MIDI input suddenly dies constantly.

manateemilitia · « **on:** May 16, 2020, 01:42:14 PM »

Anyone else have this problem? MIDI input via USB or DIN from both MIDI interfaces, my pc, and a hardware sequencer just suddenly stops working anywhere from 30 seconds to an hour after turning on. Reboots sometimes help but usually do not. Of course I'm contacting Sequential, just wondering if anyone else has had this issue.

I'm starting to think I got a lemon of a PX, I've already had to send mine in once and I can't count the number of freezes I've experienced or the number of times I've had to reinstall firmwares.

manateemilitia · « **Reply #1 on:** May 16, 2020, 02:22:49 PM »

Interestingly it's gotten worse. Now midi input quits around 5 seconds after it receives a note via USB or DIN. And it appears local keyboard control dies when the midi input dies as well.

Rimshot · « **Reply #2 on:** May 20, 2020, 12:38:12 PM »

I’m having a similar issue. When using MIDI via USB from my Push 2, the PX seems to freeze up and doesn’t respond to any MIDI note commands after a short period of time post switch on (usually within minutes). If I power off/on the PX this resolves the issue. I have switched to MIDI din and this has resolved the issue which is a workaround for now. Would appreciate it if you could share your response from the help desk before i raise a similar ticket. Thanks!

Anscorm · « **Reply #3 on:** May 21, 2020, 12:20:07 PM »

Same Here. While using MIDI Over USB with Local off and going throught my MPC Live MIDI stops working every once in a while.

manateemilitia · « **Reply #4 on:** May 21, 2020, 01:39:41 PM »

Monitoring the MIDI messages at the suggestion of Sequential support I've discovered that it quits when the parameter messages from the knobs are sent back to the PX from itself. To me, that totally sounds like a bug so I'll be updating Sequential but the fix in the mean time for me was to turn off MIDI Param Send (of course you can't record automations until this is fixed).

antongia · « **Reply #5 on:** June 21, 2020, 07:10:31 AM »

Same problem also for me. As soon as I touch any pot on the PX I lost the MIDI connection with Cubase. Now I'm gonna using the trick you suggest, but hope to have an update to solve this bug asap.
Ciao

Derek Cook · « **Reply #6 on:** June 25, 2020, 11:10:29 AM »

Quote from: manateemilitia on May 21, 2020, 01:39:41 PM

Monitoring the MIDI messages at the suggestion of Sequential support I've discovered that it quits when the parameter messages from the knobs are sent back to the PX from itself. To me, that totally sounds like a bug so I'll be updating Sequential but the fix in the mean time for me was to turn off MIDI Param Send (of course you can't record automations until this is fixed).

Hi, I have raised the same point to Sequential in my reports on SYSEX hangs when you are dumping SYSEX to the synth. I need to get back to investigating it in more detail (but too much nice weather here right now to be inside!), but I did not see the hangs when I turned off NRPNs or put them on a different MIDI point (i.e. MIDI DIN instead of USB).

I think this whole area of MIDI needs debugging.

manateemilitia · « **Reply #7 on:** July 14, 2020, 03:28:53 PM »

Still no response from Sequential on this and it has gotten worse on my unit—now even with midi param send/receive disabled, the usb midi still dies after an arbitrary amount of time when looped through both Logic Pro X and Bitwig. Makes the PX almost useless in conjunction with a DAW.

manateemilitia · « **Reply #8 on:** July 15, 2020, 12:40:29 PM »

Amazingly another update.

It's gotten even worse: No MIDI loop or anything, just external sequencing kills the input until a restart. I guess it's an opportunity to get my timing down really well and multi-track audio only but my PX is totally unable to be sequenced.

dbm · « **Reply #9 on:** July 20, 2020, 06:26:59 PM »

I think I found something!

I have also had the problem of the ProX locking up while in use in Logic Pro on a Mac. I spent some time this afternoon tracing MIDI messages to/fro using MIDI Monitor.

What I found is that when using NRPN's, there are some seemingly undocumented NRPNs being tossed around by the ProX and echoed back from the Mac. On initial startup, I can play the keyboard just fine, and perform about 2 program changes locally before the ProX locks up and no longer responds to keyboard Note-On / Note-Off events.

After locking up, the Notes continue to be transmitted to the Mac, and echoed back. But the ProX is completely silent in return.

I had Local Control disabled for Keys/Wheel, and Control Changes enabled over USB using NRPNs. I disable transmit of program changes, and disabled receipt of program changes. In Logic Pro, all the MIDI events from all sources get merged into one giant stream, and I don't like other keyboards inadvertently changing programs on the ProX.

What I found, that seems to work okay, is to *STOP USING NRPN* while in Logic Pro X. And perhaps other DAW's as well. After all, these DAWs are simply echoing back whatever they receive from the ProX. And whatever the ProX is sending becomes poison to itself after about 2 program changes.

Program Change is never sent. But when I twist the Program knob, one NRPN indicates Layer A/B Switch of 0 (layer A?), and the subsequent codes seem to be undocumented, NRPN #374 (decimal). Whatever that is, it is sent 3 times in succession, and then the ProX decides to stop responding.

Remember, the MIDI stream is simply echoing whatever the ProX sent. Logic Pro didn't just make them up. These NRPNs are sent when I twist the Program Change knob - even though I told it not to send program changes.

But if you tell the ProX to use CC instead of NRPN's then things appear to work just fine. And twisting the Program knob sends nothing to the Mac, as expected.

-------------

So, what this means, is that until I manage to construct a filter to weed out unrecognized NRPNs from the stream, we have to live with a restricted set of controllers going to the ProX. CC can only accommodate 127 possible controls. NRPN greatly expands on that. But we have to prevent the ProX from committing suicide.

dbm · « **Reply #10 on:** July 21, 2020, 07:04:43 AM »

Here is a JavaScript that can be used in a Logic Pro X Scripter Midi Plugin, ahead of the External Instrument that sees the Pro X. It passes everything except the "invalid" NRPN sequence. Using this script now allows me to run MIDI over USB with no more trouble. The Pro X remains fully functional after any number of Program Change or Bank Change operations.

// ProX Filter - elide invalid NRPN sequence, pass everything else
// DM/RAL 07/20

nrpn_state = 0;
saved_event = 0;

function HandleMIDI(event)
{
   // event.trace(); // for debugging
   if ((event instanceof ControlChange) && (event.number == 99) && (event.value == 2)) {
           nrpn_state = 1;
           saved_event = event;
   }
   else if (nrpn_state == 1) {
      if((event instanceof ControlChange) && (event.number == 98) && (event.value == 118)) {
        nrpn_state = 2;
      } else {
         nrpn_state = 0;
         saved_event.send();
         event.send();
      }
   }
   else if(nrpn_state == 2) {
      if((event instanceof ControlChange) && ((event.number == 38) || (event.number == 6))) {
         // do nothing
      } else {
         nrpn_state = 0;
         event.send();
      }
   }
   else
      event.send();
}

dbm · « **Reply #11 on:** July 21, 2020, 08:10:44 AM »

Okay... a second order problem has shown up, and it matches what the originator of this thread was talking about. I thought he was talking about the USB problem that I encountered after only 2 program changes.

But now, using USB MIDI, if I drive the USB port really hard by sweeping the Volume control to and fro, then after a while, some notes get stuck, and the USB port stops responding to any note commands. I can only be restored by power cycling the ProX. And even then, it wakes up in a funky state that can be cleared by changing the Program and back. Meanwhile, MIDI over MIDI DIN continues to function properly, even as the USB port goes deaf.

It almost seems like there is a buffer overrun in the ProX USB port manager, and it gets into a total funk on several conditions. I was watching the MIDI command stream being sent to the ProX, and no odd codes were being sent. It just seems like a buffer overrun condition.

It seems that USB in the ProX has some problems... MIDI over DIN seems to work just fine.

dbm · « **Reply #12 on:** July 21, 2020, 09:16:48 AM »

Just as an aside... It is entirely possible that the problem is really Apple's fault...

I had an experience about a year ago, where I had a MOTU 624 that acted like it was dropping its output channels. Audio was being sent out, you could see just a short 0.1s glitch of audio before it got choked off. That was on one output channel, while the other channels were completely deaf.

I went around and around with MOTU Support on this, trying different drivers, nothing worked. Then one day, I just happened to start a fresh project in Logic Pro X, and all of a sudden the 624 was working properly. There never was anything wrong with the MOTU 624. It was entirely in Apple's Logic Pro.

I haven't had that corrective experience yet, with the Prophet X. But just sayin...

Derek Cook · « **Reply #13 on:** July 21, 2020, 01:26:07 PM »

Great diagnosis. It's hard to say if it is the same problem I am experiencing when trying to send SYSEX data to the Prophet X, and the trail has gone cold (summertime is spent more outside than in when the weather is good). I need to get back to diagnosing the issue further some time......

MPCCMP · « **Reply #14 on:** August 04, 2020, 06:02:36 PM »

I am having the same issue on a Mac, I just got the board a week ago and its a real bummer. I have to constantly power cycle just to get through a session. As soon as I touch the pots I love audio while playing midi from the arranger in studio one. I sent a support ticket in and they suggested it was a midi loop....good to see im not the only one but it make me nervous that so many others have the same issue and it doesn't seem to be a known issue by support. Have you all wrote up tickets?

Derek Cook · « **Reply #15 on:** August 05, 2020, 09:25:34 AM »

Quote from: MPCCMP on August 04, 2020, 06:02:36 PM

I am having the same issue on a Mac, I just got the board a week ago and its a real bummer. I have to constantly power cycle just to get through a session. As soon as I touch the pots I love audio while playing midi from the arranger in studio one. I sent a support ticket in and they suggested it was a midi loop....good to see im not the only one but it make me nervous that so many others have the same issue and it doesn't seem to be a known issue by support. Have you all wrote up tickets?

Yes, my ticket is in. I have two to do with MIDI issues, but not sure if they all have a common cause. I do need to find some time to do some more diagnosis to help Sequential.

Kisnou · « **Reply #16 on:** September 30, 2020, 12:47:48 PM »

Hello, any news on this? I still have the same problems..

GGW · « **Reply #17 on:** October 28, 2020, 07:35:37 PM »

Same problem here. I am on Windows and have all of the PX Global parameters turned off as per the suggestions above. All I am trying to do is play MIDI data into my DAW and then edit the data in the DAW. I want to play the MIDI file back to the PX and monitor the audio as I edit. I have recorded notes, but as I edit, the PX randomly quits after some time time. I have two DAW programs. Cakewalk locks up completely without warning and loses all the work done. In Reaper, the PX stops receiving data. Reaper indicates that it is still sending MIDI but that the PX MIDI device cannot be opened. Rebooting the DAW does not help; rebooting the PX works some times; rebooting everything, including the computer seems to get it working again for another period. I bought the PX as my main controller and hardware synth so I am also waiting for some insight from Sequential.

GGW · « **Reply #18 on:** January 20, 2021, 07:31:16 PM »

I've been trying things here, but I would have hoped for a solution from Sequential by now. I hooked up both USB and DIN cables but they both perform the same. I opened MIDI-OX to have a look at what might be going on with the data stream and, when idle, the Prophet seems to be spitting out continuous Timing Clock messages. I've turned the Global Parameter - 4. MIDI Clock Mode to "Off" and the idling activity stopped. The Prophet seems to be working stable since. Can anyone confirm if this works for them?

Jr · « **Reply #19 on:** January 21, 2021, 10:57:42 AM »

Same here!
Worked fine for 3 weeks but after installing the latest firmware update I have this issue as well.. will try bypassing the midi clock..

does someone have a downgrade link for the previous firmware?

Cheers