Takashi Iwai wrote:
> At Mon, 30 Nov 2009 10:51:17 +0100,
> Yes, I meant to add some codes to the plugin :)
> It's easy to write the code to send silence to the side speakers
> except for 7.1.
Your right. I just did that. If somebody's interested here's how:
Go into alsa-plugins' mix/pcm_upmix.c.
In SND_PCM_PLUGIN_DEFINE_FUNC(upmix) add ' && channels != 8' to
if (channels != 4 && channels != 6 && channels != 0) {
so the plugin will accept the channels configuration parameter to be 8.
You also have to replace the 6 with 8 within
snd_pcm_extplug_set_param_minmax(&mix->ext, SND_PCM_EXTPLUG_HW_CHANNELS,
1, 6);
In upmix_init add an additional stype value for 8 channels config:
switch (ext->slave_channels) {
case 6:
stype = 1;
break;
case 8:
stype = 2;
break;
default:
stype = 0;
}
and allow ctype to be up to 7:
if (ctype < 0 || ctype > 7) {
Now you can extend the do_upmix matrix, which determines what function
to call on what input-output situation. Mine looks like this:
static const upmixer_t do_upmix[8][3] = {
{ upmix_1_to_40, upmix_1_to_51, upmix_1_to_71 },
{ upmix_2_to_40, upmix_2_to_51, upmix_2_to_71 },
{ upmix_3_to_40, upmix_3_to_51, upmix_3_to_51 },
{ upmix_4_to_40, upmix_4_to_51, upmix_4_to_51 },
{ upmix_4_to_40, upmix_5_to_51, upmix_5_to_51 },
{ upmix_4_to_40, upmix_6_to_51, upmix_6_to_51 },
{ upmix_4_to_40, upmix_6_to_51, upmix_6_to_51 },
{ upmix_4_to_40, upmix_6_to_51, upmix_8_to_71 },
};
The third column is for stype 2 (8 channels output), the two additional
rows are for ctypes 6 and 7 (7 and 8 channels input).
Explanation:
If there is 1 channel input, a modified copy of upmix_1_to_51 is used
with i < 8 instead of 6, so the source gets copied to the additional two
channels as well.
If there are 2 channels, a modified copy of upmix_2_to_51 is used. It
includes the additional line
snd_pcm_areas_copy(dst_areas + 6, dst_offset, src_areas, src_offset, 2,
size, SND_PCM_FORMAT_S16);
that copies front left and right to side left and right.
I've just ignored channel counts 3 and 7 since they don't make too much
sense to me.
If the source has 4 to 6 channels, the side speakers aren't used, so the
signal sounds like it is intended to.
And finally if the source has got 8 channels, a copy of upmix_6_to_51 is
used, in which param 6 is replaced with 8.
Simple thing and works great
Thanks for your tip, Takashi
------------------------------------------------------------------------------
Join us December 9, 2009 for the Red Hat Virtual Experience,
a free event focused on virtualization and cloud computing.
Attend in-depth sessions from your desk. Your couch. Anywhere.
http://p.sf.net/sfu/redhat-sfdev2dev
_______________________________________________
Alsa-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-user
