[FFmpeg-devel] GSoC mentor for audio module playback system requested

[Jai Menon]

On Sat, Apr 17, 2010 at 7:50 PM, Sebastian Vater
<cdgs.basty at googlemail.com> wrote:
> Hey Jai!
>
> Jai Menon a ?crit :
>> Is the design already fleshed out somewhere? IIRC Peter had some stuff
>> for pattern/sample storage and representation, maybe future work could
>> build upon that. Also, I hope the renderer would be module format
>> agnostic - perhaps multiple module parsers which map to some internal
>> format/superset of effect commands. Also, channel mixing code would
>> need to be independent since we could probably re-use it elsewhere
>> too.
>>
> I have already adressed an agnostic module format long time ago.
> It's called TuComposer (The united Composer).
>
> More details (including specs and source code for Amiga) can be found here:
> http://forum.cdgs-crew.com/viewforum.php?f=11

To clarify, I meant the _internal_ format we would use within ffmpeg.
Module loaders imho cannot be represented within the existing
demuxers+decoders framework. So we would (possibly) have loaders which
fill an AVModule/AVPattern/whatever, defined functions which extract
(and maybe unpack) pattern data and feed the renderer, manage
instrument/sample data etc. And of course reusing code where
possible...

>>> while we agreed on
>>> the IFF-ANIM qualification task.
>>>
>>
>> Which format exactly?
>>
> I'm looking at that currently, esp. the format which has to be reverse
> engineered (don't have the exact name in mind right now, will look at it
> if desired).

I don't remember anything in particular which required RE, but I might
be forgetting something...

-- 
Jai Menon