From: Takashi Iwai Subject: Re: [PATCH 2/3] ALSA: HDA: Rename "Front" to "Master Front" on ALC861VD and ALC887 Date: Tue, 23 Nov 2010 09:05:09 +0100 Message-ID: References: <4CE64381.4070003@canonical.com> <4CEA2D06.4010205@canonical.com> <4CEA856E.5070209@canonical.com> Mime-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka") Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mx1.suse.de (cantor.suse.de [195.135.220.2]) by alsa0.perex.cz (Postfix) with ESMTP id 08181243D8 for ; Tue, 23 Nov 2010 09:05:10 +0100 (CET) In-Reply-To: <4CEA856E.5070209@canonical.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: alsa-devel-bounces@alsa-project.org Errors-To: alsa-devel-bounces@alsa-project.org To: David Henningsson Cc: ALSA Development Mailing List List-Id: alsa-devel@alsa-project.org At Mon, 22 Nov 2010 15:59:58 +0100, David Henningsson wrote: > > On 2010-11-22 09:51, Takashi Iwai wrote: > > At Mon, 22 Nov 2010 09:42:46 +0100, > > David Henningsson wrote: > >> > >> On 2010-11-22 07:50, Takashi Iwai wrote: > >>> At Fri, 19 Nov 2010 10:29:37 +0100, > >>> David Henningsson wrote: > >>>> > >>>> Since "Front" controls headphones as well on these two codecs, > >>>> they should be labeled "Master Front" in order to be handled correct > >>>> by PulseAudio. > >>> > >>> Well, "Front" doesn't mean the physical output but rather the output > >>> channels. So, it doesn't conflict with the fact that this control > >>> actually changes all outputs, as long as the output jacks are of front > >>> channels. In that sense, even "Surround" is also a sort of Master > >>> control. > >>> > >>> That being said, I'm not happy with this rename. If any, we should > >>> try to get rid of unnecessary front volume control. But, just because > >>> PA gets confused doesn't sound like a right reason. > >>> > >>> Could you give the alsa-info.sh output for checking what mixer elements > >>> are created? > >> > >> As you know PulseAudio needs consistent naming of volume controls, and > >> this is not expressed clearly enough in the docs, so, what would you > >> call a volume control that controls a front-line-out, but does not > >> control a front-speaker or a front-headphone? > > > > What do you mean "front" in this context? Is it the front 2 channels > > out of 4.0 or 5.1 output? Or do you mean the physical location of > > the output? Front/Surround/Center/LFE are the former case. They are > > controls independent from the physical output locations (per > > definition). > > > > I guess this confusion came up because of the mix-up with front > > channel and front panel. We've used "Front Mic" in some places, for > > example, which increases the confusion. > > Good point. I think the reasoning so far is to refer to channels for > output and location for input. > > Anyway, what I meant was "Front" channel (not location), so let me > rephrase my question into two: > > 1) If I find a volume control (or mute switch) labeled "Front", should I > assume that it controls, or does not control, headphone output? How > about internal speakers? Yes, this should be applied to all outputs in general. (There might be some misuses, though...) > 2) Assume I have a volume control (or mute switch) which controls only > the front line-out, not headphones, and not speakers. How should I name it? A good question. We should have a unified prefix for the location. IIRC, "Front Line-Out" or "Front XXX" are used indeed, so far. > >> PulseAudio assumes that if > >> a control is named "Front", it does just that. So that's why I would > >> prefer a rename to "Master Front". > >> > >> As for alsa-info output, I believe anything in hda-emu/codecs/alc861vd-* > >> would do, as they have nsteps=0 on NIDs 0x0c - 0x0f. > > > > Yeah, the output with your first patch would be helpful. > > Unfortunately I don't have access to the actual hardware at the moment, > so I'm attaching a codec dump from the hardware before the patch. > Running it against current HEAD, with today's patch from Kailang in, > probably makes hda-emu take the wrong code path (since hda-emu can't > read proc coefs). Indeed, something is still fishy there. OK, we need to fix it again... thanks, Takashi