[alsa-devel] [PATCH] hda: Add dmux to STAC 9228
Added a dmux to the STAC9228 cards with DMIC support. And added a STAC_DIGITAL_INPUT_SOURCE macro for repeating mixer code. --- Signed-off-by: Matthew Ranostay mranostay@embeddedalley.com
diff -r 821a4558dd47 pci/hda/patch_sigmatel.c --- a/pci/hda/patch_sigmatel.c Thu Oct 18 17:38:17 2007 +0200 +++ b/pci/hda/patch_sigmatel.c Thu Oct 18 17:35:56 2007 -0400 @@ -430,6 +430,16 @@ static struct hda_verb stac9205_core_ini {} };
+#define STAC_DIGITAL_INPUT_SOURCE(cnt) \ + { \ + .iface = SNDRV_CTL_ELEM_IFACE_MIXER, \ + .name = "Digital Input Source", \ + .count = cnt, \ + .info = stac92xx_dmux_enum_info, \ + .get = stac92xx_dmux_enum_get, \ + .put = stac92xx_dmux_enum_put,\ + } + #define STAC_INPUT_SOURCE(cnt) \ { \ .iface = SNDRV_CTL_ELEM_IFACE_MIXER, \ @@ -482,14 +492,7 @@ static struct snd_kcontrol_new stac925x_ };
static struct snd_kcontrol_new stac9205_mixer[] = { - { - .iface = SNDRV_CTL_ELEM_IFACE_MIXER, - .name = "Digital Input Source", - .count = 1, - .info = stac92xx_dmux_enum_info, - .get = stac92xx_dmux_enum_get, - .put = stac92xx_dmux_enum_put, - }, + STAC_DIGITAL_INPUT_SOURCE(1), STAC_INPUT_SOURCE(2), STAC_ANALOG_LOOPBACK(0xFE0, 0x7E0), STAC_VOLKNOB(0x24), @@ -521,6 +524,7 @@ static struct snd_kcontrol_new stac922x_
static struct snd_kcontrol_new stac927x_mixer[] = { + STAC_DIGITAL_INPUT_SOURCE(1), STAC_INPUT_SOURCE(3), STAC_VOLKNOB(0x24), STAC_ANALOG_LOOPBACK(0xFEB, 0x7EB), @@ -2773,6 +2777,7 @@ static int patch_stac927x(struct hda_cod case 0x10280209: /* STAC 9228 */ spec->dmic_nids = stac927x_dmic_nids; spec->num_dmics = STAC927X_NUM_DMICS; + spec->dmux_nid = 0x1c; break; default: spec->num_dmics = 0;
At Thu, 18 Oct 2007 17:36:32 -0400, Matthew Ranostay wrote:
Added a dmux to the STAC9228 cards with DMIC support. And added a STAC_DIGITAL_INPUT_SOURCE macro for repeating mixer code.
Signed-off-by: Matthew Ranostay mranostay@embeddedalley.com
Thanks, applied to HG tree now.
Takashi
Matthew Ranostay wrote:
Added a dmux to the STAC9228 cards with DMIC support. And added a STAC_DIGITAL_INPUT_SOURCE macro for repeating mixer code.
I'm not having any luck with this patch on my XPS M1330. I built against Ubuntu Gutsy (2.6.22 which requires some PIT_TICK_RATE futzing in pcsp.h) from current ALSA repo bits after Takashi applied it to HG. I also removed Gusty l-u-m which contains a conflicting snd-hda-intel.
Audio output works fine. The only record source that appears to work is the external microphone (front input), though it is very faint even with mux 1 cranked to the max.
Can you describe your environment and record applications? Perhaps I'm just missing some assumption.
Tim Gardner wrote:
Matthew Ranostay wrote:
Added a dmux to the STAC9228 cards with DMIC support. And added a STAC_DIGITAL_INPUT_SOURCE macro for repeating mixer code.
I'm not having any luck with this patch on my XPS M1330. I built against Ubuntu Gutsy (2.6.22 which requires some PIT_TICK_RATE futzing in pcsp.h) from current ALSA repo bits after Takashi applied it to HG. I also removed Gusty l-u-m which contains a conflicting snd-hda-intel.
Audio output works fine. The only record source that appears to work is the external microphone (front input), though it is very faint even with mux 1 cranked to the max.
Can you describe your environment and record applications? Perhaps I'm just missing some assumption.
I think the problem is that I set the dmux to ADC1MUX instead of ADC0MUX, could you test this patch to confirm? Please let me know your results.
Thanks,
Matt Ranostay
diff -r 7cf5e23f804e pci/hda/patch_sigmatel.c --- a/pci/hda/patch_sigmatel.c Fri Oct 19 08:23:00 2007 +0200 +++ b/pci/hda/patch_sigmatel.c Fri Oct 19 13:47:35 2007 -0400 @@ -2777,7 +2777,7 @@ static int patch_stac927x(struct hda_cod case 0x10280209: /* STAC 9228 */ spec->dmic_nids = stac927x_dmic_nids; spec->num_dmics = STAC927X_NUM_DMICS; - spec->dmux_nid = 0x1c; + spec->dmux_nid = 0x1b; break; default: spec->num_dmics = 0;
This doesn't seem to make a lot of difference. The external microphone is marginally better, but still not really usable (recognizable but distorted). The Gnome ALSA mixer shows 3 capture sources plus digital and 3 mux's. Only the first capture source records anything. I verified that you must have an external mic plugged in just to make sure the built-in mic wasn't recording.
rtg
Matthew Ranostay wrote:
I think the problem is that I set the dmux to ADC1MUX instead of ADC0MUX, could you test this patch to confirm? Please let me know your results.
Thanks,
Matt Ranostay
diff -r 7cf5e23f804e pci/hda/patch_sigmatel.c --- a/pci/hda/patch_sigmatel.c Fri Oct 19 08:23:00 2007 +0200 +++ b/pci/hda/patch_sigmatel.c Fri Oct 19 13:47:35 2007 -0400 @@ -2777,7 +2777,7 @@ static int patch_stac927x(struct hda_cod case 0x10280209: /* STAC 9228 */ spec->dmic_nids = stac927x_dmic_nids; spec->num_dmics = STAC927X_NUM_DMICS;
spec->dmux_nid = 0x1c;
spec->dmux_nid = 0x1b;
Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Tim Gardner wrote:
This doesn't seem to make a lot of difference. The external microphone is marginally better, but still not really usable (recognizable but distorted). The Gnome ALSA mixer shows 3 capture sources plus digital and 3 mux's. Only the first capture source records anything. I verified that you must have an external mic plugged in just to make sure the built-in mic wasn't recording.
rtg
Matthew Ranostay wrote:
I think the problem is that I set the dmux to ADC1MUX instead of ADC0MUX, could you test this patch to confirm? Please let me know your results.
Thanks,
Matt Ranostay
diff -r 7cf5e23f804e pci/hda/patch_sigmatel.c --- a/pci/hda/patch_sigmatel.c Fri Oct 19 08:23:00 2007 +0200 +++ b/pci/hda/patch_sigmatel.c Fri Oct 19 13:47:35 2007 -0400 @@ -2777,7 +2777,7 @@ static int patch_stac927x(struct hda_cod case 0x10280209: /* STAC 9228 */ spec->dmic_nids = stac927x_dmic_nids; spec->num_dmics = STAC927X_NUM_DMICS;
spec->dmux_nid = 0x1c;
spec->dmux_nid = 0x1b;
Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
Ok it is true that there are 3 mux capture sources, but as of the moment we are only using one of them, ADC0MUX(which will be the first capture source control).
The reason that you can barely hear any sound recording from a DMIC is that there is no hardware db gain, the only way you can currently get measurable results is from software gain.
Hope that answers your questions.
Thanks,
Matt Ranostay
participants (3)
-
Matthew Ranostay -
Takashi Iwai -
Tim Gardner