[MPlayer-dev-eng] [RFC] 6-channel AAC and channel reordering

[Corey Hickey]

Dominik 'Rathann' Mierzejewski wrote:
> Reviving an old, forgotten patch...

All right, here's an update. I'm sleepy by now, so there might be
something wrong that's easily spotted with fresh eyes. I'll take a
closer look tomorrow, but it should be testable.

New stuff:
* a little cleaner; better error checking
* remapping for 6-channel WMA (in ad_dmo.c)
  This is UNTESTED. I won't apply it without testing later, but somebody
  else can test it now if they want to.
* remapping for 6-channel PCM (see below)



On 2006-07-30, Alexander 'Lazy Ranma' Ponyatikh sent a patch for
decoding 6-channel PCM directly into each correct channel. I benchmarked
it, and it seems a bit slower than using af_channels, probably because
it calls demux_read_data() repeatedly and only reads two samples at a time.

I tested on a 6-channel wav file made from The Matrix DVD. It is 8177.3
seconds long. Each number is the user-time average of three tests, but
there really wasn't much variation. The command is:
$ time mplayer -ao pcm:fast:file=/dev/null -channels 6 audiodump.wav


method                          user time     % of play time
------------------------------------------------------------
original (incorrect channels):     9.099s           0.11127%
using af_channels to remap:       15.057s           0.18413%
reading into each channel:        30.296s           0.37048%

By the way, Alexander, if you're reading this, don't think I'm
denigrating your work; the information you sent is helpful and I might
end up applying your ae_*.c patches. Also, if you come up with a faster
approach for ad_pcm, feel free to submit it and we can consider using
that instead of af_channels for PCM decoding.

-Corey
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: chan_map2.diff
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20061030/dac0c6f2/attachment.asc>