So the story seems to be this: ALSA stream reconfiguration (the sort of thing done by JACK or by an application running on top of ALSA) *while the device is running* will result in random re-arrangement of the input channel mapping. In particular, JACK does this when an xrun happens (also when dynamically changing the buffer size).
What the MOTU web app shows as a routing to "To Computer N" can end up sort of anywhere in the full range of possibilities. Result: it looks as if no signal is flowing through to the CPU, but it is ... just on the wrong channel(s).
Restarting the device at the ALSA level will change it again, and again ... until it is finally fixed.
Tentative diagnostic: ALSA driver bug.
On Tue, Aug 8, 2017 at 3:40 PM, Anders Hellquist <lau@xxxxxxxxxxxxx> wrote:
Could be but the device remembers the sample rate and everything else between bootups and the display is showing 48k before I hit start in cadence (or qjackctl) so from my point of view, sample rate change does not occur..I don't think waiting before launching jack helps but I can try that and see if it helps./AndersOn Aug 8, 2017 9:33 PM, "Paul Davis" <paul@xxxxxxxxxxxxxxxxxxxxx> wrote:That seems to be a function of the device taking a noticeably long time to switch sample rates and be "ready for action". If JACK changes the device state as a part of the startup, it takes a while before it is actually ready.On Tue, Aug 8, 2017 at 2:57 PM, Anders Hellquist <lau@xxxxxxxxxxxxx> wrote:/AndersI hope you can figure out whats causing the unreliable behaviour on your end.Ok, the only quirk with my setup that others have experienced too according to Ardour forum threads is that I often need two to four attempts to start jackd (actually using cadence in kxstudio to starti jackd) but when successfully started, it has never failed.Sometimes it starts on first attempt and sometimes I need more than 5 but usually on second or third.2017-08-08 20:18 GMT+02:00 Paul Davis <paul@xxxxxxxxxxxxxxxxxxxxx>:I got mine to work (kernel 4.9) but it seems a little unreliable and may have needed to be used on a MacOS box first. Still trying to figure out what causes the lack of signal at the CPU. Certainly not caused by config changes on the MOTU itself.On Tue, Aug 8, 2017 at 1:27 PM, Anders Hellquist <lau@xxxxxxxxxxxxx> wrote:______________________________/Anderscat /proc/asound/versionHi again and sorry for the delayuname -a
Linux 3.13.0-126-lowlatency #175-Ubuntu SMP PREEMPT Thu Jul 20 19:13:12 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
Advanced Linux Sound Architecture Driver Version k3.13.0-126-lowlatency.
_________________
Linux-audio-user mailing list
Linux-audio-user@lists.linuxaudio.org
https://lists.linuxaudio.org/listinfo/linux-audio-user
_______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx https://lists.linuxaudio.org/listinfo/linux-audio-user
