Home | History | Annotate | Download | only in doc
      1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
      2 <html>
      3 <head>
      4 
      5 <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15"/>
      6 <title>Ogg Vorbis Documentation</title>
      7 
      8 <style type="text/css">
      9 body {
     10   margin: 0 18px 0 18px;
     11   padding-bottom: 30px;
     12   font-family: Verdana, Arial, Helvetica, sans-serif;
     13   color: #333333;
     14   font-size: .8em;
     15 }
     16 
     17 a {
     18   color: #3366cc;
     19 }
     20 
     21 img {
     22   border: 0;
     23 }
     24 
     25 #xiphlogo {
     26   margin: 30px 0 16px 0;
     27 }
     28 
     29 #content p {
     30   line-height: 1.4;
     31 }
     32 
     33 h1, h1 a, h2, h2 a, h3, h3 a {
     34   font-weight: bold;
     35   color: #ff9900;
     36   margin: 1.3em 0 8px 0;
     37 }
     38 
     39 h1 {
     40   font-size: 1.3em;
     41 }
     42 
     43 h2 {
     44   font-size: 1.2em;
     45 }
     46 
     47 h3 {
     48   font-size: 1.1em;
     49 }
     50 
     51 li {
     52   line-height: 1.4;
     53 }
     54 
     55 #copyright {
     56   margin-top: 30px;
     57   line-height: 1.5em;
     58   text-align: center;
     59   font-size: .8em;
     60   color: #888888;
     61   clear: both;
     62 }
     63 </style>
     64 
     65 </head>
     66 
     67 <body>
     68 
     69 <div id="xiphlogo">
     70   <a href="http://www.xiph.org/"><img src="fish_xiph_org.png" alt="Fish Logo and Xiph.org"/></a>
     71 </div>
     72 
     73 <h1>Ogg Vorbis encoding format documentation</h1>
     74 
     75 <p><img src="wait.png" alt="wait"/>As of writing, not all the below document
     76 links are live. They will be populated as we complete the documents.</p>
     77 
     78 <h2>Documents</h2>
     79 
     80 <ul>
     81 <li><a href="packet.html">Vorbis packet structure</a></li>
     82 <li><a href="envelope.html">Temporal envelope shaping and blocksize</a></li>
     83 <li><a href="mdct.html">Time domain segmentation and MDCT transform</a></li>
     84 <li><a href="resolution.html">The resolution floor</a></li>
     85 <li><a href="residuals.html">MDCT-domain fine structure</a></li>
     86 </ul>
     87 
     88 <ul>
     89 <li><a href="probmodel.html">The Vorbis probability model</a></li>
     90 <li><a href="bitpack.html">The Vorbis bitpacker</a></li>
     91 </ul>
     92 
     93 <ul>
     94 <li><a href="oggstream.html">Ogg bitstream overview</a></li>
     95 <li><a href="framing.html">Ogg logical bitstream and framing spec</a></li>
     96 <li><a href="vorbis-stream.html">Vorbis packet->Ogg bitstream mapping</a></li>
     97 </ul>
     98 
     99 <ul>
    100 <li><a href="programming.html">Programming with libvorbis</a></li>
    101 </ul>
    102 
    103 <h2>Description</h2>
    104 
    105 <p>Ogg Vorbis is a general purpose compressed audio format
    106 for high quality (44.1-48.0kHz, 16+ bit, polyphonic) audio and music
    107 at moderate fixed and variable bitrates (40-80 kb/s/channel). This
    108 places Vorbis in the same class as audio representations including
    109 MPEG-1 audio layer 3, MPEG-4 audio (AAC and TwinVQ), and PAC.</p>
    110 
    111 <p>Vorbis is the first of a planned family of Ogg multimedia coding
    112 formats being developed as part of the Xiph.org Foundation's Ogg multimedia
    113 project. See <a href="http://www.xiph.org/">http://www.xiph.org/</a>
    114 for more information.</p>
    115 
    116 <h2>Vorbis technical documents</h2>
    117 
    118 <p>A Vorbis encoder takes in overlapping (but contiguous) short-time
    119 segments of audio data. The encoder analyzes the content of the audio
    120 to determine an optimal compact representation; this phase of encoding
    121 is known as <em>analysis</em>. For each short-time block of sound,
    122 the encoder then packs an efficient representation of the signal, as
    123 determined by analysis, into a raw packet much smaller than the size
    124 required by the original signal; this phase is <em>coding</em>.
    125 Lastly, in a streaming environment, the raw packets are then
    126 structured into a continuous stream of octets; this last phase is
    127 <em>streaming</em>. Note that the stream of octets is referred to both
    128 as a 'byte-' and 'bit-'stream; the latter usage is acceptible as the
    129 stream of octets is a physical representation of a true logical
    130 bit-by-bit stream.</p>
    131 
    132 <p>A Vorbis decoder performs a mirror image process of extracting the
    133 original sequence of raw packets from an Ogg stream (<em>stream
    134 decomposition</em>), reconstructing the signal representation from the
    135 raw data in the packet (<em>decoding</em>) and them reconstituting an
    136 audio signal from the decoded representation (<em>synthesis</em>).</p>
    137 
    138 <p>The <a href="programming.html">Programming with libvorbis</a>
    139 documents discuss use of the reference Vorbis codec library
    140 (libvorbis) produced by the Xiph.org Foundation.</p>
    141 
    142 <p>The data representations and algorithms necessary at each step to
    143 encode and decode Ogg Vorbis bitstreams are described by the below
    144 documents in sufficient detail to construct a complete Vorbis codec.
    145 Note that at the time of writing, Vorbis is still in a 'Request For
    146 Comments' stage of development; despite being in advanced stages of
    147 development, input from the multimedia community is welcome.</p>
    148 
    149 <h3>Vorbis analysis and synthesis</h3>
    150 
    151 <p>Analysis begins by seperating an input audio stream into individual,
    152 overlapping short-time segments of audio data. These segments are
    153 then transformed into an alternate representation, seeking to
    154 represent the original signal in a more efficient form that codes into
    155 a smaller number of bytes. The analysis and transformation stage is
    156 the most complex element of producing a Vorbis bitstream.</p>
    157 
    158 <p>The corresponding synthesis step in the decoder is simpler; there is
    159 no analysis to perform, merely a mechanical, deterministic
    160 reconstruction of the original audio data from the transform-domain
    161 representation.</p>
    162 
    163 <ul>
    164 <li><a href="packet.html">Vorbis packet structure</a>:
    165 Describes the basic analysis components necessary to produce Vorbis
    166 packets and the structure of the packet itself.</li>
    167 <li><a href="envelope.html">Temporal envelope shaping and blocksize</a>:
    168 Use of temporal envelope shaping and variable blocksize to minimize
    169 time-domain energy leakage during wide dynamic range and spectral energy
    170 swings. Also discusses time-related principles of psychoacoustics.</li>
    171 <li><a href="mdct.html">Time domain segmentation and MDCT transform</a>:
    172 Division of time domain data into individual overlapped, windowed
    173 short-time vectors and transformation using the MDCT</li>
    174 <li><a href="resolution.html">The resolution floor</a>: Use of frequency
    175 doamin psychoacoustics, and the MDCT-domain noise, masking and resolution
    176 floors</li>
    177 <li><a href="residuals.html">MDCT-domain fine structure</a>: Production,
    178 quantization and massaging of MDCT-spectrum fine structure</li>
    179 </ul>
    180 
    181 <h3>Vorbis coding and decoding</h3>
    182 
    183 <p>Coding and decoding converts the transform-domain representation of
    184 the original audio produced by analysis to and from a bitwise packed
    185 raw data packet. Coding and decoding consist of two logically
    186 orthogonal concepts, <em>back-end coding</em> and <em>bitpacking</em>.</p>
    187 
    188 <p><em>Back-end coding</em> uses a probability model to represent the raw numbers
    189 of the audio representation in as few physical bits as possible;
    190 familiar examples of back-end coding include Huffman coding and Vector
    191 Quantization.</p>
    192 
    193 <p><em>Bitpacking</em> arranges the variable sized words of the back-end
    194 coding into a vector of octets without wasting space. The octets
    195 produced by coding a single short-time audio segment is one raw Vorbis
    196 packet.</p>
    197 
    198 <ul>
    199 <li><a href="probmodel.html">The Vorbis probability model</a></li>
    200 <li><a href="bitpack.html">The Vorbis bitpacker</a>: Arrangement of 
    201 variable bit-length words into an octet-aligned packet.</li>
    202 </ul>
    203 
    204 <h3>Vorbis streaming and stream decomposition</h3>
    205 
    206 <p>Vorbis packets contain the raw, bitwise-compressed representation of a
    207 snippet of audio. These packets contain no structure and cannot be
    208 strung together directly into a stream; for streamed transmission and
    209 storage, Vorbis packets are encoded into an Ogg bitstream.</p>
    210 
    211 <ul>
    212 <li><a href="oggstream.html">Ogg bitstream overview</a>: High-level
    213 description of Ogg logical bitstreams, how logical bitstreams
    214 (of mixed media types) can be combined into physical bitstreams, and
    215 restrictions on logical-to-physical mapping. Note that this document is
    216 not specific only to Ogg Vorbis.</li>
    217 <li><a href="framing.html">Ogg logical bitstream and framing
    218 spec</a>: Low level, complete specification of Ogg logical
    219 bitstream pages. Note that this document is not specific only to Ogg
    220 Vorbis.</li>
    221 <li><a href="vorbis-stream.html">Vorbis bitstream mapping</a>:
    222 Specifically describes mapping Vorbis data into an
    223 Ogg physical bitstream.</li>
    224 </ul>
    225 
    226 <div id="copyright">
    227   The Xiph Fish Logo is a
    228   trademark (&trade;) of Xiph.Org.<br/>
    229 
    230   These pages &copy; 1994 - 2005 Xiph.Org. All rights reserved.
    231 </div>
    232 
    233 </body>
    234 </html>
    235