[Vorbis-dev] 5.1 surround channel coupling

[Sebastian Olter]

2007/2/24, Richard Lee <ricardo at justnet.com.au>:
> Does this mean there is an existing B-format to Vorbis encoder?

Yes, I posted an initial version of the encoder here several days ago.

> And a Vorbis to B-format decoder?
> How do I find what design decisions were made in this?
> Particularly the frequency banding and "phase" model used.

Current Vorbis Ambisonic streams do not need to be decoded into
B-Format, as they're already encoded into it. Such a design seems to
be as clear as possible--and in fact, they are perfectly decodable
even if no dedicated decoder currently exists. I will create a decoder
too (currently waiting for ffmpeg people to answer), and then it'll be
your time to discuss with others how to improve Vorbis encoding
efficiency with Ambisonics and to made all appropriate decisions.

> I was hoping to become a Vorbis guru and extend Vorbis to do Ambisonics properly.  It's not as complicated as I originally envisaged.  But I need more info.

Vorbis started without any stereo channel coupling, and it worked and
was much better than other lossy audio codecs. I believe (well,
actually I have the opportunity to know it) the same applies to
Ambisonics. The only thing that is actually needed is a single bit in
the Vorbis stream that marks it as Ambisonic; all the rest is handled
by the channels number (2 for UHJ, 3 and better for B-Format).
G-Format could simply remain unmarked, if it will be ever used.

The most interesting application of Ambisonic channel coupling is a
very low bitrate stream, such as movie sound track. So maybe you could
think about how to use Vorbis mechanisms to do UHJ; perhaps it will
take much less work than a regular multichannel coupling. UHJ is
proven to work and it may be valuable to do listening tests with it.

Regards

--
Sebastian