Playback of bulk sysex dump crashing DP 9.1

[James Steele]

This is a weird one that I discovered tonight. I have a project I use just for recording bulk sysex dumps from effects devices in my guitar rig effects rack. I record these dumps in the obvious way: rec enable a track in DP, start DP recording, then initiate a bulk sysex dump from the front panel of the device. When the bulk dump is complete, stop DP.

When you then playback that track to the device, it should send the sysex back to that device and restore all the patch memory, etc. However, in DP 9.1, as soon as the playback wiper gets to the sysex dump, DP 9.1 crashes. This does not happen in DP 9.02.

A couple of other tidbits:

• The bulk dump could have been recorded in any prior version of DP. It doesn't seem to matter.
• The output of the MIDI track must be assigned to a valid MIDI destination.

Can anyone else confirm? If so, I'll perhaps write this up for the official bug topic.

[MIDI Life Crisis]

I haven't used SYSEX in many years. Just a wild guess here, but all the SYSEX info falls on one beat if I am not mistaken. The only troubleshooting/workaround I can think of at the moment is to try setting the start and stop time of the sequence to one frame (maybe even the data frame only) and hit play. Then I'd try it with just one frame later than the data time as the end time. If that still crashes I'd also toggle the clock to internal if it's not already there.

My best guess is that perhaps DP is choking on the density of the data but that is a completely arbitrary thought. I can check out a SYSEX dump here tomorrow.

[James Steele]

I haven't used SYSEX in many years. Just a wild guess here, but all the SYSEX info falls on one beat if I am not mistaken. The only troubleshooting/workaround I can think of at the moment is to try setting the start and stop time of the sequence to one frame (maybe even the data frame only) and hit play. Then I'd try it with just one frame later than the data time as the end time. If that still crashes I'd also toggle the clock to internal if it's not already there.

My best guess is that perhaps DP is choking on the density of the data but that is a completely arbitrary thought. I can check out a SYSEX dump here tomorrow.

I've reported it to MOTU. The problem is that this has NEVER happened UNTIL DP 9.1. This is sysex that was recorded into a track in real time by placing a track into record, and then initiating a bulk dump on the front on the unit. My workaround is to use 9.02 for playing back sysex dumps. I have a number of devices that require receiving these dumps in a "real time" manner. They require the sysex messages be sent back to the device, (in the case of my Rocktron MIDI Raider), with rather specific timing. I suspect it lacks much of a buffer. It can take 3-4 minutes to dump all patch memory and the same to receive it. Even then it's touch and fails sometimes.

[James Steele]

Here's a project for anyone to use to verify:

http://motunation.com/files/SysexDumpCrash.zip

Load the project file and then assign the track to any valid MDI device, either a VI or an external module, and then hit play. When DP's playback wiper reaches the sysex message DP 9.1 will crash and spontaneously quit. This will not happen in DP 8 or DP 9.02.

[waterstrum]

I'm wondering if the "sustain pedal/cc controller" bug might be related to your sysex problem.
In each case, MIDI info is sent and received in a very specific and precise time frame.
9.1 seems to have a bit of a problem with this.
Mr. Shoosh would be the expert in this realm.

I would also look at the new Pre-gen feature which has great promise.
The current implementation is well on the way, but has some issues.
Not sure if it would mess with a sysex dump, but who knows.

[waterstrum]

BTW, I opened your example project and tried it with both a VI and then a rack mount synth.
No crashing.

That being said.....
I suspect the crash you are reporting requires having the specific MIDI destination available.
Something about the handshake between DP and your specific device.

The question becomes....

How did 9.1 change that relationship?

[James Steele]

BTW, I opened your example project and tried it with both a VI and then a rack mount synth.
No crashing.

That being said.....
I suspect the crash you are reporting requires having the specific MIDI destination available.
Something about the handshake between DP and your specific device.

The question becomes....

How did 9.1 change that relationship?

Strange. Any valid MIDI target of the MIDI track will cause DP 9.1 to crash on my system. Both on the MacBook that I use for backing tracks and on my studio computer. Very odd. What Mac OS are you using? I'm on Yosemite on my MacBook and El Cap on my studio computer.

[rickorick]

I just did an OS update for my Access Virus b. I dragged it int a track set the BPM for 70, played the track back it took 15 minutes but it worked. I'm running OS 10.8.5 and DP 7.24. SYSEX

[James Steele]

I just did an OS update for my Access Virus b. I dragged it int a track set the BPM for 70, played the track back it took 15 minutes but it worked. I'm running OS 10.8.5 and DP 7.24. SYSEX

Right... this is a bug that (for me anyway) was introduced with DP 9.1. If you tried this on DP 7.24 I'm not surprised it didn't crash.

[Robert Randolph]

Here's a project for anyone to use to verify:

http://motunation.com/files/SysexDumpCrash.zip

Load the project file and then assign the track to any valid MDI device, either a VI or an external module, and then hit play. When DP's playback wiper reaches the sysex message DP 9.1 will crash and spontaneously quit. This will not happen in DP 8 or DP 9.02.

I can play back the entire project to any MIDI device on my system here without issue. OS X 10.11.6.

However, if I try to view the sysex info, DP 9.1 slows to an absolute crawl. No crashing though.

[James Steele]

Thanks for trying. Ugh... that's the worst issue of all. When it only crashes YOUR system! I sent the crashlog to MOTU so hopefully they might be helpful as to what's going on here. This project crashes DP 9.1 every time on two different systems of mine: MacBook on Yosemite and MacPro5,1 on 10.11.6. Bizarre!

[James Steele]

On the MacBook I'm having to use older MOTU MIDI Driver. I use Eventide's H9 Control app to edit patches on my ModFactor in my guitar rack. MOTU MIDI Installer 1.6.64208 is the version I have to use for H9 Control to work and establish communication via sysex with my Mod factor. If I install 1.6.64820 or the latest 1.6.68717, the H9 Control app cannot find the ModFactor and can’t establish communication. Also reported this to MOTU.

[crescentoon55]

DP 9.1 also crashed when trying to update my virus b to 4.90...

[James Steele]

DP 9.1 also crashed when trying to update my virus b to 4.90...

Were you sending a large sysex dump? Does it crash in 9.02. In my situation it's clearly something introduced in 9.1. At least it appears to be.

[Phil O]

Crashes here in 9.1 as well, James. Works fine in 9.02. You should send that file to MOTU with your TechLink.

Phil