PINE64 dev

20 Apr 2024 02:39

[I] <m​egi> which is needed

PINE64 dev

20 Apr 2024 02:37

[M] <Y​oda> could you put it on your download page next to the other ?

PINE64 dev

20 Apr 2024 02:36

[M] <Y​oda> I assume thats what your pinephone-call-audio does for PP. Do you have one for PPP ?

PINE64 dev

20 Apr 2024 02:29

[I] <m​egi> that's why I don't use UCM, and instead define states of all controls of the codec based on a target use case in a declarative manner and not in the procedural way of "change these controls when transitioninig to this state"

PINE64 dev

20 Apr 2024 02:27

[I] <m​egi> it's just a big mess of undefined states

PINE64 dev

20 Apr 2024 02:27

[I] <m​egi> it doesn't even configure any of the remaining hpo mix inputs to any particular state, other than povolr

PINE64 dev

20 Apr 2024 02:25

[I] <m​egi> the one you shared leaves out a lot of controls, including many that would put random noise into HP output

PINE64 dev

20 Apr 2024 02:15

[I] <m​egi> you will not hear anything if you basically just turn off the speaker

PINE64 dev

20 Apr 2024 01:15

[M] <p​iggz> megi: i found that the control "HP R Playback Switch" seems responsible for the noise when trying to use the speaker ... the UCM turns this ON when activating the speaker, turning it off removes the noise .. though i still dont hear from the speaker

PINE64 dev

17 Apr 2024 00:16

[M] <N​ewbyte> megi: thanks, so the strange name is intentional?

PINE64 dev

17 Apr 2024 00:10

[M] <p​iggz> megi: these are the things that happen when enabling the speaker https://github.com/sailfish-on-dontbeevil/droid-config-pinephone/blob/pro/sparse/usr/share/alsa/ucm2/PinePhonePro/VoiceCall.conf#L178

PINE64 dev

16 Apr 2024 23:59

[I] <m​egi> it must be from somewhere else, it doesn't make sense to be affected by speaker on/off

PINE64 dev

16 Apr 2024 23:58

[M] <p​iggz> ok, keep it in mind for when you get a chance ... the PPP speaker end is fine, but the noise from the mic is wful

PINE64 dev

16 Apr 2024 23:58

[I] <m​egi> Newbyte: jack is created by simple-sound-card now, not by codec driver

PINE64 dev

16 Apr 2024 23:54

[M] <p​iggz> (from the perspective of the other caller)

PINE64 dev

20 Apr 2024 02:39

[I] <m​egi> not yet, I need to finish the testing and change the digital mic path to go via the second adc, because there's no way to control the DMIC gain independently from other analog inputs

PINE64 dev

20 Apr 2024 02:36

[I] <m​egi> yes

PINE64 dev

20 Apr 2024 02:35

[I] <m​egi> when you have all controls organized in a logical manner from input to output, and the target use cases mapped to control values via simple boolean expressions, it gets much easier to manage the complexity of routing all this: https://megous.com/dl/tmp/6415b4c94d55bc4b.png to satisfy various use cases

PINE64 dev

20 Apr 2024 02:27

[I] <m​egi> noise can plauysibly can be comming in via 10 different paths with that UCM :D

PINE64 dev

20 Apr 2024 02:27

[I] <m​egi> hpovolr

PINE64 dev

20 Apr 2024 02:26

[I] <m​egi> it has recmixr input on out mixr enabled, it doesn't configure any other inputs of out mixr to any particular state, it doesn't configure any inputs to rec mixr to any particular state, except one of them, etc. etc.

PINE64 dev

20 Apr 2024 02:23

[I] <m​egi> UCM way of setting things is meh... you just can't know how the audio will actually be routed when using it

PINE64 dev

20 Apr 2024 02:14

[I] <m​egi> that's what turns on/off the final part of the audio path to the speaker :)

PINE64 dev

17 Apr 2024 02:19

[I] <m​egi> more unfortunate, than unintentional, but also not easily avoidable

PINE64 dev

17 Apr 2024 00:11

[M] <p​iggz> maybe i should test with alsamixer, with audio on the earpice, and just toggle the bits to enable the speaker

PINE64 dev

17 Apr 2024 00:00

[I] <m​egi> the mic has a digital output signal

PINE64 dev

16 Apr 2024 23:58

[I] <m​egi> there's no noise from the mic in my tests

PINE64 dev

16 Apr 2024 23:58

[I] <m​egi> so the names are different

PINE64 dev

16 Apr 2024 23:55

[I] <m​egi> I have not tried that yet, last I worked on this I did optimizations for main codec I2S -> modem codec I2S paths

PINE64 dev

16 Apr 2024 23:54

[I] <m​egi> except PPP has two codecs and is much more complicated