[MPlayer-users] a few notes on AC3 passthrough in general (was SBLive AC3...)

[Marcus Blomenkamp]

Aubin Paul wrote:

>
>>1) Some OSS guy invented the format AFMT_AC3. Soundcard drivers 
>>implementing this format should therefore adjust it's mixers and 
>>switches to produce the desired output. Unfortunately some soundcard 
>>drivers do not support this format correctly and most do not even 
>>support it at all (including ALSA).
>>
> 
> I've verified that mplayer sets the format correctly via the ioctl,
> but what is appearing to happen is that the video loses all semblance
> of sync.


Huh? You verified mplayer setting soundcard to AFMT_AC3? I recently 
posted a patch to mplayer CVS to prevent that and use AFMT_S16LE 
instead. Or was this a misunderstanding on my side?

> I actually considered picking up the C-Media boards, simply for the
> fact that they're cheap, and appear to be supported fully. I only
> didn't because of the sound.html suggesting that there are problems
> with the card; can you tell me how well it worked for you? Any sound
> glitches or noise on AC3 or pcm data?


Card works fine. However I do have a slight problem when playing 
something at 44kHz: Once or twice a day I get a short dropout. I can't 
remember this happening using my Trident card. I do not know exactly if 
  it's samplerate dependant or card dependant since the CMedia works 
fine at 48kHz but also works fine at 44KHz when connected to my 
MD-Recorder. I suppose it's a voltage misalignment because dropout count 
remarkebly increases if I toggle the output voltage.
Remark: I made a small patch to snd-card-cmipci which allows me to 
toggle the output voltage. On one setting the optical output is working 
but the electrical output woes as described. On the other setting the 
optical output is dead but the electrical connection to my receiver is 
better (with only those 1 or 2 dropouts left). This is not in alsa CVS.

> Thanks for your help; I'm hoping someone can shed some light on the
> A-V sync issues...

What sync issues exactly? I haven't experienced any (serious): A friend 
of mine produced all my AC3 Divx movies so I can't exclude them from 
beeing the point of failure, but I have to play them with delay -0.1 sec 
to get lip-sync. Also after 1.5h there is a very small drift below the 
0.1 interval which I thus cannot correct. Again: I don't know if the 
source is 100% ok. But I did not experience any _big_ drift-away yet, 
perhaps try using '-ni' or '-nobps'.

cu