[alsa-devel] wrong channel mappings for HDMI audio

[Anssi Hannula]

On 13.11.2010 02:26, Dan Christensen wrote:
> Anssi Hannula <anssi.hannula at iki.fi> writes:
> 
>> This is a known issue, the NVIDIA MCP79/7A HDMI hardware has incorrect
>> channel mapping.
>>
>> I reported this several months ago as "Wrong channel order with
>> multichannel HDMI on MCP7A":
>> http://www.spinics.net/lists/alsa-devel/msg35948.html
>> (there are some earlier reports from 2009 as well)
> 
> Great, thanks for confirming it's not just me.
> 
>> I'm using this workaround at the moment:
>>  pcm.!hdmi {
>>      type             route
>>      slave.pcm "cards.HDA-Intel.pcm.hdmi.0:CARD=NVidia,AES0=0x4,AES1=0x82,AES2=0x0,AES3=0x2"
>>      ttable.0.0 1
>>      ttable.1.1 1
>>      ttable.2.4 1
>>      ttable.3.5 1
>>      ttable.4.2 1
>>      ttable.5.3 1
>>      ttable.6.6 1
>>      ttable.7.7 1
>>  }
> 
> I can confirm that this fixes things for me.
> 
>> (not a perfect workaround as I'm hardcoding AESx instead of using the
>> ones provided as arguments, but at least you get the idea)
> 
> I don't understand this comment, but hopefully if I just put the above
> in my asound.conf, all should be fine, right?

Yes, it is just that programs can't specify their own AES parameters. If
everything works with your receiver, it is ok.

>> As for the preferred solution to this problem, as far as I understand,
>> that would be for the driver to have some ioctl that would provide
>> alsa-lib information about the unusual channel mapping, and alsa-lib
>> could then remap the channels using a channel remapping plugin.
> 
> Alternatively, can the fix just be hardcoded for this audio card?

Unfortunately I don't see a way of doing that, as all HDA-Intel codecs
use the same HDA-Intel.conf...
Maybe someone more knowledged knows a way, though.

-- 
Anssi Hannula