On Fri, 23 Oct 2015, Clemens Ladisch wrote:
Ricard Wanderlof wrote:
I seem to recall reading posts not to long ago that refer to using an implicit feedback path; my question is, was this in fact implemented,
Yes, for USB Audio 2.0 devices.
meaning that a device like the M-16DX should function as expected also for playback?
There is some bug in the code that should detect such Roland devices.
I finally got my hands on an M-16DX yesterday, and was preparing myself for some initial research into this problem. Connected the device and used some DAW software (Ardour 3) to record and playback about half a minute of audio. No glitches whatsoever, neither in capture or playback.
So I thought, perhaps a DAW always opens the device for recording and playback and hence there's always a feedback channel open. So I tried just playing back a 7 minute wav file using aplay instead. Again, no glitches in the resulting audio.
More tests are needed - I only used 44.1 kHz sample rate for instance - but have I just been extremely lucky with the sample clocks so they match up precisely (if that even would work), or could the problem have magically fixed itself (ok, more realistically, as a result of something else getting fixed)?
/Ricard