[ogg-dev] [Vorbis] MT9 Capabilities

Silvia Pfeiffer silviapfeiffer1 at gmail.com
Mon Aug 18 15:46:11 PDT 2008


To add some more information to this thread:

* Ogg allows multiplexing of any number of audio, video, timed text
(subtitle, annotation, karaoke) etc tracks (also called: logical
bitstreams) together in one Ogg physical bitstream

* Skeleton is a header that marks for such a complex Ogg physical
bitstream what logical bitstreams are actually inside the file:
http://xiph.org/ogg/doc/skeleton.html

* also, at the last FOMS workshop we defined ROE-
http://wiki.xiph.org/index.php/ROE - a rich way to describe the
relationships between the different logical bitstreams; it would help
a player decide what to play back and how; the only implementation of
ROE is so far in Metavidwiki, e.g. if you click the little caterpillar
image at the top right at
http://metavid.ucsc.edu/wiki/index.php/Stream:Screen_cast_march_08
(http://metavid.ucsc.edu/wiki/index.php?title=Special:MvExportStream&feed_format=roe&stream_name=Screen_cast_march_08&t=0:00:00/0:14:45).
Full ROE spec as it currently stands is here:
http://svn.annodex.net/standards/roe/roe_1_0.xsd

As for trying to adjust volumes on the different tracks (as per the
original request) all you really need is an application that can
separately control for a multi-track Ogg audio file the volumes on the
different tracks (the multitrack Ogg file should come with a
skeleton).

If you wanted to store the volume changes back to a file and re-play,
you have several choices. You could extend ROE with a means to
describe the volume changes, or you could define a new annotation
track that would describe the volume changes of the different skeleton
tracks, or you could add the volume changes to the meta headers of a
CMML track (see
http://www.annodex.net/TR/draft-pfeiffer-cmml-03.html), or you could
re-encode the audio tracks at the changed volumes, and there's
probably a number more that I haven't considered yet. What's the best
option? To be honest, I don't know - that depends on your goals and
how you want to implement the application.

BTW: you could try and play around with gstreamer and filters on the
different tracks of a multi-track Ogg file to get track-level volume
changes happening. That would probably already provide what your need
for the playback part of it.

Cheers,
Silvia.


On Mon, Aug 18, 2008 at 10:25 PM, Gregory Maxwell <gmaxwell at gmail.com> wrote:
> On Mon, Aug 18, 2008 at 6:36 AM, Danny Piccirillo
> <danny.piccirillo at ubuntu.com> wrote:
>> Sorry, but i wasn't sure the best place to post this. I was looking for some
>> forums but it doesn't look like Xiph has any official ones. Anyways, i heard
>> about this MT9 think a while ago and i thought it would be awesome if FLAC,
>> Vorbis, and all could implement something like this.
>>
>> http://www.guardian.co.uk/music/2008/may/27/news.seanmichaels
>> http://en.wikipedia.org/wiki/MT9
>>
>> In combination with that, it would also be cool to have some sort of
>> subtitles-esque thing for audio which would be useful for speeches and
>> karaoke.
>>
>> These two things would be a very very big boost for free audio formats :)
>>
>> How feasible is this in the long run?
>
> Vorbis supports 255 channels in a single stream, FLAC 8. Ogg would let
> you encapsulate many streams in one file, even combining Vorbis, FLAC,
> and Speex.  The open questions are application support and metadata
> for signaling the right default settings.
> _______________________________________________
> ogg-dev mailing list
> ogg-dev at xiph.org
> http://lists.xiph.org/mailman/listinfo/ogg-dev
>


More information about the ogg-dev mailing list