|
Chris@1
|
1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
|
|
Chris@1
|
2 <html>
|
|
Chris@1
|
3 <head>
|
|
Chris@1
|
4
|
|
Chris@1
|
5 <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15"/>
|
|
Chris@1
|
6 <title>Ogg Vorbis Documentation</title>
|
|
Chris@1
|
7
|
|
Chris@1
|
8 <style type="text/css">
|
|
Chris@1
|
9 body {
|
|
Chris@1
|
10 margin: 0 18px 0 18px;
|
|
Chris@1
|
11 padding-bottom: 30px;
|
|
Chris@1
|
12 font-family: Verdana, Arial, Helvetica, sans-serif;
|
|
Chris@1
|
13 color: #333333;
|
|
Chris@1
|
14 font-size: .8em;
|
|
Chris@1
|
15 }
|
|
Chris@1
|
16
|
|
Chris@1
|
17 a {
|
|
Chris@1
|
18 color: #3366cc;
|
|
Chris@1
|
19 }
|
|
Chris@1
|
20
|
|
Chris@1
|
21 img {
|
|
Chris@1
|
22 border: 0;
|
|
Chris@1
|
23 }
|
|
Chris@1
|
24
|
|
Chris@1
|
25 #xiphlogo {
|
|
Chris@1
|
26 margin: 30px 0 16px 0;
|
|
Chris@1
|
27 }
|
|
Chris@1
|
28
|
|
Chris@1
|
29 #content p {
|
|
Chris@1
|
30 line-height: 1.4;
|
|
Chris@1
|
31 }
|
|
Chris@1
|
32
|
|
Chris@1
|
33 h1, h1 a, h2, h2 a, h3, h3 a {
|
|
Chris@1
|
34 font-weight: bold;
|
|
Chris@1
|
35 color: #ff9900;
|
|
Chris@1
|
36 margin: 1.3em 0 8px 0;
|
|
Chris@1
|
37 }
|
|
Chris@1
|
38
|
|
Chris@1
|
39 h1 {
|
|
Chris@1
|
40 font-size: 1.3em;
|
|
Chris@1
|
41 }
|
|
Chris@1
|
42
|
|
Chris@1
|
43 h2 {
|
|
Chris@1
|
44 font-size: 1.2em;
|
|
Chris@1
|
45 }
|
|
Chris@1
|
46
|
|
Chris@1
|
47 h3 {
|
|
Chris@1
|
48 font-size: 1.1em;
|
|
Chris@1
|
49 }
|
|
Chris@1
|
50
|
|
Chris@1
|
51 li {
|
|
Chris@1
|
52 line-height: 1.4;
|
|
Chris@1
|
53 }
|
|
Chris@1
|
54
|
|
Chris@1
|
55 #copyright {
|
|
Chris@1
|
56 margin-top: 30px;
|
|
Chris@1
|
57 line-height: 1.5em;
|
|
Chris@1
|
58 text-align: center;
|
|
Chris@1
|
59 font-size: .8em;
|
|
Chris@1
|
60 color: #888888;
|
|
Chris@1
|
61 clear: both;
|
|
Chris@1
|
62 }
|
|
Chris@1
|
63 </style>
|
|
Chris@1
|
64
|
|
Chris@1
|
65 </head>
|
|
Chris@1
|
66
|
|
Chris@1
|
67 <body>
|
|
Chris@1
|
68
|
|
Chris@1
|
69 <div id="xiphlogo">
|
|
Chris@1
|
70 <a href="http://www.xiph.org/"><img src="fish_xiph_org.png" alt="Fish Logo and Xiph.Org"/></a>
|
|
Chris@1
|
71 </div>
|
|
Chris@1
|
72
|
|
Chris@1
|
73 <h1>Ogg logical bitstream framing</h1>
|
|
Chris@1
|
74
|
|
Chris@1
|
75 <h2>Ogg bitstreams</h2>
|
|
Chris@1
|
76
|
|
Chris@1
|
77 <p>The Ogg transport bitstream is designed to provide framing, error
|
|
Chris@1
|
78 protection and seeking structure for higher-level codec streams that
|
|
Chris@1
|
79 consist of raw, unencapsulated data packets, such as the Vorbis audio
|
|
Chris@1
|
80 codec or Theora video codec.</p>
|
|
Chris@1
|
81
|
|
Chris@1
|
82 <h2>Application example: Vorbis</h2>
|
|
Chris@1
|
83
|
|
Chris@1
|
84 <p>Vorbis encodes short-time blocks of PCM data into raw packets of
|
|
Chris@1
|
85 bit-packed data. These raw packets may be used directly by transport
|
|
Chris@1
|
86 mechanisms that provide their own framing and packet-separation
|
|
Chris@1
|
87 mechanisms (such as UDP datagrams). For stream based storage (such as
|
|
Chris@1
|
88 files) and transport (such as TCP streams or pipes), Vorbis uses the
|
|
Chris@1
|
89 Ogg bitstream format to provide framing/sync, sync recapture
|
|
Chris@1
|
90 after error, landmarks during seeking, and enough information to
|
|
Chris@1
|
91 properly separate data back into packets at the original packet
|
|
Chris@1
|
92 boundaries without relying on decoding to find packet boundaries.</p>
|
|
Chris@1
|
93
|
|
Chris@1
|
94 <h2>Design constraints for Ogg bitstreams</h2>
|
|
Chris@1
|
95
|
|
Chris@1
|
96 <ol>
|
|
Chris@1
|
97 <li>True streaming; we must not need to seek to build a 100%
|
|
Chris@1
|
98 complete bitstream.</li>
|
|
Chris@1
|
99 <li>Use no more than approximately 1-2% of bitstream bandwidth for
|
|
Chris@1
|
100 packet boundary marking, high-level framing, sync and seeking.</li>
|
|
Chris@1
|
101 <li>Specification of absolute position within the original sample
|
|
Chris@1
|
102 stream.</li>
|
|
Chris@1
|
103 <li>Simple mechanism to ease limited editing, such as a simplified
|
|
Chris@1
|
104 concatenation mechanism.</li>
|
|
Chris@1
|
105 <li>Detection of corruption, recapture after error and direct, random
|
|
Chris@1
|
106 access to data at arbitrary positions in the bitstream.</li>
|
|
Chris@1
|
107 </ol>
|
|
Chris@1
|
108
|
|
Chris@1
|
109 <h2>Logical and Physical Bitstreams</h2>
|
|
Chris@1
|
110
|
|
Chris@1
|
111 <p>A <em>logical</em> Ogg bitstream is a contiguous stream of
|
|
Chris@1
|
112 sequential pages belonging only to the logical bitstream. A
|
|
Chris@1
|
113 <em>physical</em> Ogg bitstream is constructed from one or more
|
|
Chris@1
|
114 than one logical Ogg bitstream (the simplest physical bitstream
|
|
Chris@1
|
115 is simply a single logical bitstream). We describe below the exact
|
|
Chris@1
|
116 formatting of an Ogg logical bitstream. Combining logical
|
|
Chris@1
|
117 bitstreams into more complex physical bitstreams is described in the
|
|
Chris@1
|
118 <a href="oggstream.html">Ogg bitstream overview</a>. The exact
|
|
Chris@1
|
119 mapping of raw Vorbis packets into a valid Ogg Vorbis physical
|
|
Chris@1
|
120 bitstream is described in the Vorbis I Specification.</p>
|
|
Chris@1
|
121
|
|
Chris@1
|
122 <h2>Bitstream structure</h2>
|
|
Chris@1
|
123
|
|
Chris@1
|
124 <p>An Ogg stream is structured by dividing incoming packets into
|
|
Chris@1
|
125 segments of up to 255 bytes and then wrapping a group of contiguous
|
|
Chris@1
|
126 packet segments into a variable length page preceded by a page
|
|
Chris@1
|
127 header. Both the header size and page size are variable; the page
|
|
Chris@1
|
128 header contains sizing information and checksum data to determine
|
|
Chris@1
|
129 header/page size and data integrity.</p>
|
|
Chris@1
|
130
|
|
Chris@1
|
131 <p>The bitstream is captured (or recaptured) by looking for the beginning
|
|
Chris@1
|
132 of a page, specifically the capture pattern. Once the capture pattern
|
|
Chris@1
|
133 is found, the decoder verifies page sync and integrity by computing
|
|
Chris@1
|
134 and comparing the checksum. At that point, the decoder can extract the
|
|
Chris@1
|
135 packets themselves.</p>
|
|
Chris@1
|
136
|
|
Chris@1
|
137 <h3>Packet segmentation</h3>
|
|
Chris@1
|
138
|
|
Chris@1
|
139 <p>Packets are logically divided into multiple segments before encoding
|
|
Chris@1
|
140 into a page. Note that the segmentation and fragmentation process is a
|
|
Chris@1
|
141 logical one; it's used to compute page header values and the original
|
|
Chris@1
|
142 page data need not be disturbed, even when a packet spans page
|
|
Chris@1
|
143 boundaries.</p>
|
|
Chris@1
|
144
|
|
Chris@1
|
145 <p>The raw packet is logically divided into [n] 255 byte segments and a
|
|
Chris@1
|
146 last fractional segment of < 255 bytes. A packet size may well
|
|
Chris@1
|
147 consist only of the trailing fractional segment, and a fractional
|
|
Chris@1
|
148 segment may be zero length. These values, called "lacing values" are
|
|
Chris@1
|
149 then saved and placed into the header segment table.</p>
|
|
Chris@1
|
150
|
|
Chris@1
|
151 <p>An example should make the basic concept clear:</p>
|
|
Chris@1
|
152
|
|
Chris@1
|
153 <pre>
|
|
Chris@1
|
154 <tt>
|
|
Chris@1
|
155 raw packet:
|
|
Chris@1
|
156 ___________________________________________
|
|
Chris@1
|
157 |______________packet data__________________| 753 bytes
|
|
Chris@1
|
158
|
|
Chris@1
|
159 lacing values for page header segment table: 255,255,243
|
|
Chris@1
|
160 </tt>
|
|
Chris@1
|
161 </pre>
|
|
Chris@1
|
162
|
|
Chris@1
|
163 <p>We simply add the lacing values for the total size; the last lacing
|
|
Chris@1
|
164 value for a packet is always the value that is less than 255. Note
|
|
Chris@1
|
165 that this encoding both avoids imposing a maximum packet size as well
|
|
Chris@1
|
166 as imposing minimum overhead on small packets (as opposed to, eg,
|
|
Chris@1
|
167 simply using two bytes at the head of every packet and having a max
|
|
Chris@1
|
168 packet size of 32k. Small packets (<255, the typical case) are
|
|
Chris@1
|
169 penalized with twice the segmentation overhead). Using the lacing
|
|
Chris@1
|
170 values as suggested, small packets see the minimum possible
|
|
Chris@1
|
171 byte-aligned overheade (1 byte) and large packets, over 512 bytes or
|
|
Chris@1
|
172 so, see a fairly constant ~.5% overhead on encoding space.</p>
|
|
Chris@1
|
173
|
|
Chris@1
|
174 <p>Note that a lacing value of 255 implies that a second lacing value
|
|
Chris@1
|
175 follows in the packet, and a value of < 255 marks the end of the
|
|
Chris@1
|
176 packet after that many additional bytes. A packet of 255 bytes (or a
|
|
Chris@1
|
177 multiple of 255 bytes) is terminated by a lacing value of 0:</p>
|
|
Chris@1
|
178
|
|
Chris@1
|
179 <pre><tt>
|
|
Chris@1
|
180 raw packet:
|
|
Chris@1
|
181 _______________________________
|
|
Chris@1
|
182 |________packet data____________| 255 bytes
|
|
Chris@1
|
183
|
|
Chris@1
|
184 lacing values: 255, 0
|
|
Chris@1
|
185 </tt></pre>
|
|
Chris@1
|
186
|
|
Chris@1
|
187 <p>Note also that a 'nil' (zero length) packet is not an error; it
|
|
Chris@1
|
188 consists of nothing more than a lacing value of zero in the header.</p>
|
|
Chris@1
|
189
|
|
Chris@1
|
190 <h3>Packets spanning pages</h3>
|
|
Chris@1
|
191
|
|
Chris@1
|
192 <p>Packets are not restricted to beginning and ending within a page,
|
|
Chris@1
|
193 although individual segments are, by definition, required to do so.
|
|
Chris@1
|
194 Packets are not restricted to a maximum size, although excessively
|
|
Chris@1
|
195 large packets in the data stream are discouraged; the Ogg
|
|
Chris@1
|
196 bitstream specification strongly recommends nominal page size of
|
|
Chris@1
|
197 approximately 4-8kB (large packets are foreseen as being useful for
|
|
Chris@1
|
198 initialization data at the beginning of a logical bitstream).</p>
|
|
Chris@1
|
199
|
|
Chris@1
|
200 <p>After segmenting a packet, the encoder may decide not to place all the
|
|
Chris@1
|
201 resulting segments into the current page; to do so, the encoder places
|
|
Chris@1
|
202 the lacing values of the segments it wishes to belong to the current
|
|
Chris@1
|
203 page into the current segment table, then finishes the page. The next
|
|
Chris@1
|
204 page is begun with the first value in the segment table belonging to
|
|
Chris@1
|
205 the next packet segment, thus continuing the packet (data in the
|
|
Chris@1
|
206 packet body must also correspond properly to the lacing values in the
|
|
Chris@1
|
207 spanned pages. The segment data in the first packet corresponding to
|
|
Chris@1
|
208 the lacing values of the first page belong in that page; packet
|
|
Chris@1
|
209 segments listed in the segment table of the following page must begin
|
|
Chris@1
|
210 the page body of the subsequent page).</p>
|
|
Chris@1
|
211
|
|
Chris@1
|
212 <p>The last mechanic to spanning a page boundary is to set the header
|
|
Chris@1
|
213 flag in the new page to indicate that the first lacing value in the
|
|
Chris@1
|
214 segment table continues rather than begins a packet; a header flag of
|
|
Chris@1
|
215 0x01 is set to indicate a continued packet. Although mandatory, it
|
|
Chris@1
|
216 is not actually algorithmically necessary; one could inspect the
|
|
Chris@1
|
217 preceding segment table to determine if the packet is new or
|
|
Chris@1
|
218 continued. Adding the information to the packet_header flag allows a
|
|
Chris@1
|
219 simpler design (with no overhead) that needs only inspect the current
|
|
Chris@1
|
220 page header after frame capture. This also allows faster error
|
|
Chris@1
|
221 recovery in the event that the packet originates in a corrupt
|
|
Chris@1
|
222 preceding page, implying that the previous page's segment table
|
|
Chris@1
|
223 cannot be trusted.</p>
|
|
Chris@1
|
224
|
|
Chris@1
|
225 <p>Note that a packet can span an arbitrary number of pages; the above
|
|
Chris@1
|
226 spanning process is repeated for each spanned page boundary. Also a
|
|
Chris@1
|
227 'zero termination' on a packet size that is an even multiple of 255
|
|
Chris@1
|
228 must appear even if the lacing value appears in the next page as a
|
|
Chris@1
|
229 zero-length continuation of the current packet. The header flag
|
|
Chris@1
|
230 should be set to 0x01 to indicate that the packet spanned, even though
|
|
Chris@1
|
231 the span is a nil case as far as data is concerned.</p>
|
|
Chris@1
|
232
|
|
Chris@1
|
233 <p>The encoding looks odd, but is properly optimized for speed and the
|
|
Chris@1
|
234 expected case of the majority of packets being between 50 and 200
|
|
Chris@1
|
235 bytes (note that it is designed such that packets of wildly different
|
|
Chris@1
|
236 sizes can be handled within the model; placing packet size
|
|
Chris@1
|
237 restrictions on the encoder would have only slightly simplified design
|
|
Chris@1
|
238 in page generation and increased overall encoder complexity).</p>
|
|
Chris@1
|
239
|
|
Chris@1
|
240 <p>The main point behind tracking individual packets (and packet
|
|
Chris@1
|
241 segments) is to allow more flexible encoding tricks that requiring
|
|
Chris@1
|
242 explicit knowledge of packet size. An example is simple bandwidth
|
|
Chris@1
|
243 limiting, implemented by simply truncating packets in the nominal case
|
|
Chris@1
|
244 if the packet is arranged so that the least sensitive portion of the
|
|
Chris@1
|
245 data comes last.</p>
|
|
Chris@1
|
246
|
|
Chris@1
|
247 <h3>Page header</h3>
|
|
Chris@1
|
248
|
|
Chris@1
|
249 <p>The headering mechanism is designed to avoid copying and re-assembly
|
|
Chris@1
|
250 of the packet data (ie, making the packet segmentation process a
|
|
Chris@1
|
251 logical one); the header can be generated directly from incoming
|
|
Chris@1
|
252 packet data. The encoder buffers packet data until it finishes a
|
|
Chris@1
|
253 complete page at which point it writes the header followed by the
|
|
Chris@1
|
254 buffered packet segments.</p>
|
|
Chris@1
|
255
|
|
Chris@1
|
256 <h4>capture_pattern</h4>
|
|
Chris@1
|
257
|
|
Chris@1
|
258 <p>A header begins with a capture pattern that simplifies identifying
|
|
Chris@1
|
259 pages; once the decoder has found the capture pattern it can do a more
|
|
Chris@1
|
260 intensive job of verifying that it has in fact found a page boundary
|
|
Chris@1
|
261 (as opposed to an inadvertent coincidence in the byte stream).</p>
|
|
Chris@1
|
262
|
|
Chris@1
|
263 <pre><tt>
|
|
Chris@1
|
264 byte value
|
|
Chris@1
|
265
|
|
Chris@1
|
266 0 0x4f 'O'
|
|
Chris@1
|
267 1 0x67 'g'
|
|
Chris@1
|
268 2 0x67 'g'
|
|
Chris@1
|
269 3 0x53 'S'
|
|
Chris@1
|
270 </tt></pre>
|
|
Chris@1
|
271
|
|
Chris@1
|
272 <h4>stream_structure_version</h4>
|
|
Chris@1
|
273
|
|
Chris@1
|
274 <p>The capture pattern is followed by the stream structure revision:</p>
|
|
Chris@1
|
275
|
|
Chris@1
|
276 <pre><tt>
|
|
Chris@1
|
277 byte value
|
|
Chris@1
|
278
|
|
Chris@1
|
279 4 0x00
|
|
Chris@1
|
280 </tt></pre>
|
|
Chris@1
|
281
|
|
Chris@1
|
282 <h4>header_type_flag</h4>
|
|
Chris@1
|
283
|
|
Chris@1
|
284 <p>The header type flag identifies this page's context in the bitstream:</p>
|
|
Chris@1
|
285
|
|
Chris@1
|
286 <pre><tt>
|
|
Chris@1
|
287 byte value
|
|
Chris@1
|
288
|
|
Chris@1
|
289 5 bitflags: 0x01: unset = fresh packet
|
|
Chris@1
|
290 set = continued packet
|
|
Chris@1
|
291 0x02: unset = not first page of logical bitstream
|
|
Chris@1
|
292 set = first page of logical bitstream (bos)
|
|
Chris@1
|
293 0x04: unset = not last page of logical bitstream
|
|
Chris@1
|
294 set = last page of logical bitstream (eos)
|
|
Chris@1
|
295 </tt></pre>
|
|
Chris@1
|
296
|
|
Chris@1
|
297 <h4>absolute granule position</h4>
|
|
Chris@1
|
298
|
|
Chris@1
|
299 <p>(This is packed in the same way the rest of Ogg data is packed; LSb
|
|
Chris@1
|
300 of LSB first. Note that the 'position' data specifies a 'sample'
|
|
Chris@1
|
301 number (eg, in a CD quality sample is four octets, 16 bits for left
|
|
Chris@1
|
302 and 16 bits for right; in video it would likely be the frame number.
|
|
Chris@1
|
303 It is up to the specific codec in use to define the semantic meaning
|
|
Chris@1
|
304 of the granule position value). The position specified is the total
|
|
Chris@1
|
305 samples encoded after including all packets finished on this page
|
|
Chris@1
|
306 (packets begun on this page but continuing on to the next page do not
|
|
Chris@1
|
307 count). The rationale here is that the position specified in the
|
|
Chris@1
|
308 frame header of the last page tells how long the data coded by the
|
|
Chris@1
|
309 bitstream is. A truncated stream will still return the proper number
|
|
Chris@1
|
310 of samples that can be decoded fully.</p>
|
|
Chris@1
|
311
|
|
Chris@1
|
312 <p>A special value of '-1' (in two's complement) indicates that no packets
|
|
Chris@1
|
313 finish on this page.</p>
|
|
Chris@1
|
314
|
|
Chris@1
|
315 <pre><tt>
|
|
Chris@1
|
316 byte value
|
|
Chris@1
|
317
|
|
Chris@1
|
318 6 0xXX LSB
|
|
Chris@1
|
319 7 0xXX
|
|
Chris@1
|
320 8 0xXX
|
|
Chris@1
|
321 9 0xXX
|
|
Chris@1
|
322 10 0xXX
|
|
Chris@1
|
323 11 0xXX
|
|
Chris@1
|
324 12 0xXX
|
|
Chris@1
|
325 13 0xXX MSB
|
|
Chris@1
|
326 </tt></pre>
|
|
Chris@1
|
327
|
|
Chris@1
|
328 <h4>stream serial number</h4>
|
|
Chris@1
|
329
|
|
Chris@1
|
330 <p>Ogg allows for separate logical bitstreams to be mixed at page
|
|
Chris@1
|
331 granularity in a physical bitstream. The most common case would be
|
|
Chris@1
|
332 sequential arrangement, but it is possible to interleave pages for
|
|
Chris@1
|
333 two separate bitstreams to be decoded concurrently. The serial
|
|
Chris@1
|
334 number is the means by which pages physical pages are associated with
|
|
Chris@1
|
335 a particular logical stream. Each logical stream must have a unique
|
|
Chris@1
|
336 serial number within a physical stream:</p>
|
|
Chris@1
|
337
|
|
Chris@1
|
338 <pre><tt>
|
|
Chris@1
|
339 byte value
|
|
Chris@1
|
340
|
|
Chris@1
|
341 14 0xXX LSB
|
|
Chris@1
|
342 15 0xXX
|
|
Chris@1
|
343 16 0xXX
|
|
Chris@1
|
344 17 0xXX MSB
|
|
Chris@1
|
345 </tt></pre>
|
|
Chris@1
|
346
|
|
Chris@1
|
347 <h4>page sequence no</h4>
|
|
Chris@1
|
348
|
|
Chris@1
|
349 <p>Page counter; lets us know if a page is lost (useful where packets
|
|
Chris@1
|
350 span page boundaries).</p>
|
|
Chris@1
|
351
|
|
Chris@1
|
352 <pre><tt>
|
|
Chris@1
|
353 byte value
|
|
Chris@1
|
354
|
|
Chris@1
|
355 18 0xXX LSB
|
|
Chris@1
|
356 19 0xXX
|
|
Chris@1
|
357 20 0xXX
|
|
Chris@1
|
358 21 0xXX MSB
|
|
Chris@1
|
359 </tt></pre>
|
|
Chris@1
|
360
|
|
Chris@1
|
361 <h4>page checksum</h4>
|
|
Chris@1
|
362
|
|
Chris@1
|
363 <p>32 bit CRC value (direct algorithm, initial val and final XOR = 0,
|
|
Chris@1
|
364 generator polynomial=0x04c11db7). The value is computed over the
|
|
Chris@1
|
365 entire header (with the CRC field in the header set to zero) and then
|
|
Chris@1
|
366 continued over the page. The CRC field is then filled with the
|
|
Chris@1
|
367 computed value.</p>
|
|
Chris@1
|
368
|
|
Chris@1
|
369 <p>(A thorough discussion of CRC algorithms can be found in <a
|
|
Chris@1
|
370 href="http://www.ross.net/crc/download/crc_v3.txt">"A
|
|
Chris@1
|
371 Painless Guide to CRC Error Detection Algorithms"</a> by Ross
|
|
Chris@1
|
372 Williams <a href="mailto:ross@ross.net">ross@ross.net</a>.)</p>
|
|
Chris@1
|
373
|
|
Chris@1
|
374 <pre><tt>
|
|
Chris@1
|
375 byte value
|
|
Chris@1
|
376
|
|
Chris@1
|
377 22 0xXX LSB
|
|
Chris@1
|
378 23 0xXX
|
|
Chris@1
|
379 24 0xXX
|
|
Chris@1
|
380 25 0xXX MSB
|
|
Chris@1
|
381 </tt></pre>
|
|
Chris@1
|
382
|
|
Chris@1
|
383 <h4>page_segments</h4>
|
|
Chris@1
|
384
|
|
Chris@1
|
385 <p>The number of segment entries to appear in the segment table. The
|
|
Chris@1
|
386 maximum number of 255 segments (255 bytes each) sets the maximum
|
|
Chris@1
|
387 possible physical page size at 65307 bytes or just under 64kB (thus
|
|
Chris@1
|
388 we know that a header corrupted so as destroy sizing/alignment
|
|
Chris@1
|
389 information will not cause a runaway bitstream. We'll read in the
|
|
Chris@1
|
390 page according to the corrupted size information that's guaranteed to
|
|
Chris@1
|
391 be a reasonable size regardless, notice the checksum mismatch, drop
|
|
Chris@1
|
392 sync and then look for recapture).</p>
|
|
Chris@1
|
393
|
|
Chris@1
|
394 <pre><tt>
|
|
Chris@1
|
395 byte value
|
|
Chris@1
|
396
|
|
Chris@1
|
397 26 0x00-0xff (0-255)
|
|
Chris@1
|
398 </tt></pre>
|
|
Chris@1
|
399
|
|
Chris@1
|
400 <h4>segment_table (containing packet lacing values)</h4>
|
|
Chris@1
|
401
|
|
Chris@1
|
402 <p>The lacing values for each packet segment physically appearing in
|
|
Chris@1
|
403 this page are listed in contiguous order.</p>
|
|
Chris@1
|
404
|
|
Chris@1
|
405 <pre><tt>
|
|
Chris@1
|
406 byte value
|
|
Chris@1
|
407
|
|
Chris@1
|
408 27 0x00-0xff (0-255)
|
|
Chris@1
|
409 [...]
|
|
Chris@1
|
410 n 0x00-0xff (0-255, n=page_segments+26)
|
|
Chris@1
|
411 </tt></pre>
|
|
Chris@1
|
412
|
|
Chris@1
|
413 <p>Total page size is calculated directly from the known header size and
|
|
Chris@1
|
414 lacing values in the segment table. Packet data segments follow
|
|
Chris@1
|
415 immediately after the header.</p>
|
|
Chris@1
|
416
|
|
Chris@1
|
417 <p>Page headers typically impose a flat .25-.5% space overhead assuming
|
|
Chris@1
|
418 nominal ~8k page sizes. The segmentation table needed for exact
|
|
Chris@1
|
419 packet recovery in the streaming layer adds approximately .5-1%
|
|
Chris@1
|
420 nominal assuming expected encoder behavior in the 44.1kHz, 128kbps
|
|
Chris@1
|
421 stereo encodings.</p>
|
|
Chris@1
|
422
|
|
Chris@1
|
423 <div id="copyright">
|
|
Chris@1
|
424 The Xiph Fish Logo is a
|
|
Chris@1
|
425 trademark (™) of Xiph.Org.<br/>
|
|
Chris@1
|
426
|
|
Chris@1
|
427 These pages © 1994 - 2005 Xiph.Org. All rights reserved.
|
|
Chris@1
|
428 </div>
|
|
Chris@1
|
429
|
|
Chris@1
|
430 </body>
|
|
Chris@1
|
431 </html>
|
