[vorbis] Multiple tracks per file, or continuous audio support

[Michael Smith]

>Is this the same thing though?  Just because it
>truncates the file to a certain length, can you be
>sure the waveform will match properly with the
>beginning of the next track?  Wouldn't it be better if
>there was a special way to tell the Ogg encoder what
>the first couple of milliseconds of the next track
>were, so the frames could be encoded more towards
>meeting the signal later?  Maybe that's overkill.  I'm
>not sure what influence the zero-padding has backwards
>on the signal.

While this is technically correct, in practice it's inaudible. 
I did some semi-extensive experiments with this a while back, 
the transition between the tracks isn't noticable. 

Also note that the padding isn't actualyl zero-padding, the encoder
extrapolates more intelligently than that.

This isn't the only way to do things - it's actually possible to 
encode the entire thing as one file, and then later do sample-
precise editing of the file to cut it into multiple sub-files - this
way you get the same as just encoding them seperately, but the 
overlap region is guaranteed to be 'smooth'. There's even a tool to
do the cutting (though not fully automatically), called 'vcut'.

>
>I guess I'm skeptical of the first solution.  I *did*
>encode two tracks that play continuously, and played
>them back in Ogg-enabled Winamp.  There's a small pop.
> This could be the player, though, cheating with its
>own zero frames.

This sounds like a pretty nasty player bug, it definately
doesn't happen with any of the players I use. I'd send a 
bug report to the plugin author, and/or the winamp people.

Michael

--- >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-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.