This is basically the NeXT format. I have no objections to the
fundamental format, but I *do* have a comment on the number of encodings
supported.
Basically, I think we should support as few formats as we can get away
with, and try to insist that if you claim to speak audio, you should
support all of these formats.
Specifically, I propose:
renaming the proposal to be a definition for the "AUDIO/BASIC" type
limiting the set of valid data encodings to type "1" (ISDN u-law)
and type "23" (CCITT G.721 ADPCM).
a sample rate of 8000, using 1 channel
The current spec is just too broad. It is the set of things that
the next can support, but it is unreasonable to expect all people
to be able to deal with all types.
I could even be talked into eliminating type "1" (ISDN u-law) from
the acceptable list, and limiting things to exactly one audio encoding.
Neil Katin