Very odd, and I have no idea who is the bad boy in this combination but:
Hardware:
MacBook
Moog Sub 37 connected via USB
Virus TI2 Desktop connected via USB
Software
Apple MainStage (with the Sub 37 configured as MIDI source of a keyboard)
Access Virus TI Control Plugin (AU)
MIDI Monitor
Everything is fine and the Sub 37 sends MIDI events (as seen on MIDI Monitor and in MainStage), but as soon as the Virus Control plugin is instantiated, the Sub 37 stops to send MIDI output. During instantiation of Virus Control, no specific traffic is seen on the MIDI Monitor! After the VC is completely loaded and established connection to the Virus TI2, no more events are sent from the Sub 37. Power cycling the Sub 37 restores its MIDI sending capability.
It’s rather frustrating that nothing appears on MIDI Monitor, although something must have been sent to the Sub 37 which killed its MIDI output. How could I proceed to further investigate?
Additional frustration is provided by the fact, that the problem is not 100% reproducible. Especially if trying to reproduce it, I get a 20-30% error rate.
It seems to me like a setup related issue (rather than a bug). I might be wrong, but I can’t think of any reason why the Sub 37 would stop sending midi in that case. As I can’t test it locally, maybe you’ll be able to further investigate this ?
Thank you for your answer. I will investigate, but I ran out of ideas how I can see what exactly is sent to the Sub 37 that kills its MIDI out. While trying to figure it out, I also started the Editor hoping it would show something or give me additional access to the settings on the Sub 37. The connection to the Editor was still fine and changes from the Editor were sent to and recognized by the Sub 37, but no changes on the Sub 37 were reflected on the Editor.
Any hint how I could see the MIDI messages going out to the Sub 37 aside from MIDI Monitor? Do you know of a SysEx byte combination which shuts off the MIDI output on the Sub 37?
My first guess would be that somehow the Virus control app is “stepping on” the Sub 37 USB driver within OSX CoreMIDI.
One way to test this would be to look at the DIN (5-pin) MIDI output from the Sub 37, using a separate DIN MIDI interface if you have one… or by connecting the DIN MIDI output from the Sub 37 directly to another keyboard, might be even better…
basically, set everything up as you normally would, but additionally connect the Sub 37 DIN output directly to another synth so you can see that the other synth is triggered when you play keys on the Sub 37. Fire up the Virus control app and see if it causes the problem you described… if it does, is the DIN MIDI output from the Sub 37 affected also, or no?
-Amos
PS) I realize that there is no fix suggested here… but if it can be shown that the problem is only on the USB side, then it’s probably happening on the Mac (as opposed to the Sub 37) and this could provide a direction for further efforts.
Thank you very much for taking the time to think about this issue. It seems very logical that the culprit is the Virus MIDI Driver. I will try as you suggested when I find the time, and report the bug to Access Music.
For the time being, I downgraded the Software (including MIDI Driver) and the TI OS to an earlier version which seems to have resolved most of the issues I found with their 5.1.3.0 release.
Your assumption was correct: Only the MIDI out via USB is broken, MIDI via DIN keeps working. So only USB is affected.
However, interesting is that I have to reboot the Sub 37 (hardware). Unplugging the USB is not sufficient. I wonder why… how are the MIDI drivers loaded and unloaded on the Mac OS X? Any ideas how I could help to narrow the problem down?