Sysex fault with PolyEvolver editor

Sysex fault with PolyEvolver editor

Postby gilesjuk » 19 Sep 2015, 22:01

Hi,

I'm using Copperlan on OSX Yosemite with AL-88. When using the Soundtower Poly Evolver sound editor and its "Morph" feature the AL-88 goes funny.

This morph function must send a batch of sysex very quickly and as a result the AL-88s MIDI out light just periodically flashes and when I try to play notes the Poly Evolver doesn't receive any notes. If I press the reset button on the front of the AL-88 it then starts to work again.

This is most annoying and severely limits the usefulness of Copperlan for me.
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Re: Sysex fault with PolyEvolver editor

Postby CopperPhil » 19 Sep 2015, 22:13

Can you capture the sysex messages (ideally with timestamp information) so I can try to reproduce the problem?

Thanks
CopperPhil
 
Posts: 480
Joined: 30 Mar 2011, 15:02
Location: Brussels

Re: Sysex fault with PolyEvolver editor

Postby gilesjuk » 19 Sep 2015, 23:01

Can you tell me how I'd record the data?

I'm using a small network switch as I have two AL 8 port devices, could that be causing a problem? I guess I could bypass it.
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Re: Sysex fault with PolyEvolver editor

Postby CopperPhil » 19 Sep 2015, 23:20

No the problem is not related to your network.

I suspect a new problem in sysex transport through CopperLan. We had to tweak this feature many times in the past, each use-case bringing new issues. By "use-case", I mean "software sending sysex" and/or "destination synth". We had timing problems, packet size problem, we had to make possible to insert pauses in the middle of the sysex stream to avoid crashing old synths needing specific timings... Transporting Sysex through a high-speed network is not a problem, but handling all the sysex source & destination oddities is a nightmare. And, as far I know it is the first time we have a feedback related to PolyEvolver editor...

To record the sysex stream, you can direct it to a virtual MIDI port instead of the AL-88, then listen to it in a MIDI monitoring app (I'm using MIDI-Ox on PC, you should find something similar on Mac). Please capture the complete sequence causing the AL-88 failure, ideally with time position for each byte or block of byte (if possible in the MIDI monitoring).
CopperPhil
 
Posts: 480
Joined: 30 Mar 2011, 15:02
Location: Brussels

Re: Sysex fault with PolyEvolver editor

Postby gilesjuk » 19 Sep 2015, 23:47

Hi,

I've attached a session with me playing around with the morph feature. At times the notes wouldn't play. But eventually after waiting for the sysex to stop sending (which takes a good few seconds) it did seem to work. Perhaps monitoring the output with MIDI Monitor helped with timing?
Attachments
PolyEvolver.zip
(4.58 KiB) Downloaded 245 times
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Re: Sysex fault with PolyEvolver editor

Postby CopperPhil » 21 Sep 2015, 10:39

Well I made a small fix in the Sysex handling code. Can you try this version?

https://dl.dropboxusercontent.com/u/215 ... %29%5D.dmg
CopperPhil
 
Posts: 480
Joined: 30 Mar 2011, 15:02
Location: Brussels

Re: Sysex fault with PolyEvolver editor

Postby gilesjuk » 21 Sep 2015, 20:18

Thanks, I'll give it a go.

One thing I noticed, I swapped MIDI controller to see if my controller was acting up. I used the replacement for the test. Swapped back and noticed my original controller was sending MIDI clock data. That might have been creating issues.
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Re: Sysex fault with PolyEvolver editor

Postby gilesjuk » 21 Sep 2015, 23:46

Hi,

It seemed better, but still I end up with the notes not being received and some sysex message keeps repeating (at the end of the file, just above the final note on/offs). Once those sysex have stopped the synth plays a note on its own and then I can continue.

I might try another MIDI interface just to rule out some other cause.

I should add that I have a rack and keyboard, hence different port numbers in this file. My rack has gone haywire tonight (needs PLCC chips reseating argh). But the same issue happens with the PEK.

Had to revert to the last release of Copperlan Manager too, seemed to be causing issue with my Microwave 1.
Attachments
pek.zip
(2.81 KiB) Downloaded 208 times
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Re: Sysex fault with PolyEvolver editor

Postby CopperPhil » 22 Sep 2015, 09:56

I am very curious about all of these duplicated sysex messages... CopperLan is not able to duplicate messages, it is just transferring what it receives on the MIDI->CopperLan bridge to the destination CopperLan->MIDI bridge. So I am wondering if the MIDI content itself is correct... I want to be sure that the problem is well located somewhere in the CopperLan stack.

=> Can you check that it is working fine using a simple MIDI USB interface between the PolyEvolver editor and your synth ?
CopperPhil
 
Posts: 480
Joined: 30 Mar 2011, 15:02
Location: Brussels

Re: Sysex fault with PolyEvolver editor

Postby gilesjuk » 22 Sep 2015, 12:42

Yeah I'll give that a go, I've got a few USB MIDI interfaces.

I'm also confused about the duplicated messages :D It was more noticeable with the Microwave as it has a MIDI receive light. The only way to stop it was to power down the AL88. The MIDI monitor didn't show anything after a while.

My Macbook Pro has no ethernet port, so I'm using Thunderbolt to a Thunderbolt dock, maybe I need to take the out of the equation too.
gilesjuk
 
Posts: 12
Joined: 24 Feb 2015, 11:00

Next

Return to Bug report

cron