[Vorbis-dev] Mapping = 1 Ambisonic Vorbis flag

Oliver Oli oliver.oli+0815 at gmail.com
Mon Sep 8 09:26:43 PDT 2008


On Mon, Sep 8, 2008 at 3:00 PM, Paul Martin <pm at nowster.org.uk> wrote:
> On Mon, Sep 08, 2008 at 12:33:07PM +0100, Paul Martin wrote:
>> On Mon, Sep 08, 2008 at 09:25:08AM +0000, Oliver Oli wrote:
>> > 1:  W
>> > 2:  WX
>> > 3:  WXY
>
>> (Incidentally, you don't need a separate representation for middle/side, as
>> Ambisonic's "WX" is equivalent.)
>
> As Richard Furse has pointed out off-list, WY is the correct encoding for
> mid-side. Suggestion 2 above would therefore be better as WY rather than WX.
>
> The axes of the first order components are:
>
> W is omnidirectional.
> X is +front -back
> Y is +left -right
> Z is +up -down

You are right. I always wondered why X is front/back and messed it up
again. When I'm thinking about now, it does not make much sense to
support 1 and 2 channels at all, that's not really Ambisonics. But why
not use the OggPCM channel mapping definition, instead of inventing
something new?

My proposal is as simple and limited ad the type 0 channel mapping
definition. OggPCM spec defines a full feature channel mapping set,
but I don't know how well it fits into Vorbis. I'm not involved in Ogg
and Vorbis development and I don't have the knowledge to decide what's
the best solution for extending Vorbis.


More information about the Vorbis-dev mailing list