[ogg-dev] OggPCM family

[Sampo Syreeni]

On 2007-10-19, Martin Leese wrote:

>    OggPCM Draft3

Draft 3 is obviously a joke. Draft 2 is what most of the people agreed 
upon the last time around, with the channel maps left unfinished. Draft 
1 was abandoned by most people in favour of draft 2.

> I suggest this because somebody has started making changes to OggPCM 
> Draft2.

That someone is me. I've asked about this on-list a while back 
(http://lists.xiph.org/pipermail/ogg-dev/2007-October/000644.html), and 
tried to spur some interest on #vorbis as well. Since nobody seemed too 
interested but didn't object to my suggestion of helping out either, I 
went ahead and added a proposed channel mapping section, to see if I can 
get some discussion going that way. I haven't disturbed the existing 
channel map section, eventhough it lacks channel_type definitions, and 
I've tried to keep my changes conservative otherwise. Apparently you 
noticed my changes right after I made them, and before I had time to ask 
for comments on-list.

Against that background, do you have suggestions wrt to the channel map 
section? What I've tried to do is 1) add channel definitions from a 
number of sources (like 5.1, 6.1, 7.1, 10.2, 22.2, Apple's Core Audio 
spec, and so on) so that roundtrip compatibility becomes possible, 2) 
keep the existing mapping approach, because at least Jean-Marc Valen has 
indicated (on IRC) preference for it over the restrictions I suggested a 
couple of years ago, 3) add the multiple, preferential mapping approach 
because at least John Koleszar agreed with it the last time around and 
nobody objected, and 4) cover as many of the special cases, nits, 
canonicalization issues, and so on, as I can think of.
-- 
Sampo Syreeni, aka decoy - mailto:decoy at iki.fi, tel:+358-50-5756111
student/math+cs/helsinki university, http://www.iki.fi/~decoy/front
openpgp: 050985C2/025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2