Re: [PATCH 1/2] ALSA: usb-audio: Add a volume scale quirk for AudioQuest DragonFly

[Anssi Hannula <anssi.hannula@iki.fi>]

Hi,

17.08.2015, 17:16, Takashi Iwai kirjoitti:
> On Sun, 16 Aug 2015 14:50:12 +0200,
> Anssi Hannula wrote:
>>
>> AudioQuest DragonFly DAC reports a volume control range of 0..50
>> (0x0000..0x0032) which in USB Audio means a range of 0 .. 0.2dB, which
>> is obviously incorrect and causes software using the dB information in
>> e.g. volume sliders to have a massive volume difference in 100..102%
>> range.
>>
>> The actual volume mapping seems to be neither linear volume nor linear
>> dB scale, but instead quite close to the cubic mapping e.g. alsamixer
>> uses, with a range of -53...0 dB.
>>
>> Add a quirk for DragonFly to use a custom dB mapping, based on my
>> measurements, using a 10-item range TLV (so it still fits in alsa-lib
>> MAX_TLV_RANGE_SIZE).
>>
>> Tested on AudioQuest DragonFly HW v1.2. The quirk is only applied if the
>> range is 0..50, so if this gets fixed/changed in later HW revisions it
>> will no longer be applied.
>>
>> Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi>
>> Cc: <stable@vger.kernel.org>
>> ---
>>  sound/usb/mixer_quirks.c | 37 +++++++++++++++++++++++++++++++++++++
>>  1 file changed, 37 insertions(+)
>>
>> diff --git a/sound/usb/mixer_quirks.c b/sound/usb/mixer_quirks.c
>> index 337c317ead6f..39d7f34e44e6 100644
>> --- a/sound/usb/mixer_quirks.c
>> +++ b/sound/usb/mixer_quirks.c
>> @@ -37,6 +37,7 @@
>>  #include <sound/control.h>
>>  #include <sound/hwdep.h>
>>  #include <sound/info.h>
>> +#include <sound/tlv.h>
>>  
>>  #include "usbaudio.h"
>>  #include "mixer.h"
>> @@ -1733,6 +1734,38 @@ static int snd_microii_controls_create(struct usb_mixer_interface *mixer)
>>  	return 0;
>>  }
>>  
>> +static void snd_dragonfly_quirk_db_scale(struct usb_mixer_interface *mixer)
>> +{
>> +	struct usb_mixer_elem_list *list;
>> +	struct usb_mixer_elem_info *cval;
>> +	static const int unit_id = 7;
>> +
>> +	/* approximation using 10 ranges based on output measurement on hw v1.2 */
>> +	static const DECLARE_TLV_DB_RANGE(scale,
>> +		 0,  1, TLV_DB_MINMAX_ITEM(-5300, -4970),
>> +		 2,  5, TLV_DB_MINMAX_ITEM(-4710, -4160),
>> +		 6,  7, TLV_DB_MINMAX_ITEM(-3884, -3710),
>> +		 8, 14, TLV_DB_MINMAX_ITEM(-3443, -2560),
>> +		15, 16, TLV_DB_MINMAX_ITEM(-2475, -2324),
>> +		17, 19, TLV_DB_MINMAX_ITEM(-2228, -2031),
>> +		20, 26, TLV_DB_MINMAX_ITEM(-1910, -1393),
>> +		27, 31, TLV_DB_MINMAX_ITEM(-1322, -1032),
>> +		32, 40, TLV_DB_MINMAX_ITEM(-968, -490),
>> +		41, 50, TLV_DB_MINMAX_ITEM(-441, 0),
>> +	);
>> +
>> +	for (list = mixer->id_elems[unit_id]; list; list = list->next_id_elem) {
>> +		cval = (struct usb_mixer_elem_info *)list;
>> +		if (cval->control == UAC_FU_VOLUME &&
>> +		    cval->min == 0 && cval->max == 50) {
>> +			usb_audio_info(mixer->chip, "applying DragonFly dB scale quirk\n");
>> +			list->kctl->tlv.p = scale;
>> +			list->kctl->vd[0].access |=  SNDRV_CTL_ELEM_ACCESS_TLV_READ;
>> +			list->kctl->vd[0].access &= ~SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK;
>> +		}
>> +	}
> 
> Instead of looking through the list, just hooking at
> build_feature_ctl() would be simpler in the end, I think.
> E.g. something like:
> 
> --- a/sound/usb/mixer.c
> +++ b/sound/usb/mixer.c
> @@ -1334,6 +1334,7 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
>  				SNDRV_CTL_ELEM_ACCESS_TLV_READ |
>  				SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK;
>  		}
> +		mixer_fu_apply_quirk(state->mixer, cval, unitid, kctl);
>  	}
>  
>  	range = (cval->max - cval->min) / cval->res;
> 
> ... and the quirk implementation in mixer_quirks.c like
> 
> static void snd_dragonfly_quirk_db_scale(mixer, kctl)
> {
> 	usb_audio_info(mixer->chip, "applying DragonFly dB scale quirk\n");
> 	kctl->tlv.p = scale;
> 	kctl->vd[0].access |=  SNDRV_CTL_ELEM_ACCESS_TLV_READ;
> 	list->kctl->vd[0].access &= ~SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK;	
> }
> 
> void mixer_fu_apply_quirk(mixer, cval, unitid, kctl)
> {
> 	switch (mixer->chip->usb_id) {
> 	case USB_ID(0x21b4, 0x0081): /* AudioQuest DragonFly */
> 		if (unitid == 7 && cval->min == 0 && cval->max == 50)
> 			snd_dragonfly_quirk_db_scale(mixer, kctl);
> 		break;
> 	}
> }

OK, seems like a good idea.

However, I just noticed another volume quirk for DragonFly has already
been merged since I started looking into this:
https://git.kernel.org/cgit/linux/kernel/git/tiwai/sound.git/commit/?id=2d1cb7f658fb9c3ba8f9dab8aca297d4dfdec835

It sets a fixed dB linear range map of 0...50dB via mixer_maps.c, which
doesn't seem 100% right to me. While it is much better than the
unquirked 0...0.2dB, it causes e.g. pulseaudio mixer to show the nearly
inaudible raw 0 as the "base" level. And, unless I made some silly
mistake, the volume scale is not actually linear dB AFAICS.

Yao-Wen, did you have some basis for the assumption "dB conversion
factor is 1" on DragonFly other than that it sounded approximately right?

Takashi, should I add removal of that "duplicate" quirk in the same
commit (or a separate one)? (assuming my quirk turns out to be actually
better/correct, of course)

-- 
Anssi Hannula