[Intel,HDA,ALC3226] Too much white noise from microphone

Commit Message

Shawn Starr Feb. 6, 2015, 3:22 p.m. UTC

On Friday, February 06, 2015 04:48:09 PM Alexander E. Patrakov wrote:
> 05.02.2015 02:13, Shawn Starr wrote:
> > Hello ALSA devs,
> > 
> > I spoke to patrakov on Freenode from the #Alsa IRC channel and they
> > mentioned that the noise background I am experiencing is not normal. I
> > tried two different headsets, both headsets have two separate jacks, the
> > same result.
> > 
> > I will test any patches/quirks for kernel as needed.
> > 
> > Here is the alsa-info:
> > 
> > http://www.alsa-project.org/db/?f=29d3c45e2d66c58be03a10bd0186067a9d735a11
> 
> I am not an expert on Intel HDA cards, but... some microphones are picky
> about the reference voltage supplied by the soundcard. You can change
> this temporarily, for debugging purposes, with hda-analyzer. See general
> instructions at
> 
> http://www.alsa-project.org/main/index.php/HDA_Analyzer
> 
> Note: it needs the "snd-hwdep" module loaded, and needs to be run as root.
> 
> In your case, you need to select Node[0x18] PIN in the left pane, and,
> at the bottom of the right pane, select 50 or 100 instead of 80 in the
> VREF box. See https://imgur.com/F84raSN for a screenshot based on your
> alsa-info.
> 
> If this doesn't help, just click around randomly. Maybe you'll find the
> magic checkbox that makes the noise go away and the voice stay :) - if
> that happens, please email the resulting diff file that hda-analyzer
> creates.

Hello Alexander,

Indeed, adjusting VREF to 50 cuts the background noise noticeably, it's still 
there but it is lesser.

Resulting 10 second audio with difference: http://turl.ca/accwc 
Old audio: http://turl.ca/zrhbcs

Both using same Logitech headset for test.

Thanks,
Shawn


-- CUT --

Diff for codec 1/0 (0x10ec0292):

Comments

Alexander Patrakov Feb. 6, 2015, 3:34 p.m. UTC | #1

06.02.2015 20:22, Shawn Starr wrote:
>>> http://www.alsa-project.org/db/?f=29d3c45e2d66c58be03a10bd0186067a9d735a11

> Hello Alexander,
>
> Indeed, adjusting VREF to 50 cuts the background noise noticeably, it's still
> there but it is lesser.

The voice also becomes more quiet, but the SNR indeed improves. Although 
I'd still consider the situation as too noisy.

So, the original request stands.

Patch

--- 
+++ 
@@ -189,27 +189,27 @@ 
   Unsolicited: tag=0x02, enabled=1
   Power: setting=D0, actual=D0
   Connection: 2
      0x0c* 0x0d
 Node 0x17 [Vendor Defined Widget] wcaps 0xf00000: Mono
 Node 0x18 [Pin Complex] wcaps 0x40058f: Stereo Amp-In Amp-Out
   Control: name="Front Mic Boost Volume", index=0, device=0
     ControlAmp: chs=3, dir=1, idx=0, ofs=0
-  Control: name="Front Mic Jack", index=0, device=0
+  Control: iface="card", name="Front Mic Jack", index=0, device=0
   Amp-In caps: ofs=0x00, nsteps=0x03, stepsize=0x27, mute=0
   Amp-In vals: [0x01 0x01]
   Amp-Out caps: ofs=0x00, nsteps=0x00, stepsize=0x00, mute=1
   Amp-Out vals: [0x80 0x80]
   Pincap 0x00003734: IN OUT Detect
     Vref caps: HIZ 50 GRD 80 100
   Pin Default 0x02a19031: [Jack] Mic at Ext Front
     Conn = 1/8, Color = Pink
     DefAssociation = 0x3, Sequence = 0x1
-  Pin-ctls: 0x24: IN VREF_80
+  Pin-ctls: 0x21: IN VREF_50
   Unsolicited: tag=0x03, enabled=1
   Power: setting=D0, actual=D0
   Connection: 2
      0x0c* 0x0d
 Node 0x19 [Pin Complex] wcaps 0x40048b: Stereo Amp-In
   Control: name="Dock Mic Boost Volume", index=0, device=0
     ControlAmp: chs=3, dir=1, idx=0, ofs=0
   Control: name="Dock Mic Jack", index=0, device=