[mythtv] [patch] multi channel audio patch 1
Michael T. Dean
mtdean at thirdcontact.com
Fri Dec 30 13:38:38 EST 2005
Mark Spieth wrote:
> attached is the next version of the multi channel audio patch.
> I have tested it with ALSA and it works up to 6 channels with
> timestretch with a AC3.5.1 avi I have. I have only listened to 2
> channels connected to my sound card though so this needs more testing.
> Ive limited it to 6 channels at the moment.
> AC3 passthrough support is not handled yet so you need to use all 3
> jacks from your audio card to test currently. So dont use it yet (ac3
> passthrough).
>
> however from my alsa reading you can get alsa to reencode te 6 channel
> audio back to AC3 but I dont know how this is done. If this is true, I
> may just leave it at that and not reencode in myth (easier for me :) ).
> There is a bit of code for support of ac3 reencode in audio layer
> which I may back out if an ALSA setup is viable.
Instead of re-encoding to AC-3, ALSA will allow 6-channel PCM output via
the digital connection. This is actually preferrable to AC-3
re-encoding because AC-3 is a lossy compression format (and PCM is
uncompressed and, therefore, not lossy).
That being said, I haven't tested your patch (holiday visitors) and I
only have 2 speakers in my setup (although I could "fake" the others
with creative shuffling). As soon as the visitors are gone (next week),
I'll finally have a chance to work on my HDTV setup (complete rebuild),
and I'll do some testing (and write up an ALSA config, if required).
(I'll need to do this because I can't watch TV without timestretch,
anymore, and the only way I'll do more than 2-channel output is if
timestretch works with it--so thanks for the timestretch and for the
multi-channel work. :)
The only benefit I can see of fixing the AC-3 would be allowing the user
to offload AC-3 decoding to an external A/V receiver. However, keeping
the AC-3 decoding on the Myth box should make A/V sync easier to manage
(because we can pretty much assume PCM decoding will be "instantaneous",
whereas AC-3 decoding may not be). Therefore, I would recommend you
save your time and don't worry about AC-3--at least, not yet. (I'm even
volunteering to reply to all the "How come it won't do AC-3 with
timestretch? I want AC-3 for better sound quality..." posts. :)
Mike
More information about the mythtv-dev
mailing list