I'm curious if anyone has ever noticed fewer xruns when starting jack from a command line than via qjackctl, because (and I'm still testing this), that appears to be the case for me.
there was a time that I had convinced myself that when qjackclt was in verbose mode and writing xrun info to the message window that the load of doing such would cause more xruns - sort of a positive feedback thing. Unfortunately I can't recall what the exact configuration was. I was recently plagued by the reoccurrence of xruns - about 1 per minute. I disabled either the "no memory lock" or "unlock memory" and they stopped immediately and I have not had one since. This was counter intuitive to me. I've gotten pretty good at only changing one thing at a time when troubleshooting and it definitely solved my issue. I did not investigate any further, as that machine is due for a re-build soon. Perhaps you could see if you notice the same behaviour - You may have already done so....
Well, I just realized that it's called "QjackCtl" because it's a qt interface, which I think uses a relatively large chunk of resources. But QjackCtl in verbose mode using up even more resources makes sense to me either way.
All QjackCtl does is write .jackdrc and run it, so for now I've decided to use QjackCtl to play with settings whenever I need to and start jack manually based on those settings.
Actually Qt doesn't use up to many resources, and QJackCTL can do much more than simply write out a configuration file. Not commenting one way or another on the apparent differences in xruns, just correcting those comments.
Your subscriptions & donations are critical help that make it possible for full-time development of Ardour to continue. Your support is critical and much appreciated.
May goal US$4500 (US$54k/yr)
Subscribers 804 US$2076/month
Support Ardour, get free upgrades: pay $1, $4, $10 or $50/month for a year: