Spotify - tiny gaps/pauses in the sound

@koyaanisqatsi, good to know this improved the problem! It’s in line with my testing and a few other testers also reported the issue being fixed by this PR so I’ll go ahead and merge it. About the 25 second delay, I know exactly what you mean but from my experience it’s related to the FIFO snapcast file being written faster than it’s being read for a brief period of time, so samples accumulate over there and hence the delay. Now it’s weird if this happens with the new PR because it does not use FIFO files, so definitely worth keeping an eye out for that.

@garciasteve, I just tried the DWB and it does not seem to work indeed. Not sure what happened there. In any case, I’ll merge the PR now and the changes should be in the master branch so you can test directly from there. Still interested in the results btw!

1 Like

So, long-term… I’ve been playing music on and off for several hours at a time. I connect iTunes on Windows to AirPlay.

Overall latency has settled in at about 17 seconds. That’s kinda strange. I’m going to reboot all my devices and see if that changes at all. But honestly, I only deal with it once in a great while, and it really doesn’t bother me. Playback is clean for many hours at a time on RPi 4Bs. Had a few sync issues, but they always recover after a few seconds, and it was only once per many hours of playback. RPi Zero Ws can’t handle it. They play for a while, but always have long periods of chops and skips.

I found a steal of a deal on a Phase Technologies PC Sub 10-inch forward-firing, so that has really improved my listening experience as well. :slight_smile:

Let us know how it goes @koyaanisqatsi. And enjoy that new subwoofer : )

The 17 second thing went away with the reboots. It seems to show up after a bit of up-time, like after maybe a day or two, which I don’t have right now, because I’m messing with things a lot.

The sub has been fun. It’s way superior quality to any of my other speakers. I found its permanent location in the house last night. Now for some good cables… :loud_sound:

EDIT: The old RPi B+ holds up only if I run it on the headphone jack and a custom build that eliminates all unused containers. Too many containers swamps the low IO capabilities of the B+. Running on HDMI audio is too much, because it has to shuttle all the video data to the HDMI port as well, even though there is no monitor. With nothing on the HDMI port, and using the headphone jack, it works great. I may see what happens with a DAC on that board. It also helps to shut off the vpn and logs in balenaCloud.

The Zero Ws work great on the full application, but with everything unused disabled via environment vars. Although I’m deploying the scaled down app to them anyway, because they’re in the same application as the B+.
/EDIT

I haven’t had much time this week to work on the project, so the system has several days of uptime, and it has settled in again with about a 12-second latency. After a reboot, latency is around the expected 2-3 seconds. But it slowly creeps up during periods of non-usage. I have not seen anything in the logs that could be related (I also have not looked very closely). It may be creeping up while music is playing, but I haven’t been able to witness it, even through several hours of constant playback.

This behavior exists whether I deploy the default application or my modified version that removes pulseaudio from audio output, and uses it only for the plugin inputs.

The long latency really isn’t that big a deal to me. Although it may bother others. Playback quality has been rock-solid, and that is the priority as far as I’m concerned.

Let me know if you’d like me to gather any info around this. I’ll have some time this weekend to mess around.

Cheers!
Mark-

I can second this. After 11 days of uptime I have 16-25 seconds of latency. Better than skips and pauses, but a problem if the volume is accidentally set much too high. Let me know if you guys have found any solutions to this!

Thank you, @Mularkey for the additional data point on this. I have been doing some things to try and identify the cause. But I’ve only determined so far that restarting the multiroom-server container resets the latency back to the normal ~2 seconds. But nothing else about it so far. Logs aren’t showing anything out of the ordinary (at least that I can identify). I have some suspicions. But I don’t want to mislead anyone in their own diags.

To be clear, I haven’t put any major time into figuring out the latency. I’ve done some very basic things. But the ramp-up to the holiday season has kept me busy the last couple weeks. And, I listen to music most of the day every day, and I don’t want to bring down my fleet to troubleshoot.

Let us know here if you find anything you think might help.\

EDIT: Fixed typo from multiroom-client to multiroom-server.

Cheers!
Mark-

For the high playback latency issue, I created Long latency after days of fleet uptime, so we don’t veer off topic on this thread.