trinity-devel@lists.pearsoncomputing.net

Message: previous - next
Month: February 2018

Re: Re: kmix, pulseaudio and bluetooth

From: deloptes <deloptes@...>
Date: Wed, 28 Feb 2018 18:59:51 +0100
Luke Kenneth Casson Leighton wrote:

>> what do you mean? I installed pulse and let it take control. alsamixer
>> works just fine. When alsamixer is invoked without argument, it shows
>> pulseaudio as audiocard.
> 
> ... so pulseaudio is configured to "take over" ALSA, and to offer an
> alternative ALSA interface.
> 
>> When I run it with -c 0 it shows the audiocard as when pulse
>> was not installed
> 
> ok yes, so that tells us that pulseaudio is trying to "take over"
> ALSA, then offer a 2nd ALSA "gateway" interface.
> 
> 

yes it becomes a layer above alsa

>> when killing PA, kmix also goes at 100% CPU - need to quit kmix. When I
>> press volume key kmix starts again
> 
> interesting.
> 
> 
>> Well, I appreciate your opinion,
> 
> ... experience.  pulseaudio is involved in most of the audio f***-ups
> i've been forced to deal with.
> 
> 

Yes, forsome time in the beggining there were issues and might be there are
issues now as well, but there are many advantages, which make it inevitable
evil. BT and ofono are just few examples.

>> but there must be a solution, so that kmix
>> does not go 100% CPU.
> 
> it's the first time i've ever heard of kmix doing that... but then i
> never ran it with pulseaudio.
> 
> what happens if you run it without pulseaudio?  is it fine then?
> 

well I stated that problem occurs when something changes in pulse audio or
pulseaudio is shut down. What do you expect if kmix runs without
pulseaudio?!

> also, have you ever installed jackd (and the jackd pulseaudio
> plugin).  i now run with jackd->pulseaudio because i have a MIDI
> application that i occasionally run... but also it turns out that
> applications are a bit more stable going through jackd.
> 

might be 10y+ since last time - as stated above I don't think it will fit
(A2DP, HSP/HFP etc). 

> the advantage is that f****g pulseaudio gets the f*** off the damn
> ALSA interface (because it detects that jackd is running and managing
> it), and jackd does a hell of a lot better job of directing audio
> through to ALSA.

to me it sounds like hack a day or two task. I just want to have working BT,
besides I can't avoid PA in other context, so it is pointless to try to
eliminate PA.

I think kmix needs to be updated somehow to be able to interact properly
with PA. I need to have a closer look at the code probably.

regards