USB MIDI jittery

[Lunatic Sound]

I would like to ask about this once more:

How stable and tight is your Brute's USB MIDI, when triggering the MatrixBrute via USB? For me it is really wobbly. I don't have these problems with other synths, and the Brute is on its own  USB port.

[standingwave]

Mine is VERY jittery syncing with midiclock via USB or MIDI din. It's not even usable.

[Lunatic Sound]

Interesting. Thank you for answering.

I will investigate more torrow about stability using the Din MIDI In.

What is your clock source, when you encounter those problems?

[standingwave]

Protools 12 through focusrite 18i20 DIN port, Novation slmk2 port, or USB direct to Matrixbrute. I've plugged direct to the CPU (no hub) with the same result.
 My MIDIclock has a certain amount of Jitter across the whole system, but with the Matrixbrute it shows the worst. In general this is something that can be improved as a whole in my studio; I'm interested to see what you and others are doing to improve MIDIclock stability.
 Cheers..

[standingwave]

I just ordered a Expert Sleepers USAMO. When I have sample accurate and jitter free clock output, I'll report back.
 Cheers.

[standingwave]

I just got the USAMO up and running in my system to send clock. It has tightened up the sync considerably, but still not perfect. On my Prophet 12 it stays at a steady 120bpm while the matrixbrute fluctuates between 119.8 and 120.2. Its within reason for most purposes, but needs some fixing to be jitter free.

[endreola]

I've also noticed a little fluctuation on mine as well, roughly +/- .2
Am controlling it from a Squarp Pyramid using Midi DIN through a mio10.

[TonyToot]

Hi everyone,

Well, after a lot of head scratching, and trying all sorts of different permutations, I've arrived back to how it was originally shipped, I'm using the delay in Logics midi settings set to -18 ms and seems to have tightened things up. The Matrixbrute's display still shows fluctuations of .1 of a bpm, but to my ears when played against Logics metronome or drum track, all seems good.

I need to do more testing to be fully satisfied, but so far, I'm a lot more happier than I have been. I need to record some long passages against stuff that I know is bang on to make a final judgement.

[Lunatic Sound]

Ok, since there was a lot of people having issues with their sync I did some in depth research.

First,  I want to adress,  that there have been discussions about the Sync Clock settings in the MMC and that 24ppq would be advisable. I am preeeety sure that this value is ONLY relevant, when you are using the CV clock in as a clock source, which as far as I can tell none of the folks experiencing trouble where. I am sure DrJustice could confirm this, because he is a wizard, but I also just ran a lot of very detailed test and REGARDLESS of the Sync Clock In settings my MatrixBrute shows about 3 to 4 ms worth of fluctuations at 96 BpM when the sequencer is running, synced to my CuBase 10 pro with an RME Babyface.

And here we get to the bad part:

This fluctuation is present in any setting I could conceive. I tried USBMidi, Midi via DinJack, various MMC settings, differing buffer sizes for my interface (down to 2ms latency on 64 samples)....

Which is to say, I am afraid it is an inherent property,  if not to say flaw, of the MatrixBrute. I hope I am overlooking something, though. Please feel free to prove me wrong.

When I find the time I will test one more time with an MPCLive as clock source, and I will report those results as well.

And if you think, 3 to 4ms at 96 BpM fluctuation is not that bad: I only started this rather tedious operation, because I was not happy with the sync musically before...

[standingwave]

At least I know it's not just mine that is broken.
 The USAMO I'm using to send MIDIclock is clocked with my audio (its audio to MIDI converter) and my other synths are locked solid to the signal. I can confirm that I have up to 2bpm fluctuation when sync'd to a sample accurate and jitter free clock.
 This I hope can be fixed with firmware and is not a hardware issue. I'm close to returning my brute if we don't get some communication from Arturia regarding bug fixes asap.
 I've opened a ticket and did get a response from Edouard.  He just gave me general information about MIDIclock jitter. If he understood my first request for support, he would have known I have already solved any jitter in my system.

[Lunatic Sound]

But StandingWave, is it really 2BpM, or rather 0.2 BpM? The first would be crazy, everyone else here is experiencing 0.2 rather...

[standingwave]

Sorry to all for misinformation. I just checked my video footage of the sync. Lunatic is correct:
 When I attempted 120bpm sync the Matrixbrute drifted from 119.9 to 120.1.
  So yes  .2bpm variance in temp sync.
 Either way though,  not sync'd.

[TonyToot]

Ok, since there was a lot of people having issues with their sync I did some in depth research.

First,  I want to adress,  that there have been discussions about the Sync Clock settings in the MMC and that 24ppq would be advisable. I am preeeety sure that this value is ONLY relevant, when you are using the CV clock in as a clock source, which as far as I can tell none of the folks experiencing trouble where. I am sure DrJustice could confirm this, because he is a wizard, but I also just ran a lot of very detailed test and REGARDLESS of the Sync Clock In settings my MatrixBrute shows about 3 to 4 m worth of fluctuations at 96 BpM when the sequencer is running, synced to my CuBase 10 pro with an RME Babyface.

And here we get to the bad part:

This fluctuation is present in any setting I could conceive. I tried USBMidi, Midi via DinJack, various MMC settings, differing buffer sizes for my interface (down to 2ms latency on 64 samples)....

Which is to say, I am afraid it is an inherent property,  if not to say flaw, of the MatrixBrute. I hope I am overlooking something, though. Please feel free to prove me wrong.

When I find the time I will test one more time with an MPCLive as clock source, and I will report those results as well.

And if you think, 3 to 4ms at 96 BpM fluctuation is not that bad: I only started this rather tedious operation, because I was not happy with the sync musically before...

I've tried all sorts of different settings yet get the same results.

Not the greatest test, but I thought I'd put a bpm counter on a track recorded from the Matrixbrute. Although it doesn't show up the constant fluctuations that I'm seeing on the synth, the plugin does show the occasional dropped bpm.

[Lunatic Sound]

The +.1 and -.1 fluctuations on their own are not that unusual, when syncing hardware gear with a midi clock, but the problem is, that the lack of tightness IS audible...

[standingwave]

My other synths don't exhibit the behavior from the same clock source and yes, it is audible.
 Point is that it should be sync'd, not almost sync'd. 
This is a synth that costs $2999.99 CAD before tax, $3388.00 after tax. This stuff should definitely work.