[alsa-devel] Nvidia hdmi problems

[Takashi Iwai]

At Fri, 24 Oct 2008 09:32:43 +0200,
Nir Tzachar wrote:
> 
> On Fri, Oct 24, 2008 at 9:27 AM, Takashi Iwai <tiwai at suse.de> wrote:
> > At Thu, 23 Oct 2008 22:08:15 +0200,
> > Nir Tzachar wrote:
> >>
> >> On Thu, Oct 23, 2008 at 11:31 PM, Takashi Iwai <tiwai at suse.de> wrote:
> >> > At Thu, 23 Oct 2008 20:34:30 +0200,
> >> > Nir Tzachar wrote:
> >> >>
> >> >> On Thu, Oct 23, 2008 at 10:04 AM, Takashi Iwai <tiwai at suse.de> wrote:
> >> >> > At Wed, 22 Oct 2008 19:09:41 +0200,
> >> >> > Nir Tzachar wrote:
> >> >> >>
> >> >> >> On Wed, Oct 22, 2008 at 8:15 AM, Takashi Iwai <tiwai at suse.de> wrote:
> >> >> >> > At Wed, 22 Oct 2008 06:43:06 +0200,
> >> >> >> > Nir Tzachar wrote:
> >> >> >> >>
> >> >> >> >> On Wed, Oct 22, 2008 at 4:42 AM, Wei Ni <wni at nvidia.com> wrote:
> >> >> >> >> > Hi, Takshi
> >> >> >> >> > We have no official plan to support this nforce 630i HDMI audio in linux
> >> >> >> >> > now,
> >> >> >> >> > and we didn't do any test on it yet. So we can't provide the official
> >> >> >> >> > patch for this addition.
> >> >> >> >> >
> >> >> >> >> > I think if someone interest in it and do some test, he can generate
> >> >> >> >> > patch and submit it.
> >> >> >> >>
> >> >> >> >> As I am interested, and willing o do the tests, I would be happy to
> >> >> >> >> know which tests are needed.
> >> >> >> >
> >> >> >> > If you can play both audio streams and non-audio streams over the device
> >> >> >> > by some applications, it's basically confirmed to work :)
> >> >> >>
> >> >> >> What do you mean by non-audio stream?
> >> >> >
> >> >> > The passthrough of AC3 stream or so.
> >> >> >
> >> >> >> Can you give specific commands
> >> >> >> which should work?
> >> >> >
> >> >> > I usually test using ac3dec program included in alsa-tools.
> >> >>
> >> >> Tried  ac3decode, and got a perfect sound out. I am using the
> >> >> following asound.conf (if it is relevant):
> >> >
> >> > Well, I guess this won't be a non-audio mode.
> >> > Try to get the latest alsa-tools git tree.  I changed ac3dec and
> >> > now it has -H option.  Specify this together with -C,
> >> >
> >> >        % ac3dec -C -H foo.ac3
> >> >
> >> > This will be the AC3 passthru testing.
> 
> This is the only one that works:
> 
> ./ac3dec -R --hdmi /tmp/ac3_the_other_side_48khz.ac3
> 5.1 Mode 48.0 KHz 640 kbps Complete Main Audio Service
> Using PCM device 'plug:hdmi:{AES0 0x0 AES1 0x82 AES2 0x0 AES3 0x2}'
> 
> -P and -C give me a strange mewling..
> 
> BTW, -H does not work, only --hdmi.

The patch below should fix the missing -H option.

With the option -R, ac3dec itself decodes the AC3 stream, and send
it as PCM.  To make your receiver decoding AC3, -C/-P is needed.

Can you show alsa-info.sh output (with --no-upload option)?
If you have two "IEC958 Playback Default" controls with index 0 and 1,
the latter is likely the one for HDMI.

Try to adjust it manually via iecset program in alsa-utils
	% iecset -c0 -n1 pro off audio off
then run ac3dec
	% ac3dec -H -C foo.ac3


thanks,

Takashi

diff --git a/ac3dec/ac3dec.c b/ac3dec/ac3dec.c
index c45148c..293a95c 100644
--- a/ac3dec/ac3dec.c
+++ b/ac3dec/ac3dec.c
@@ -134,7 +134,7 @@ int main(int argc,char *argv[])
 	while (1) {
 		int c;
 
-		if ((c = getopt_long(argc, argv, "hvc:D:46CPRZ:q", long_option, NULL)) < 0)
+		if ((c = getopt_long(argc, argv, "hvc:D:46HCPRZ:q", long_option, NULL)) < 0)
 			break;
 		switch (c) {
 		case 'h':