On 26/01/2023 11:09, Peter P. wrote:
* Fons Adriaensen <fons@xxxxxxxxxxxxxx> [2023-01-26 11:04]:
On Thu, Jan 26, 2023 at 10:47:14AM +0100, Michael Jarosch wrote:
Am 26.01.23 um 10:03 schrieb Peter P.:
Can anyone comment on the
usefulness or experiences with changing it (where would one change it?)?
Sounds like a good idea.
I'm not so sure. If the audio processing is done in real-time
threads and these threads do not interact with X in any way
that could block them (as things should be), then this won't
help.
The only way X could interrupt audio in that case is by having
real-time threads as well, and the nice value won't affect
those.
Thanks Fons,
I my case I am trying to optimize Pd, which has a GUI that is somehow
interlocked with its audio thread.
Not specific to the subjet line but...
Is/are your patch(es) 'gui-heavy', which in Pd would be e.g. lots of
bangs or changing number-boxes? That in my experience has been cause of
audio drop-outs. Also make sure to run pd with the -rt flag and if you
are on a laptop make sure to disable WiFi.
Knowing more about the patch(es) is always very useful when trying to
debug Pd IMHO :-)
Lorenzo
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@xxxxxxxxxxxxxxxxxxxx
https://lists.linuxaudio.org/listinfo/linux-audio-user
