifdef HAVE_CONFIG_H 30 #include "config.h" 31 #endif 32 33 #include <stdio.h> 34 #include <stdlib.h> 35 #include <math.h> 36 #include <string.h> 37 #include "opus.h" 38 #include "debug.h" 39 #include "opus_types.h" 40 #include "opus_private.h" 41 #include "opus_multistream.h" 42 43 #define MAX_PACKET 1500 44 45 void print_usage( char* argv[] ) 46 { 47 fprintf(stderr, "Usage: %s [-e] <application> <sampling rate (Hz)> <channels (1/2)> " 48 "<bits per second> [options] <input> <output>\n", argv[0]); 49 fprintf(stderr, " %s -d <sampling rate (Hz)> <channels (1/2)> " 50 "[options] <input> <output>\n\n", argv[0]); 51 fprintf(stderr, "application: voip | audio | restricted-lowdelay\n" ); 52 fprintf(stderr, "options:\n" ); 53 fprintf(stderr, "-e : only runs the encoder (output the bit-stream)\n" ); 54 fprintf(stderr, "-d : only runs the decoder (reads the bit-stream as input)\n" ); 55 fprintf(stderr, "-cbr : enable constant bitrate; default: variable bitrate\n" ); 56 fprintf(stderr, "-cvbr : enable constrained variable bitrate; default: unconstrained\n" ); 57 fprintf(stderr, "-delayed-decision : use look-ahead for speech/music detection (experts only); default: disabled\n" ); 58 fprintf(stderr, "-bandwidth <NB|MB|WB|SWB|FB> : audio bandwidth (from narrowband to fullband); default: sampling rate\n" ); 59 fprintf(stderr, "-framesize <2.5|5|10|20|40|60|80|100|120> : frame size in ms; default: 20 \n" ); 60 fprintf(stderr, "-max_payload <bytes> : maximum payload size in bytes, default: 1024\n" ); 61 fprintf(stderr, "-complexity <comp> : complexity, 0 (lowest) ... 10 (highest); default: 10\n" ); 62 fprintf(stderr, "-inbandfec : enable SILK inband FEC\n" ); 63 fprintf(stderr, "-forcemono : force mono encoding, even for stereo input\n" ); 64 fprintf(stderr, "-dtx : enable SILK DTX\n" ); 65 fprintf(stderr, "-loss <perc> : simulate packet loss, in percent (0-100); default: 0\n" ); 66 } 67 68 static void int_to_char(opus_uint32 i, unsigned char ch[4]) 69 { 70 ch[0] = i>>24; 71 ch[1] = (i>>16)&0xFF; 72 ch[2] = (i>>8)&0xFF; 73 ch[3] = i&0xFF; 74 } 75 76 static opus_uint32 char_to_int(unsigned char ch[4]) 77 { 78 return ((opus_uint32)ch[0]<<24) | ((opus_uint32)ch[1]<<16) 79 | ((opus_uint32)ch[2]<< 8) | (opus_uint32)ch[3]; 80 } 81 82 #define check_encoder_option(decode_only, opt) do {if (decode_only) {fprintf(stderr, "option %s is only for encoding\n", opt); goto failure;}} while(0) 83 84 static const int silk8_test[][4] = { 85 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960*3, 1}, 86 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960*2, 1}, 87 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960, 1}, 88 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 480, 1}, 89 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960*3, 2}, 90 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960*2, 2}, 91 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960, 2}, 92 {MODE_SILK_ONLY, OPUS_BANDWIDTH_NARROWBAND, 480, 2} 93 }; 94 95 static const int silk12_test[][4] = { 96 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960*3, 1}, 97 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960*2, 1}, 98 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960, 1}, 99 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 480, 1}, 100 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960*3, 2}, 101 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960*2, 2}, 102 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 960, 2}, 103 {MODE_SILK_ONLY, OPUS_BANDWIDTH_MEDIUMBAND, 480, 2} 104 }; 105 106 static const int silk16_test[][4] = { 107 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960*3, 1}, 108 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960*2, 1}, 109 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960, 1}, 110 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 480, 1}, 111 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960*3, 2}, 112 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960*2, 2}, 113 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960, 2}, 114 {MODE_SILK_ONLY, OPUS_BANDWIDTH_WIDEBAND, 480, 2} 115 }; 116 117 static const int hybrid24_test[][4] = { 118 {MODE_SILK_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 960, 1}, 119 {MODE_SILK_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 480, 1}, 120 {MODE_SILK_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 960, 2}, 121 {MODE_SILK_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 480, 2} 122 }; 123 124 static const int hybrid48_test[][4] = { 125 {MODE_SILK_ONLY, OPUS_BANDWIDTH_FULLBAND, 960, 1}, 126 {MODE_SILK_ONLY, OPUS_BANDWIDTH_FULLBAND, 480, 1}, 127 {MODE_SILK_ONLY, OPUS_BANDWIDTH_FULLBAND, 960, 2}, 128 {MODE_SILK_ONLY, OPUS_BANDWIDTH_FULLBAND, 480, 2} 129 }; 130 131 static const int celt_test[][4] = { 132 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 960, 1}, 133 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 960, 1}, 134 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960, 1}, 135 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960, 1}, 136 137 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 480, 1}, 138 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 480, 1}, 139 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 480, 1}, 140 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 480, 1}, 141 142 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 240, 1}, 143 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 240, 1}, 144 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 240, 1}, 145 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 240, 1}, 146 147 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 120, 1}, 148 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 120, 1}, 149 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 120, 1}, 150 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 120, 1}, 151 152 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 960, 2}, 153 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 960, 2}, 154 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 960, 2}, 155 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 960, 2}, 156 157 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 480, 2}, 158 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 480, 2}, 159 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 480, 2}, 160 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 480, 2}, 161 162 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 240, 2}, 163 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 240, 2}, 164 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 240, 2}, 165 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 240, 2}, 166 167 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 120, 2}, 168 {MODE_CELT_ONLY, OPUS_BANDWIDTH_SUPERWIDEBAND, 120, 2}, 169 {MODE_CELT_ONLY, OPUS_BANDWIDTH_WIDEBAND, 120, 2}, 170 {MODE_CELT_ONLY, OPUS_BANDWIDTH_NARROWBAND, 120, 2}, 171 172 }; 173 174 static const int celt_hq_test[][4] = { 175 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 960, 2}, 176 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 480, 2}, 177 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 240, 2}, 178 {MODE_CELT_ONLY, OPUS_BANDWIDTH_FULLBAND, 120, 2}, 179 }; 180 181 #if 0 /* This is a hack that replaces the normal encoder/decoder with the multistream version */ 182 #define OpusEncoder OpusMSEncoder 183 #define OpusDecoder OpusMSDecoder 184 #define opus_encode opus_multistream_encode 185 #define opus_decode opus_multistream_decode 186 #define opus_encoder_ctl opus_multistream_encoder_ctl 187 #define opus_decoder_ctl opus_multistream_decoder_ctl 188 #define opus_encoder_create ms_opus_encoder_create 189 #define opus_decoder_create ms_opus_decoder_create 190 #define opus_encoder_destroy opus_multistream_encoder_destroy 191 #define opus_decoder_destroy opus_multistream_decoder_destroy 192 193 static OpusEncoder *ms_opus_encoder_create(opus_int32 Fs, int channels, int application, int *error) 194 { 195 int streams, coupled_streams; 196 unsigned char mapping[256]; 197 return (OpusEncoder *)opus_multistream_surround_encoder_create(Fs, channels, 1, &streams, &coupled_streams, mapping, application, error); 198 } 199 static OpusDecoder *ms_opus_decoder_create(opus_int32 Fs, int channels, int *error) 200 { 201 int streams; 202 int coupled_streams; 203 unsigned char mapping[256]={0,1}; 204 streams = 1; 205 coupled_streams = channels==2; 206 return (OpusDecoder *)opus_multistream_decoder_create(Fs, channels, streams, coupled_streams, mapping, error); 207 } 208 #endif 209 210 int main(int argc, char *argv[]) 211 { 212 int err; 213 char *inFile, *outFile; 214 FILE *fin=NULL; 215 FILE *fout=NULL; 216 OpusEncoder *enc=NULL; 217 OpusDecoder *dec=NULL; 218 int args; 219 int len[2]; 220 int frame_size, channels; 221 opus_int32 bitrate_bps=0; 222 unsigned char *data[2] = {NULL, NULL}; 223 unsigned char *fbytes=NULL; 224 opus_int32 sampling_rate; 225 int use_vbr; 226 int max_payload_bytes; 227 int complexity; 228 int use_inbandfec; 229 int use_dtx; 230 int forcechannels; 231 int cvbr = 0; 232 int packet_loss_perc; 233 opus_int32 count=0, count_act=0; 234 int k; 235 opus_int32 skip=0; 236 int stop=0; 237 short *in=NULL; 238 short *out=NULL; 239 int application=OPUS_APPLICATION_AUDIO; 240 double bits=0.0, bits_max=0.0, bits_act=0.0, bits2=0.0, nrg; 241 double tot_samples=0; 242 opus_uint64 tot_in, tot_out; 243 int bandwidth=OPUS_AUTO; 244 const char *bandwidth_string; 245 int lost = 0, lost_prev = 1; 246 int toggle = 0; 247 opus_uint32 enc_final_range[2]; 248 opus_uint32 dec_final_range; 249 int encode_only=0, decode_only=0; 250 int max_frame_size = 48000*2; 251 size_t num_read; 252 int curr_read=0; 253 int sweep_bps = 0; 254 int random_framesize=0, newsize=0, delayed_celt=0; 255 int sweep_max=0, sweep_min=0; 256 int random_fec=0; 257 const int (*mode_list)[4]=NULL; 258 int nb_modes_in_list=0; 259 int curr_mode=0; 260 int curr_mode_count=0; 261 int mode_switch_time = 48000; 262 int nb_encoded=0; 263 int remaining=0; 264 int variable_duration=OPUS_FRAMESIZE_ARG; 265 int delayed_decision=0; 266 int ret = EXIT_FAILURE; 267 268 if (argc < 5 ) 269 { 270 print_usage( argv ); 271 goto failure; 272 } 273 274 tot_in=tot_out=0; 275 fprintf(stderr, "%s\n", opus_get_version_string()); 276 277 args = 1; 278 if (strcmp(argv[args], "-e")==0) 279 { 280 encode_only = 1; 281 args++; 282 } else if (strcmp(argv[args], "-d")==0) 283 { 284 decode_only = 1; 285 args++; 286 } 287 if (!decode_only && argc < 7 ) 288 { 289 print_usage( argv ); 290 goto failure; 291 } 292 293 if (!decode_only) 294 { 295 if (strcmp(argv[args], "voip")==0) 296 application = OPUS_APPLICATION_VOIP; 297 else if (strcmp(argv[args], "restricted-lowdelay")==0) 298 application = OPUS_APPLICATION_RESTRICTED_LOWDELAY; 299 else if (strcmp(argv[args], "audio")!=0) { 300 fprintf(stderr, "unknown application: %s\n", argv[args]); 301 print_usage(argv); 302 goto failure; 303 } 304 args++; 305 } 306 sampling_rate = (opus_int32)atol(argv[args]); 307 args++; 308 309 if (sampling_rate != 8000 && sampling_rate != 12000 310 && sampling_rate != 16000 && sampling_rate != 24000 311 && sampling_rate != 48000) 312 { 313 fprintf(stderr, "Supported sampling rates are 8000, 12000, " 314 "16000, 24000 and 48000.\n"); 315 goto failure; 316 } 317 frame_size = sampling_rate/50; 318 319 channels = atoi(argv[args]); 320 args++; 321 322 if (channels < 1 || channels > 2) 323 { 324 fprintf(stderr, "Opus_demo supports only 1 or 2 channels.\n"); 325 goto failure; 326 } 327 328 if (!decode_only) 329 { 330 bitrate_bps = (opus_int32)atol(argv[args]); 331 args++; 332 } 333 334 /* defaults: */ 335 use_vbr = 1; 336 max_payload_bytes = MAX_PACKET; 337 complexity = 10; 338 use_inbandfec = 0; 339 forcechannels = OPUS_AUTO; 340 use_dtx = 0; 341 packet_loss_perc = 0; 342 343 while( args < argc - 2 ) { 344 /* process command line options */ 345 if( strcmp( argv[ args ], "-cbr" ) == 0 ) { 346 check_encoder_option(decode_only, "-cbr"); 347 use_vbr = 0; 348 args++; 349 } else if( strcmp( argv[ args ], "-bandwidth" ) == 0 ) { 350 check_encoder_option(decode_only, "-bandwidth"); 351 if (strcmp(argv[ args + 1 ], "NB")==0) 352 bandwidth = OPUS_BANDWIDTH_NARROWBAND; 353 else if (strcmp(argv[ args + 1 ], "MB")==0) 354 bandwidth = OPUS_BANDWIDTH_MEDIUMBAND; 355 else if (strcmp(argv[ args + 1 ], "WB")==0) 356 bandwidth = OPUS_BANDWIDTH_WIDEBAND; 357 else if (strcmp(argv[ args + 1 ], "SWB")==0) 358 bandwidth = OPUS_BANDWIDTH_SUPERWIDEBAND; 359 else if (strcmp(argv[ args + 1 ], "FB")==0) 360 bandwidth = OPUS_BANDWIDTH_FULLBAND; 361 else { 362 fprintf(stderr, "Unknown bandwidth %s. " 363 "Supported are NB, MB, WB, SWB, FB.\n", 364 argv[ args + 1 ]); 365 goto failure; 366 } 367 args += 2; 368 } else if( strcmp( argv[ args ], "-framesize" ) == 0 ) { 369 check_encoder_option(decode_only, "-framesize"); 370 if (strcmp(argv[ args + 1 ], "2.5")==0) 371 frame_size = sampling_rate/400; 372 else if (strcmp(argv[ args + 1 ], "5")==0) 373 frame_size = sampling_rate/200; 374 else if (strcmp(argv[ args + 1 ], "10")==0) 375 frame_size = sampling_rate/100; 376 else if (strcmp(argv[ args + 1 ], "20")==0) 377 frame_size = sampling_rate/50; 378 else if (strcmp(argv[ args + 1 ], "40")==0) 379 frame_size = sampling_rate/25; 380 else if (strcmp(argv[ args + 1 ], "60")==0) 381 frame_size = 3*sampling_rate/50; 382 else if (strcmp(argv[ args + 1 ], "80")==0) 383 frame_size = 4*sampling_rate/50; 384 else if (strcmp(argv[ args + 1 ], "100")==0) 385 frame_size = 5*sampling_rate/50; 386 else if (strcmp(argv[ args + 1 ], "120")==0) 387 frame_size = 6*sampling_rate/50; 388 else { 389 fprintf(stderr, "Unsupported frame size: %s ms. " 390 "Supported are 2.5, 5, 10, 20, 40, 60, 80, 100, 120.\n", 391 argv[ args + 1 ]); 392 goto failure; 393 } 394 args += 2; 395 } else if( strcmp( argv[ args ], "-max_payload" ) == 0 ) { 396 check_encoder_option(decode_only, "-max_payload"); 397 max_payload_bytes = atoi( argv[ args + 1 ] ); 398 args += 2; 399 } else if( strcmp( argv[ args ], "-complexity" ) == 0 ) { 400 check_encoder_option(decode_only, "-complexity"); 401 complexity = atoi( argv[ args + 1 ] ); 402 args += 2; 403 } else if( strcmp( argv[ args ], "-inbandfec" ) == 0 ) { 404 use_inbandfec = 1; 405 args++; 406 } else if( strcmp( argv[ args ], "-forcemono" ) == 0 ) { 407 check_encoder_option(decode_only, "-forcemono"); 408 forcechannels = 1; 409 args++; 410 } else if( strcmp( argv[ args ], "-cvbr" ) == 0 ) { 411 check_encoder_option(decode_only, "-cvbr"); 412 cvbr = 1; 413 args++; 414 } else if( strcmp( argv[ args ], "-delayed-decision" ) == 0 ) { 415 check_encoder_option(decode_only, "-delayed-decision"); 416 delayed_decision = 1; 417 args++; 418 } else if( strcmp( argv[ args ], "-dtx") == 0 ) { 419 check_encoder_option(decode_only, "-dtx"); 420 use_dtx = 1; 421 args++; 422 } else if( strcmp( argv[ args ], "-loss" ) == 0 ) { 423 packet_loss_perc = atoi( argv[ args + 1 ] ); 424 args += 2; 425 } else if( strcmp( argv[ args ], "-sweep" ) == 0 ) { 426 check_encoder_option(decode_only, "-sweep"); 427 sweep_bps = atoi( argv[ args + 1 ] ); 428 args += 2; 429 } else if( strcmp( argv[ args ], "-random_framesize" ) == 0 ) { 430 check_encoder_option(decode_only, "-random_framesize"); 431 random_framesize = 1; 432 args++; 433 } else if( strcmp( argv[ args ], "-sweep_max" ) == 0 ) { 434 check_encoder_option(decode_only, "-sweep_max"); 435 sweep_max = atoi( argv[ args + 1 ] ); 436 args += 2; 437 } else if( strcmp( argv[ args ], "-random_fec" ) == 0 ) { 438 check_encoder_option(decode_only, "-random_fec"); 439 random_fec = 1; 440 args++; 441 } else if( strcmp( argv[ args ], "-silk8k_test" ) == 0 ) { 442 check_encoder_option(decode_only, "-silk8k_test"); 443 mode_list = silk8_test; 444 nb_modes_in_list = 8; 445 args++; 446 } else if( strcmp( argv[ args ], "-silk12k_test" ) == 0 ) { 447 check_encoder_option(decode_only, "-silk12k_test"); 448 mode_list = silk12_test; 449 nb_modes_in_list = 8; 450 args++; 451 } else if( strcmp( argv[ args ], "-silk16k_test" ) == 0 ) { 452 check_encoder_option(decode_only, "-silk16k_test"); 453 mode_list = silk16_test; 454 nb_modes_in_list = 8; 455 args++; 456 } else if( strcmp( argv[ args ], "-hybrid24k_test" ) == 0 ) { 457 check_encoder_option(decode_only, "-hybrid24k_test"); 458 mode_list = hybrid24_test; 459 nb_modes_in_list = 4; 460 args++; 461 } else if( strcmp( argv[ args ], "-hybrid48k_test" ) == 0 ) { 462 check_encoder_option(decode_only, "-hybrid48k_test"); 463 mode_list = hybrid48_test; 464 nb_modes_in_list = 4; 465 args++; 466 } else if( strcmp( argv[ args ], "-celt_test" ) == 0 ) { 467 check_encoder_option(decode_only, "-celt_test"); 468 mode_list = celt_test; 469 nb_modes_in_list = 32; 470 args++; 471 } else if( strcmp( argv[ args ], "-celt_hq_test" ) == 0 ) { 472 check_encoder_option(decode_only, "-celt_hq_test"); 473 mode_list = celt_hq_test; 474 nb_modes_in_list = 4; 475 args++; 476 } else { 477 printf( "Error: unrecognized setting: %s\n\n", argv[ args ] ); 478 print_usage( argv ); 479 goto failure; 480 } 481 } 482 483 if (sweep_max) 484 sweep_min = bitrate_bps; 485 486 if (max_payload_bytes < 0 || max_payload_bytes > MAX_PACKET) 487 { 488 fprintf (stderr, "max_payload_bytes must be between 0 and %d\n", 489 MAX_PACKET); 490 goto failure; 491 } 492 493 inFile = argv[argc-2]; 494 fin = fopen(inFile, "rb"); 495 if (!fin) 496 { 497 fprintf (stderr, "Could not open input file %s\n", argv[argc-2]); 498 goto failure; 499 } 500 if (mode_list) 501 { 502 int size; 503 fseek(fin, 0, SEEK_END); 504 size = ftell(fin); 505 fprintf(stderr, "File size is %d bytes\n", size); 506 fseek(fin, 0, SEEK_SET); 507 mode_switch_time = size/sizeof(short)/channels/nb_modes_in_list; 508 fprintf(stderr, "Switching mode every %d samples\n", mode_switch_time); 509 } 510 511 outFile = argv[argc-1]; 512 fout = fopen(outFile, "wb+"); 513 if (!fout) 514 { 515 fprintf (stderr, "Could not open output file %s\n", argv[argc-1]); 516 goto failure; 517 } 518 519 if (!decode_only) 520 { 521 enc = opus_encoder_create(sampling_rate, channels, application, &err); 522 if (err != OPUS_OK) 523 { 524 fprintf(stderr, "Cannot create encoder: %s\n", opus_strerror(err)); 525 goto failure; 526 } 527 opus_encoder_ctl(enc, OPUS_SET_BITRATE(bitrate_bps)); 528 opus_encoder_ctl(enc, OPUS_SET_BANDWIDTH(bandwidth)); 529 opus_encoder_ctl(enc, OPUS_SET_VBR(use_vbr)); 530 opus_encoder_ctl(enc, OPUS_SET_VBR_CONSTRAINT(cvbr)); 531 opus_encoder_ctl(enc, OPUS_SET_COMPLEXITY(complexity)); 532 opus_encoder_ctl(enc, OPUS_SET_INBAND_FEC(use_inbandfec)); 533 opus_encoder_ctl(enc, OPUS_SET_FORCE_CHANNELS(forcechannels)); 534 opus_encoder_ctl(enc, OPUS_SET_DTX(use_dtx)); 535 opus_encoder_ctl(enc, OPUS_SET_PACKET_LOSS_PERC(packet_loss_perc)); 536 537 opus_encoder_ctl(enc, OPUS_GET_LOOKAHEAD(&skip)); 538 opus_encoder_ctl(enc, OPUS_SET_LSB_DEPTH(16)); 539 opus_encoder_ctl(enc, OPUS_SET_EXPERT_FRAME_DURATION(variable_duration)); 540 } 541 if (!encode_only) 542 { 543 dec = opus_decoder_create(sampling_rate, channels, &err); 544 if (err != OPUS_OK) 545 { 546 fprintf(stderr, "Cannot create decoder: %s\n", opus_strerror(err)); 547 goto failure; 548 } 549 } 550 551 552 switch(bandwidth) 553 { 554 case OPUS_BANDWIDTH_NARROWBAND: 555 bandwidth_string = "narrowband"; 556 break; 557 case OPUS_BANDWIDTH_MEDIUMBAND: 558 bandwidth_string = "mediumband"; 559 break; 560 case OPUS_BANDWIDTH_WIDEBAND: 561 bandwidth_string = "wideband"; 562 break; 563 case OPUS_BANDWIDTH_SUPERWIDEBAND: 564 bandwidth_string = "superwideband"; 565 break; 566 case OPUS_BANDWIDTH_FULLBAND: 567 bandwidth_string = "fullband"; 568 break; 569 case OPUS_AUTO: 570 bandwidth_string = "auto bandwidth"; 571 break; 572 default: 573 bandwidth_string = "unknown"; 574 break; 575 } 576 577 if (decode_only) 578 fprintf(stderr, "Decoding with %ld Hz output (%d channels)\n", 579 (long)sampling_rate, channels); 580 else 581 fprintf(stderr, "Encoding %ld Hz input at %.3f kb/s " 582 "in %s with %d-sample frames.\n", 583 (long)sampling_rate, bitrate_bps*0.001, 584 bandwidth_string, frame_size); 585 586 in = (short*)malloc(max_frame_size*channels*sizeof(short)); 587 out = (short*)malloc(max_frame_size*channels*sizeof(short)); 588 /* We need to allocate for 16-bit PCM data, but we store it as unsigned char. */ 589 fbytes = (unsigned char*)malloc(max_frame_size*channels*sizeof(short)); 590 data[0] = (unsigned char*)calloc(max_payload_bytes,sizeof(unsigned char)); 591 if ( use_inbandfec ) { 592 data[1] = (unsigned char*)calloc(max_payload_bytes,sizeof(unsigned char)); 593 } 594 if(delayed_decision) 595 { 596 if (frame_size==sampling_rate/400) 597 variable_duration = OPUS_FRAMESIZE_2_5_MS; 598 else if (frame_size==sampling_rate/200) 599 variable_duration = OPUS_FRAMESIZE_5_MS; 600 else if (frame_size==sampling_rate/100) 601 variable_duration = OPUS_FRAMESIZE_10_MS; 602 else if (frame_size==sampling_rate/50) 603 variable_duration = OPUS_FRAMESIZE_20_MS; 604 else if (frame_size==sampling_rate/25) 605 variable_duration = OPUS_FRAMESIZE_40_MS; 606 else if (frame_size==3*sampling_rate/50) 607 variable_duration = OPUS_FRAMESIZE_60_MS; 608 else if (frame_size==4*sampling_rate/50) 609 variable_duration = OPUS_FRAMESIZE_80_MS; 610 else if (frame_size==5*sampling_rate/50) 611 variable_duration = OPUS_FRAMESIZE_100_MS; 612 else 613 variable_duration = OPUS_FRAMESIZE_120_MS; 614 opus_encoder_ctl(enc, OPUS_SET_EXPERT_FRAME_DURATION(variable_duration)); 615 frame_size = 2*48000; 616 } 617 while (!stop) 618 { 619 if (delayed_celt) 620 { 621 frame_size = newsize; 622 delayed_celt = 0; 623 } else if (random_framesize && rand()%20==0) 624 { 625 newsize = rand()%6; 626 switch(newsize) 627 { 628 case 0: newsize=sampling_rate/400; break; 629 case 1: newsize=sampling_rate/200; break; 630 case 2: newsize=sampling_rate/100; break; 631 case 3: newsize=sampling_rate/50; break; 632 case 4: newsize=sampling_rate/25; break; 633 case 5: newsize=3*sampling_rate/50; break; 634 } 635 while (newsize < sampling_rate/25 && bitrate_bps-abs(sweep_bps) <= 3*12*sampling_rate/newsize) 636 newsize*=2; 637 if (newsize < sampling_rate/100 && frame_size >= sampling_rate/100) 638 { 639 opus_encoder_ctl(enc, OPUS_SET_FORCE_MODE(MODE_CELT_ONLY)); 640 delayed_celt=1; 641 } else { 642 frame_size = newsize; 643 } 644 } 645 if (random_fec && rand()%30==0) 646 { 647 opus_encoder_ctl(enc, OPUS_SET_INBAND_FEC(rand()%4==0)); 648 } 649 if (decode_only) 650 { 651 unsigned char ch[4]; 652 num_read = fread(ch, 1, 4, fin); 653 if (num_read!=4) 654 break; 655 len[toggle] = char_to_int(ch); 656 if (len[toggle]>max_payload_bytes || len[toggle]<0) 657 { 658 fprintf(stderr, "Invalid payload length: %d\n",len[toggle]); 659 break; 660 } 661 num_read = fread(ch, 1, 4, fin); 662 if (num_read!=4) 663 break; 664 enc_final_range[toggle] = char_to_int(ch); 665 num_read = fread(data[toggle], 1, len[toggle], fin); 666 if (num_read!=(size_t)len[toggle]) 667 { 668 fprintf(stderr, "Ran out of input, " 669 "expecting %d bytes got %d\n", 670 len[toggle],(int)num_read); 671 break; 672 } 673 } else { 674 int i; 675 if (mode_list!=NULL) 676 { 677 opus_encoder_ctl(enc, OPUS_SET_BANDWIDTH(mode_list[curr_mode][1])); 678 opus_encoder_ctl(enc, OPUS_SET_FORCE_MODE(mode_list[curr_mode][0])); 679 opus_encoder_ctl(enc, OPUS_SET_FORCE_CHANNELS(mode_list[curr_mode][3])); 680 frame_size = mode_list[curr_mode][2]; 681 } 682 num_read = fread(fbytes, sizeof(short)*channels, frame_size-remaining, fin); 683 curr_read = (int)num_read; 684 tot_in += curr_read; 685 for(i=0;i<curr_read*channels;i++) 686 { 687 opus_int32 s; 688 s=fbytes[2*i+1]<<8|fbytes[2*i]; 689 s=((s&0xFFFF)^0x8000)-0x8000; 690 in[i+remaining*channels]=s; 691 } 692 if (curr_read+remaining < frame_size) 693 { 694 for (i=(curr_read+remaining)*channels;i<frame_size*channels;i++) 695 in[i] = 0; 696 if (encode_only || decode_only) 697 stop = 1; 698 } 699 len[toggle] = opus_encode(enc, in, frame_size, data[toggle], max_payload_bytes); 700 nb_encoded = opus_packet_get_samples_per_frame(data[toggle], sampling_rate)*opus_packet_get_nb_frames(data[toggle], len[toggle]); 701 remaining = frame_size-nb_encoded; 702 for(i=0;i<remaining*channels;i++) 703 in[i] = in[nb_encoded*channels+i]; 704 if (sweep_bps!=0) 705 { 706 bitrate_bps += sweep_bps; 707 if (sweep_max) 708 { 709 if (bitrate_bps > sweep_max) 710 sweep_bps = -sweep_bps; 711 else if (bitrate_bps < sweep_min) 712 sweep_bps = -sweep_bps; 713 } 714 /* safety */ 715 if (bitrate_bps<1000) 716 bitrate_bps = 1000; 717 opus_encoder_ctl(enc, OPUS_SET_BITRATE(bitrate_bps)); 718 } 719 opus_encoder_ctl(enc, OPUS_GET_FINAL_RANGE(&enc_final_range[toggle])); 720 if (len[toggle] < 0) 721 { 722 fprintf (stderr, "opus_encode() returned %d\n", len[toggle]); 723 goto failure; 724 } 725 curr_mode_count += frame_size; 726 if (curr_mode_count > mode_switch_time && curr_mode < nb_modes_in_list-1) 727 { 728 curr_mode++; 729 curr_mode_count = 0; 730 } 731 } 732 733 #if 0 /* This is for testing the padding code, do not enable by default */ 734 if (len[toggle]<1275) 735 { 736 int new_len = len[toggle]+rand()%(max_payload_bytes-len[toggle]); 737 if ((err = opus_packet_pad(data[toggle], len[toggle], new_len)) != OPUS_OK) 738 { 739 fprintf(stderr, "padding failed: %s\n", opus_strerror(err)); 740 goto failure; 741 } 742 len[toggle] = new_len; 743 } 744 #endif 745 if (encode_only) 746 { 747 unsigned char int_field[4]; 748 int_to_char(len[toggle], int_field); 749 if (fwrite(int_field, 1, 4, fout) != 4) { 750 fprintf(stderr, "Error writing.\n"); 751 goto failure; 752 } 753 int_to_char(enc_final_range[toggle], int_field); 754 if (fwrite(int_field, 1, 4, fout) != 4) { 755 fprintf(stderr, "Error writing.\n"); 756 goto failure; 757 } 758 if (fwrite(data[toggle], 1, len[toggle], fout) != (unsigned)len[toggle]) { 759 fprintf(stderr, "Error writing.\n"); 760 goto failure; 761 } 762 tot_samples += nb_encoded; 763 } else { 764 opus_int32 output_samples; 765 lost = len[toggle]==0 || (packet_loss_perc>0 && rand()%100 < packet_loss_perc); 766 if (lost) 767 opus_decoder_ctl(dec, OPUS_GET_LAST_PACKET_DURATION(&output_samples)); 768 else 769 output_samples = max_frame_size; 770 if( count >= use_inbandfec ) { 771 /* delay by one packet when using in-band FEC */ 772 if( use_inbandfec ) { 773 if( lost_prev ) { 774 /* attempt to decode with in-band FEC from next packet */ 775 opus_decoder_ctl(dec, OPUS_GET_LAST_PACKET_DURATION(&output_samples)); 776 output_samples = opus_decode(dec, lost ? NULL : data[toggle], len[toggle], out, output_samples, 1); 777 } else { 778 /* regular decode */ 779 output_samples = max_frame_size; 780 output_samples = opus_decode(dec, data[1-toggle], len[1-toggle], out, output_samples, 0); 781 } 782 } else { 783 output_samples = opus_decode(dec, lost ? NULL : data[toggle], len[toggle], out, output_samples, 0); 784 } 785 if (output_samples>0) 786 { 787 if (!decode_only && tot_out + output_samples > tot_in) 788 { 789 stop=1; 790 output_samples = (opus_int32)(tot_in - tot_out); 791 } 792 if (output_samples>skip) { 793 int i; 794 for(i=0;i<(output_samples-skip)*channels;i++) 795 { 796 short s; 797 s=out[i+(skip*channels)]; 798 fbytes[2*i]=s&0xFF; 799 fbytes[2*i+1]=(s>>8)&0xFF; 800 } 801 if (fwrite(fbytes, sizeof(short)*channels, output_samples-skip, fout) != (unsigned)(output_samples-skip)){ 802 fprintf(stderr, "Error writing.\n"); 803 goto failure; 804 } 805 tot_out += output_samples-skip; 806 } 807 if (output_samples<skip) skip -= output_samples; 808 else skip = 0; 809 } else { 810 fprintf(stderr, "error decoding frame: %s\n", 811 opus_strerror(output_samples)); 812 } 813 tot_samples += output_samples; 814 } 815 } 816 817 if (!encode_only) 818 opus_decoder_ctl(dec, OPUS_GET_FINAL_RANGE(&dec_final_range)); 819 /* compare final range encoder rng values of encoder and decoder */ 820 if( enc_final_range[toggle^use_inbandfec]!=0 && !encode_only 821 && !lost && !lost_prev 822 && dec_final_range != enc_final_range[toggle^use_inbandfec] ) { 823 fprintf (stderr, "Error: Range coder state mismatch " 824 "between encoder and decoder " 825 "in frame %ld: 0x%8lx vs 0x%8lx\n", 826 (long)count, 827 (unsigned long)enc_final_range[toggle^use_inbandfec], 828 (unsigned long)dec_final_range); 829 goto failure; 830 } 831 832 lost_prev = lost; 833 if( count >= use_inbandfec ) { 834 /* count bits */ 835 bits += len[toggle]*8; 836 bits_max = ( len[toggle]*8 > bits_max ) ? len[toggle]*8 : bits_max; 837 bits2 += len[toggle]*len[toggle]*64; 838 if (!decode_only) 839 { 840 nrg = 0.0; 841 for ( k = 0; k < frame_size * channels; k++ ) { 842 nrg += in[ k ] * (double)in[ k ]; 843 } 844 nrg /= frame_size * channels; 845 if( nrg > 1e5 ) { 846 bits_act += len[toggle]*8; 847 count_act++; 848 } 849 } 850 } 851 count++; 852 toggle = (toggle + use_inbandfec) & 1; 853 } 854 855 if(decode_only && count > 0) 856 frame_size = (int)(tot_samples / count); 857 count -= use_inbandfec; 858 if (tot_samples >= 1 && count > 0 && frame_size) 859 { 860 /* Print out bitrate statistics */ 861 double var; 862 fprintf (stderr, "average bitrate: %7.3f kb/s\n", 863 1e-3*bits*sampling_rate/tot_samples); 864 fprintf (stderr, "maximum bitrate: %7.3f kb/s\n", 865 1e-3*bits_max*sampling_rate/frame_size); 866 if (!decode_only) 867 fprintf (stderr, "active bitrate: %7.3f kb/s\n", 868 1e-3*bits_act*sampling_rate/(1e-15+frame_size*(double)count_act)); 869 var = bits2/count - bits*bits/(count*(double)count); 870 if (var < 0) 871 var = 0; 872 fprintf (stderr, "bitrate standard deviation: %7.3f kb/s\n", 873 1e-3*sqrt(var)*sampling_rate/frame_size); 874 } else { 875 fprintf(stderr, "bitrate statistics are undefined\n"); 876 } 877 silk_TimerSave("opus_timing.txt"); 878 ret = EXIT_SUCCESS; 879 failure: 880 opus_encoder_destroy(enc); 881 opus_decoder_destroy(dec); 882 free(data[0]); 883 free(data[1]); 884 if (fin) 885 fclose(fin); 886 if (fout) 887 fclose(fout); 888 free(in); 889 free(out); 890 free(fbytes); 891 return ret; 892 } 893
