Hi All!
I’m working on improvements to the Sonic Potions LXR drum machine and one of the things I’m trying to squish is a problem syncing the LXR to the OT.
What seems to happen for all users who set the OT as master and the LXR as slave is this: The sequencer on the LXR will run in time to the OT, but any steps programmed into the LXR sequencer have about a 25% chance of actually playing.
The weird thing is, if I run the MIDI clock from the OT through my MIDI interface (a MOTU Micro Express) instead of directly connecting them, everything suddenly becomes totally happy.
My theory is that the OT is sending out some proprietary sync signal in addition to the standard clock realtime messages, and that the LXR isn’t ignoring this soon enough. However, since my interface is filtering this anomalous message out, I can’t record it on my computer and I don’t have a data logger handy so…
Does anyone know what this weird signal might be? It would be great to know what the status byte is so I don’t have to start taking random guesses on what I need to be ignoring!
I don’t think the OT is doing anything weird or out-of-spec. Have you used a Midi Monitor program on your computer to look at what comes out of the OT?
This could be an assembly or component issue in the XLR. Did you build it yourself?
[ul]
[li]I think your interface doesn´t filter anything out unless you´ve specifically set it up to do such. So what you see in a midi monitor should be what´s actually coming out from your OT. No more, no less.[/li]
[li]The OT is sending SPP (song position pointer), which I´ve found out for myself that it was to much for my AxeFxII. Where it upon reception of SPP would deactivate the function of the input LEDs (input signal). With no obvious way to turn them back on, unless power cycling the unit. Clearly a bug in the AxeFxII. But when filtering SPP out, it became stable again.[/li]
[/ul]
Given this, it might be that your LXR isn´t to happy about SPP? Does it react to PC# (program change messages)? That could also be something that your OT is providing (i e if you are switching patterns).
Please advise of any discoveries in this quest to find the OT midi goblin! I am working with another manufacturer of midi equipment and the project is on hold until this can be solved. Elektron have no idea what could be wrong either. So frustrating. http://www.elektronauts.com/t/wtf-supwit-ot-midi-to-trigger-units/5807/49548
I think the first step here is to use a Midi Monitor app (there’s a good one for OS X with exactly that name) and to compare what comes out of the MIDI OUT port on the OT with what comes out of the MIDI OUT port of the Micro Express.
That way you can be sure whether or not the Micro Express filters out any of the MIDI messages or not, or if it maybe changes the order or timing or some of the messages.
If the output is identical, then the problem has nothing to do with the MIDI date being transmitted, but might be related to a lower-level issue.
I’ve done a fair amount of synth DIY and I’ve noticed that in some cases you can end up in a situation where some stuff you’ve built simply doesn’t work correctly with specific other MIDI gear.
So, for example, I at one point built a Shruthi that worked fine when I played it from my Juno-1, but that dropped notes when I used a ESI MIDIMATE II to sequence it from my Mac. The fix was to replace the optocoupler chip.
MIDI is a serial data protocol, so “on the wire” it’s just a rapid series of pulses. These pulses however are never perfectly square, and don’t always go up to the same voltage level. There might also be variation between different devices.
What I expect to be happening is that (your) OT sends out a MIDI signal with pulses that are slightly “slanted” or “rounded” or maybe don’t go up to the same level compared with the signal that comes out of the Micro Express, and that this trips up your XLR.
As an aside, this is actually a big issue with some of the DIY schematics for MIDI input that float around on the web, especially in the Arduino word. These use “passive” optocouplers such as the 4N28 which simply don’t work with some MIDI signals. You’ll want to use an optocoupler with amplification and nice sharp edges such as the 6N137.