Universal Ambisonic in context

[e deleflie]

Gentlemen,

I'd just like to state how I see Universal Ambisonic in the context of
these discussions.

I think Universal Ambisonic has 2 strengths

1) it is an end-to-end specification... from what encoders must
satisfy to what decoders must satisfy etc.
2) it is very focused on software integration

I no longer believe that it makes sense to offer specific mixed orders
... I think it goes against ambisonic's promise. Or, at least, the
mixed orders should occur _after_ the encoding ... at the very end of
the chain ... when a file is being delivered to a specific array....
that's when it makes sense to break a file down to specific mixed
orders.

There is an elegant simplicity with just having 16 channels... for everything.

There are 2 challenges:
1) how to mix first order with higher order
2) an ambisonic specific compression scheme (maybe the compression
scheme just drops higher orders... when delivering to an array it
knows only has 1st order speakers)

I'm thinking of putting up a challenge to the Ogg Vorbis people to
design a specific ambisonic compression scheme ...

Etienne