When playing MPV's, I've repeatedly had a problem that I've never had before (and I have 7 running Linux distros installed right now, and a year's history with this pc) that the sound seems to go awry, providing staticy, not in sync sound. This can be reset by killing the pulseaudio process.
My hardware is an onboard sound chip, I'm using device #2 in the paste below.

inxi -A
Audio:
Device-1: AMD Baffin HDMI/DP Audio [Radeon RX 550 640SP / RX 560/560X]
driver: snd_hda_intel
Device-2: AMD Family 17h HD Audio driver: snd_hda_intel
Sound Server: ALSA v: k5.1.14-1-next

Can you post the SysInfo? See https://kaosx.us/docs.
A log from right after the sound goes bad will be helpful too.
What kernel do your other installs run?

Strangely (bearing in mind I just got up and am on my first video of the day), it hasn't acted up since I killed the pulseaudio daemon yesterday (though machine was shutdown during my sleep). I have a variety of kernels ranging from 4.18 to 5.2pre. I will add the requested information if it does reoccur today (or tomorrow).

Ok, I got it again and noticed there is another commonality, so far at least, it's always occurred when playing a video from plex in firefox and/or falkon. Just to clarify the problem once triggered, remains for all sounds.
I played an entire movie this morning with mpv this morning and had no issues.Pastebin/sysinfo

I went through the journalctl logs and found nothing at all useful. The only thing that even could be pertinent is a series of messages originating with kde_powerdevil

Jul 03 08:38:44 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Enforcing inhibition from ":1.11" "firefox" with cookie 5 and reason "video-playing"
Jul 03 08:38:44 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Added change screen settings
Jul 03 08:38:44 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Added interrupt session
Jul 03 08:38:44 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Disabling DPMS due to inhibition
Jul 03 08:38:44 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Can't contact ck
Jul 03 08:38:45 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Enforcing inhibition from ":1.11" "firefox" with cookie 6 and reason "audio-playing"
Jul 03 08:38:45 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Added change screen settings
Jul 03 08:38:45 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Added interrupt session
Jul 03 08:38:45 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Can't contact ck
Jul 03 08:38:58 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Releasing inhibition with cookie  6
Jul 03 08:38:58 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Can't contact ck
Jul 03 08:38:58 kaos-desktop plasmashell[1703]: qml: applet destroyed
Jul 03 08:38:58 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Releasing inhibition with cookie  5
Jul 03 08:38:58 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Restoring DPMS features after inhibition release
Jul 03 08:38:58 kaos-desktop org_kde_powerdevil[1849]: powerdevil: Can't contact ck

So far logs & SysInfo give no clue yet to what is the cause.
Can you only reproduce with remote files? Tested playing (large file) from local source?
Reproducible with the regular kernel, which in core is now newer then linux-next? Build repo has linux-next at 5.2RC7.

And please use code tags for cli output (enclose with 3 backticks ```), done for you this time.

I tried multiple times to get the code formatting right..and it wasn't working (as in I didn't do it right)..I'll give it another try next time.
Yes, I've played two entire movies locally via mpv and experienced no issues. I'm watching one in Falkon now, so far at 18 minutes and successful. Maybe it only happens with Firefox (will try that again), earlier it triggered within a few minutes (all three times)?
I'll also try to update the kernel to the most current.

Admittedly a small sample, but, I watched another hour of shows with the regular kernel, and had no issues. I'll monitor for the next couple of days and count it solved if it does not reoccur.
On a more pleasant note, I like the attention to detail in the distribution. My first encounter with KaOS was 7/1.

-

It happened again...suggestions on how to isolate it?

  • demm replied to this.

    Danny It happened again

    With a local file?

      No, still has only happened with remote files. I can copy down what I want to watch/listen to tomorrow and try to only play local files to see if it's always firefox?

      Well I played about 4 hours of audio (including white-noise for booming in the neighborhood) and had no issues. Perhaps they were on holiday? 🙂 I also played at least 2 hours of remote video in an effort to make something happen, nothing did. There's always today though.

      6 days later

      I've been sick, so haven't been doing anything...but I'm back. Today I did again encounter remote played audio (in Firefox) causing pulseaudio corruption. I've played many hours of local and no issues.
      Edit: Interesting...I now had it manifest when Android messaging did an alert (also web based).

      Two more occurences yesterday, all still pointing at things being played in Firefox (including a Facebook messenger sound).
      I'm thinking I should install another browser (chromium based I suppose) and see if this continues with it or if FF is the culprit maybe?

      That could be right on the money for my problem (bookmarking and will follow it). Thank you Nezzie!

      Unless my understanding of those (limited) is flawed, I can't see how either. I loaded Opera today to see if things perform ok there or not.
      It's always possible that local sound is affected too and I just haven't triggered it.

      I pursued some changes to /etc/pulse/daemon.conf default values and have also (today) updated to 5.2.0 kernel and will see what happens.

      Ok, so....I had noticed something when playing videos from cmdline, the video and the audio would be detected to be out of sync and the player would try many times to get back in sync.
      As mentioned above, I did upgrade/update including kernel, but I also made changes to /etc/pulse/daemon.conf based on a recomendation in Arch Linux:
      default-sample-rate= 44100
      alternate-sample-rate=48000
      default-fragment-size-msec =5
      It's been most of three days and nothing has triggered anything odd. Tentatively think I'll call it solved.