[vorbis-dev] Metadata streams

rillian at telus.net rillian at telus.net
Mon Apr 16 18:54:15 PDT 2001



On Monday, April 16, 2001, at 02:08 , Aaron Plattner wrote:

>> You can do something similar... a stream can have only very sporadic
>> packets sprinkled through the stream or even the first packet halfway
>> through... so long as it's declared in the TOC.  This will all be
>> nailed down soon.  No mechanism can be all-inclusive.

> Oh... THAT'S what I thought was prohibited by the spec.

Me too. Monty's being wierd. :-)

Part of it works under the here-to-for-declared scheme, anyway. The 
interleave rate can certainly be small and variable. The important point 
is whether all substream are bounded by 'beginning-of-stream' and 
'end-of-stream' pages.

A more concrete example: You're an audio player receiving a stream that 
doesn't initially contain any audio data. If substreams can start at any 
point, you don't have a good place to decide 'this is not an audio file' 
and error out. Maybe the startup data for the other streams is just 
unreasonably large? You don't know. But under the current spec you do 
know, because every substream must begin together. Once you see the 
first non-bos page in the stream, you know you've seen all the data 
types you're going to, and if there's no vorbis, there's no vorbis. 
Technically, be cause such bundles ('groups') of substreams can be 
concatenated ('chained') you might want to keep looking and try again at 
the next boundary, but in the meantime you can at least warn and offer 
to abort.

Best I could come up with so far,
  -ralph

--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.



More information about the Vorbis-dev mailing list