volume mute at 15%

[yelo3]

pulseaudio thinks that the volume is not mute when the hardware control reaches 0% (-48 dB in alsa), so there is a small amount of percentage (0-15% in pulseaudio volume) in which the volume is mute.

Alsa (in the mailing list) says that it reports gain values accordingly to the hardware specs of the registers: -48 is the minimum achievable value in that audio chip.

[coling]

The mailing list posts you refer to are from Raymond who pretty much manages to add massive confusion to any topic. Annoyingly he's quite often got good points to make, which makes dismissing his comments out-of-hand unwise.

I'll leave Lennart to comment on the specific values, but 0 should certainly be reported as a much larger value than -48dB... it should be approaching -inf dB. But perhaps we need to change our definition of "-inf dB" to be inline iwth this -48dB if it's standard. I don't know, so will leave to Lennart to comment.

[starT_T]

>>> set-sink-volume 0 9175

D: alsa-sink.c: Requested volume: 0:  14% 1:  14%
D: alsa-sink.c: Got hardware volume: 0:  14% 1:  14%
D: alsa-sink.c: Calculated software volume: 0:  99% 1:  99% (accurate-enough=yes)

>>> list-sinks
1 sink(s) available.
  * index: 0
	name: <alsa_output.0.analog-stereo>
	driver: <module-alsa-sink.c>
	flags: HARDWARE HW_MUTE_CTRL HW_VOLUME_CTRL DECIBEL_VOLUME LATENCY FLAT_VOLUME 
	state: RUNNING
	suspend cause: 
	priority: 9009
	volume: 0:  14% 1:  14%
	        0: -51.23 dB 1: -51.23 dB
	        balance 0.00
	base volume:  10%
	             -60.00 dB
	volume steps: 65537

I tested, set the volume to 14%, -51.23 dB, I am sure that I can hear the sound. -48 dB isn't mute. But which value is good for mute?

[tanuk]

Replying to starT_T:

I tested, set the volume to 14%, -51.23 dB, I am sure that I can hear the sound.

Great.

-48 dB isn't mute.

Yep. It isn't.

But which value is good for mute?

In Pulseaudio volume scale (0-65536) 0 means mute. In decibel scale -inf dB means mute. Did that answer your question?

I'm a bit confused about your message - are you just reporting that you can't reproduce this bug? This bug is actually an alsa bug, so whether you can reproduce this or not depends on your hardware.

Hmm, since this is an alsa bug, I guess nobody has anything against me resolving this ticket as "elsewhere".

[coling]

Yes I think this is a valid closer Tanu.

For reference, I recently helped to solve this problem on my own h/w.

The error was a combination of three problems: one in the h/w itself, one in the alsa driver and one in the alsa userspace library.

• The hardware did not follow the HDA spec and muted the device when it was the slider was set to 0.
• The driver did not have a quirk defined for this.
• The library did not pay attention to the quirk when it was added.

So there are a few commits needed to fix it:

• kernel: de8c85 and a74cce (but obviously this is specific to my h/w - yours may need similar fixes if it doesn't have them already)
• alsa-lib: 2f6206