Chromebook (How to use and put Puppy on a Chromebook)

[vtpup]

vtpup,


Correction:
Should be /usr/share/alsa/ucm/chtmax9890

Nope, it's 98090. At least on the R11.

[bigpup]

Yes it is 98090.

I was actually pointing out that you left out the alsa.
Should be /usr/share/alsa/ucm/chtmax98090

[bigpup]

vtpup,

You getting sound OK in a browser?

I am using Pale Moon and get no sound.

Really strange.
With the browser running.
Multiple Sound Card Wizard->Test sound will not work.
Shutdown browser and the test works

Deedbeef sound works.

With all those Retrovol Mixer settings.
Probably, something not set correctly, maybe!

[vtpup]

Same problem here with Seamonkey.....

Interesting, because other programs like, mpv play movies (with sound) fine.

I thought it might have something to do with needing to specify apulse when invoking the browser, but that doesn't solve it.

I still wonder if pulseaudio is part of the problem with current mozilla based browsers. Seamonkey 2.53 definitely has switched to the recent firefox engine that requires pulseaudio.

ps, don't mess with the mixers, too much, ESPECIALLY DC Speaker Blocking (as I'm guessing that one may prevent speaker melting problems mentioned earlier).

Also thanks for the alsa correction, I edited the post.

[bigpup]

Could you try it with Pale Moon browser?

I stopped using Firefox and Seamonkey a long time now.
I am sure apulse is needed for Firefox.
Bionicpup64 8.0 has already got the apulse stuff, for the Firefox, you can get from Quickpet.

I do have Bionicpup64 8.0 updated by Quickpet->Info->Bionicpup updates.

[vtpup]

No sound: Seamonkey2.49.4, Pale Moon 28.4.0, whether or not started with apulse

Sound okay: Chromium 70.0.x, and Opera 68.0.xx


btw Firefox Seamonkey and Pale Moon are all Mozilla based browsers.

Firefox current and Seamonkey 2.52 or greater no longer support ALSA, and require pulse audio. However 2.49.4 should still work.....

For Firefox and newer Seamonkey, normally, running "apulse Firefox" (for example) solves this. But it doesn't here, possibly because we have a newer kernel than standard Bionicpup? Maybe apulse needs to be newer....?

This is definitely a specific browser related problem, since Chromium and Opera do have sound.

I also noticed that the Flashplayer installer doesn't work. I did install flash manually, and also tried the VLC player browser plugin. Neither solved the browser sound issue.

[bigpup]

Google Chromium
A Chromebook.
I wonder if there could be a inside connection, to why sound will work

Well, so far still no browser sound working for me.
Chromium or Opra.

Did you make any changes in the Retrovol Mixer?

[vtpup]

no bigpup, totally stock system and files, settings, etc (with the above listed mods of course).

I created a brand new frugal install and pupsave so that I could test my mods without anything else changed.

Have you tried opening a video file on your computer by typing the location into your browser address bar?

eg.

Code: Select all

file:///mnt/sdb1/videos/myvideo.avi

Of course pick one that has a sound track.

[vtpup]

When I play a file video in the browser, it has sound. It plays through a helper application, defaultvideo (mpv). When I go to Youtube or Vimeo videos play through THEIR player, and there is no sound.

I have a hunch that they are trying to play through Alsa card 0 device 1, instead of the intended default card 0 device 0.

If you go to Multiple sound card wizard and select the "Deep Buffer Audio" card, that's card 0 device 1. And it doesn't play sound when you test it. So I'm guessing that is what Mozilla browsers are trying to play through, when what we need is for them to use card 0 device 0.

The default 0 0 SHOULD be enforced by /etc/asound.conf:

Code: Select all

defaults.pcm.card 0
defaults.pcm.device 0
defaults.ctl.card 0

But maybe somehow the browser players are getting around that.
Maybe we can enforce it through an /root/.asoundrc file.
OR, better, but harder, find out why the Deep Buffer Audio (card 0 device 1) choice doesn't work. If it worked, and the browsers were choosing that device, we'd get sound.

Another possibility:

Maybe these video sites are looking at the current Mozilla useragent string and interpreting that the computer is equipped with pulseaudio, and altering their player's audio output method to suit? Though then, using apulse should solve that -- unless there's something wrong with apulse and our new kernel combination.

The only reason I mention that possibility is that Opera and Chromium would put out different useragent strings, and that might explain the difference.

Finally:

I suspect that if we had pulseaudio, things would just work.

[vtpup]

bigpup, wondering about the difference in behavior for yours and my chromebooks, maybe you have a slightly different 98090 chipset.

Here's the GalliumOS Baytrail git repo. Notice the 2 different UCM sets.
https://github.com/GalliumOS/galliumos- ... e/alsa/ucm
I wouldn't be surprised if the byt-max98090 version stands for baytrail. Maybe try that set out for UCM instead of the chtmax98090 version I need for my Braswell chipset..

Edit:

Actually, thinking about it this morning, bigpup, seems obvious in retrospect, the "byt" may stand for bay trail and the "cht" may stand for Braswell with cherry trail. My chromebook has the cherrytrail chipset and yours has the baytrail chipset.

Peronally, I would add both UCM's to usr/share/alsa/ucm and let the OS decide which it needs (or if it needs both).

[rcrsn51]

It's been my experience that browsers want to play their audio through the device that is declared in ALSA as "default". But that default device must be working properly. Open alsamixer for that device. For example, if it says "No controls found", there will be a problem.

A possible solution is to set up the ALSA equalizer as the default device, and set the actual working device as its slave. Browsers seem to accept the equalizer as a standard audio output.

[vtpup]

Thanks greatly for your help rcrsn51. Would you set that in /etc/asound.conf?

And could you suggest the code for doing that?

Present contents are:

Code: Select all

defaults.pcm.card 0
defaults.pcm.device 0
defaults.ctl.card 0

and when I do an

Code: Select all

alsamixer -c 0 

I get the mixer okay.

[rcrsn51]

Code: Select all

### these three lines enable the regular sound card ###

defaults.pcm.card 0
defaults.pcm.device 0
defaults.ctl.card 0

##################

ctl.equal {
 type equal;
}

pcm.plugequal {
  type equal;
  slave.pcm "plughw:0,0";
}

pcm.equal{
  type plug;
  slave.pcm plugequal;
}

### these two lines enable the equalizer ###

#pcm.!default pcm.equal
#ctl.!default ctl.equal

Comment OUT the top three lines and UNcomment the bottom two lines. If necessary, change "plughw:0,0" to match the desired device.

[vtpup]

Thanks rcrsn51, testing shortly ...

[vtpup]

I couldn't test the Acer R11 sound this afternoon because my daughter needed it for online classes. Will return to it tomorrow.

But I was able to resolve the internal microphone issue I had on the AcerC720 Chtomebook(Haswell, PEPPY).

To make the internal microphone work, I actually had to:
1.) In alsamixer: mute AND lower the levels to zero for BOTH "Internal Microphone" and "Internal Microphone Boost".
2.) In alsamixer: un-mute "Microphone" and "Microphone Boost" and raise their levels, as appropriate.

The was highly counter-intuitive and difficult to arrive at, as no combination where internal mic wasn't completely muted, AND zeroed worked.
Of course, "Capture" needed to be enabled and set, but that's to be expected.

So, now the C720 is fully functional, as far as I know, using stock Bionicpup64.

[rcrsn51]

As much as people around here love to hate Pulse Audio, I suspect that it could handle this situation better than raw ALSA.

[bigpup]

I am still here.

I too am starting to think Pulse Audio may just work better.

All those options, in alsamixer and Retrovol Mixer, seem to be provided by the HiFi.conf file in /usr/share/alsa/ucm/chtmax98090

Notice, the top few entries in the HiFi.conf, look like some kind of commands, that could be used to setup sound options.

Code: Select all

# command-line sequence to switch playback/capture
# alsaucm -c chtmax98090 set _verb HiFi set _enadev Headphone
# alsaucm -c chtmax98090 set _verb HiFi set _enadev Speakers
# alsaucm -c chtmax98090 set _verb HiFi set _enadev HeadsetMic
# alsaucm -c chtmax98090 set _verb HiFi set _enadev InternalMic

I do remember, in my earlier efforts getting sound working.
Other Linux OS's, help topics, talked about using these commands.

[bigpup]

I wouldn't be surprised if the byt-max98090 version stands for baytrail. Maybe try that set out for UCM instead of the chtmax98090 version I need for my Braswell chipset

But why would my sound device show as a chtmax98090 if it is actually a byt-max98090?
Plus, I do get some sound, just not in browsers.
Well, I will try it and see what byt-max98090 added to ucm will do

Looking at both HiFi.conf files, to see if there is any difference.

I am starting to think, HP put this Chromebook together, with whatever parts they had, at the time.

My processor is suppose to be baytrail, according to Intel web site. But my sound device is identified as chtmax98090.

[vtpup]

But why would my sound device show as a chtmax98090 if it is actually a byt-max98090?

It was just a guess, bigpup, and no evidence at all -- just thought it might offer another possibility to try.

I agree also that pulseaudio seems to solve many problems like these that I've read about on other forums. I don't know what's entailed in implementing it. Don't know much about it. I'd be happy to try it if I knew how.

Hoping to try out rcrsn51's changes as soon as I get that R11 back. I'm going to make a clean pupsave to try it, so nothing else I've done affects it adversely. Too many changes in searching for answers confuses things.

[bigpup]

I hope I did not make you feel that your suggestion was not helpful!
It was not a question about why would you suggest such a thing.
It was more of an observation of what I was seeing.

Well, I can now tell you for sure that my sound device is a chtmax98090.
Any commands dealing with byt-max98090 come up with errors.
Commands for chtmax98090 get no errors and seem to actually do something.

I think my problems are all about this HP chromebook made out of whatever parts they had at the time.
Maybe there is such a thing as a braswell motherboard and they put a bay trail processor in it.

ChromeOS did seem to have it all working. It must be possible.