Re: ALC883 recording troubles...

[Takashi Iwai <tiwai@suse.de>]

At Thu, 12 Jun 2008 15:58:30 +0100,
Daniel J Blueman wrote:
> 
> On Thu, Jun 12, 2008 at 3:23 PM, Takashi Iwai <tiwai@suse.de> wrote:
> > At Thu, 12 Jun 2008 16:18:14 +0200,
> > Vegard Nossum wrote:
> >>
> >> On Thu, Jun 12, 2008 at 4:02 PM, Takashi Iwai <tiwai@suse.de> wrote:
> >> > At Thu, 12 Jun 2008 15:55:45 +0200,
> >> > Vegard Nossum wrote:
> >> >>
> >> >> On Thu, Jun 12, 2008 at 3:52 PM, Takashi Iwai <tiwai@suse.de> wrote:
> >> >> > At Thu, 12 Jun 2008 15:48:32 +0200,
> >> >> > Vegard Nossum wrote:
> >> >> >>
> >> >> >> The external mic:
> >> >> >>
> >> >> >> Node 0x18 [Pin Complex] wcaps 0x40018f: Stereo Amp-In Amp-Out
> >> >> >>   Amp-In caps: ofs=0x00, nsteps=0x02, stepsize=0x4f, mute=0
> >> >> >>   Amp-In vals:  [0x00 0x00]
> >> >> >
> >> >> > The input volume is zero.  Try to raise "Mic Boost" (if present)?
> >> >>
> >> >> They're already all up.
> >> >>
> >> >> Line In Boost = 100
> >> >> Mic Boost = 100
> >> >> Internal Mic Boost = 100
> >> >>
> >> >> Right after changing all these to max, I've read the value from
> >> >> card0/codec#0 again:
> >> >>
> >> >> Node 0x18 [Pin Complex] wcaps 0x40018f: Stereo Amp-In Amp-Out
> >> >>   Amp-In caps: ofs=0x00, nsteps=0x02, stepsize=0x4f, mute=0
> >> >>   Amp-In vals:  [0x00 0x00]
> >> >>
> >> >> There doesn't seem to be any difference. Even when changing them all
> >> >> to 0 in alsamixer, this output stays the same.
> >> >
> >> > Weird.  What happens if you execute the verb directly via hda-verb
> >> > program like below?
> >> >        hda-verb /dev/snd/hwC0D0 0x18 SET_AMP 0x7002
> >> >
> >>
> >> Yep, this has the same effect has changing the "Mic Boost" control in
> >> the Playback section of alsamixer.
> >>
> >> What I discovered now, though, is that I get a much wider range of
> >> values from the external mic (so playback of the captured data
> >> actually sounds correct). It actually works, but there is still a very
> >> strong DC offset that seems dependent on the level of the "Capture"
> >> control. This offset still makes it unusable for recording, since even
> >> slightly louder noises will clip the samples.
> >>
> >> It also seems that this control is reset each time I start Audacity
> >> (even though the "Mic boost" level in alsamixer doesn't change). But I
> >> guess this unrelated to the DC offset issue.
> >>
> >> Changing the voltage level now (0x21, 0x22, 0x24, with Mic Boost at
> >> 0x02) doesn't change the DC offset, but has the same effect as before
> >> (the values fluctuate, but eventually stabilize at the same level).
> >
> > Hmm, then it's all what I can guess right now.
> > There might be some vendor-specific settings I don't know of.
> 
> Maybe. I can readily dump the PCI config space of the Intel HD-audio
> controller in Windows and list differences; the datasheet from Intel
> allows us to decode the registers. Will this help us any? Of course,
> we really need to dump the target behind the HD-bus, but that's
> another matter...

I don't think it's an issue of PCI register but rather codec setups.
PCI registers cannot influence on an analog thing like DC offset.
So, we need to peek the HD-audio codec registers if really doing so.


Takashi