[Vorbis-dev] Ambisonics

[Oliver Thuns]

2009/3/20 Richard Furse <rf015d9821 at blueyonder.co.uk>:
> Hi there - we've not discussed Ambisonics here for a while and I was
> wondering what the appetite is like to get stuff done.
>
> A fair bit of discussion has happened around the place (e.g. on the sursound
> list and beyond) on exactly what encoding (i.e. channel interpretations)
> should be used for Ambisonics, resulting in a number of new proposals, some
> of which have been or are being written up as relatively formal documents.
> There are also existing standards (e.g. in WAVE-EX and Ogg-PCM) and some
> folk are getting together in June to discuss things further. I have views on
> this, but I don't really want to start a format war off here right now ;-)

To avoid a format war and to stay focused, I suggest implementing only
1st order Ambisonics in the first step. First order B-Format is still
the standard and will be used for a long time. I feel an API for
encoding from B-Format and decoding to B-Format is a must-have, even
if the Ambisonics community will agree on a new format that is
incompatible with B-Format.

(When we talk about different Ambisonic formats, it's a little bit
like the different 5.1 conventions. LRC vs. LCR vs. CLR and LFE with
0db or 10db)

> How would we actually go about sorting out some decent Ambisonic support for
> Vorbis? Are there folk here ready and enthusiastic to get into the
> development work? If so, do they have a clear idea of what might be needed?
> If not, I should be able to make myself available to help code this up,
> though I probably wouldn't have time to come up with a good fast solution to
> the core multichannel compression problem. However, I've looked at this
> before and can probably help.

IIRC, Gregory Maxwell implemented Ambisonics encoding in Vorbis for
testing purposes. The sources have been lost, but I guess he is the
one who knows how to do it. Gregory, are you following the Ambisonics
discussion on this list?