Unable to set buffer time error under certain conditions about squeezelite HOT 8 CLOSED

Forgot to add that it happens, whether I set the parameters myself or just start squeezelite with -o jack.

from squeezelite.

Is anybody reading this or is there someplace else for issues I am not aware of?

from squeezelite.

Yes I read the issues raised here, but I don't have a solution for you. It appears that the alsa devs have changed alsa-lib in recent versions. Have you asked on the alsa dev list?
I've never used jack with squeezelite and I haven't had time to setup a system with jack installed to try to reproduce your issue.

from squeezelite.

Do you continue to get the error if you run this command beforehand?
You will need to change card0 to match the device you're using.
echo 1024 > /proc/asound/card0/pcm0p/sub0/prealloc

It is a libasound issue. See this thread for full details. https://alsa-devel.alsa-project.narkive.com/vB7E5vUT/snd-pcm-hw-params-set-buffer-time-near-invalid-argument

You can also adjust the buffer time value using the squeezelite -a option to try and find a value that works.

from squeezelite.

Oh, thx for still remembering this. Just recently I updated the entire buildroot to a recent version and so I am having all newer software versions all over the place. So far it is up and running and I am in the process of debugging anyway. Give me a couple of days and I will look into it.

from squeezelite.

Had some time to set it up and test it and it is still the same behaviour with current alsa lib and alsa-plugins V1.2.2, also when echoing 1024 into the proc filesystem as you suggested.

from squeezelite.

I also read this but my system is too old: It’s an Ubuntu Studio 14.04.6 (still, sigh), using alsa-base 1.0.25 and a wild mixture of ALSA, PulseAudio and JACK2, plus lots of audio devices.

SqueezeLite (1.9.6-1198) runs fine on this box (just used squeezelite -n somename), even through the JACK bridge.

Will test again as soon as I get my system up to Ubuntu or Linux Mint 20, this may take a few weeks though.

from squeezelite.

I tested some more: Switching back to alsa 1.1.6 again solves the issue. SO I don't think it is realy related to this 9 year old bug posted by ralph-irving as it should be present in all versions throughout.

Also I can report that the problem is solveable by using different buffer values for squeezelite's output, so it would appear to be related to the old bug but then again, why would 1.1.6 solve this and break again with newer versions?

from squeezelite.