[Vorbis] Some ideas about embeding data in ogg files...

[Jason Brower]

Greetings, Name's Jason,

I am thinking of some kind of software I can develop as a final project
at my school before I graduate.
My idea is to make an application that would let people manually enter
"beat streams" that can follow along with the music.  These "beat
streams" how ever they are processed could be used in many ways.
Visualizations, strobe lights, mood guessing, beats staying the same as
they crass fade from one song to another, etc...
Ogg is a great format, and I think if we could integrate this sequencing
system into an ogg file we could really come up with something that lets
music and other sounds move beyond just playing and into better quality
through better integration.
In my project I would make 4 parts...
The sequence encoder, something that can play the music and record the
beats that I manually enter.
The documentation and implemenation of the "beat stream" that will be
incoded with the ogg.
A basic Visualization: whether from something now, or something new, I
want to show that the visuals work better with this idea.  A proof of
concept if you will.
And last, a website that can store these files and include instructions
on how this is all done with video demonstrations and other information
on how we can use this idea.  Additionally, I want to include a basic
rating system to help keep the good tracks going.  We could include both
free music and "match tracks" that can be imported with your player to
include the beat stream with your encoded oggs.
So, my question is, is there a way to embed data into an ogg file?  And
if not, could we have something like that with backwards compatibility?
Personally, I think this data can be more vital then the current meta
data that we have. (Songs still sound good with bad in accurate meta
data, but to have a beat stream or the possibilities!"
So what you do you think? Something worth going forward with?
Best Regards,
Jason Brower
PS
Other places were this squencing could become using if by declaring
chapters in a podcast, or sides in a conversation.  Other ideas will
emerge over time I am sure. :)