I’ve seen several reports now of the DN2 locking up when under external control. This definitely concerns me, as I absolutely have plans to use the DN2 in tandem with the Hapax onstage, and will also be playing it from an external keyboard.
Just curious if you guys are working with a direct connection from your sequencers to the DN2, or if you’re running MIDI through your computers, a USB hub, using OB, or what have you?
And is it locking up from basic gate-trig sequences (i.e. drums) or is it mostly happening on denser melodic sequences using CC automations?
In the case I mentioned I was testing out the built-in sequencer on the Akai MPK Mini Plus. So while it was sequencing drums I was also playing some chords from the keys. No CC’s, just notes. This was direct from the Akai’s 5-pin midi out to the DN2 5-pin midi in.
I’ve also had it lock up in other scenarios, where it is on a Midi merge box with other Elektron’s and I am playing keys manually.
I dunno man, just feels like sometimes the DNII just straight up says “too many notes”.
Haven’t tried that, but it was a simple kick, hi-hat snare pattern and a 3 voice chord, maybe 4 voices on accident if the pinky finger got loose … it shouldn’t lock up on that.
Oh, I agree that it shouldn’t be a problem, but that’s assuming that the voice allocation algorithm is, in fact, working correctly. If it’s not, then even two voices could be a problem.
After all my work on the Tempest (which remains fraught with voice stealing issues), I have learned to objectively account for what is happening, not what should be happening.
The fundamental difference between trigs on the internal sequencer verses notes being received from an external source lies in the machine’s ability to look ahead: i.e. incoming notes from an external source (be it a sequencer or a keyboard) are effectively a surprise, which can confuse a buggy allocation algo.
To which end, assigning voices should (presumably) eliminate any real-time decision making… Just a theory, of course.
Once a sequence is programmed on the Tempest, for instance, the algo works perfectly. Playing the pads or sending it notes from an external source, on the other hand, is a crapshoot. Assigning sounds to voices solves this problem.
Anyway, I’m not saying that’s the issue, but I figure it might be worth a try.
If you do try it, I’d be curious what the results are.
In my case, I was working directly from the Hapax to the DN2.
Sequence wise, I had three tracks running mélodies. One track had one lfo, another had two. I also had a midi harmoniser effect on one track.
Combined with the use of unison on the DN2, voice allocation was getting crazy. I remember looking at the voice menu and being a bit hypnosed by the bouncing white dots.
No obligation, of course, but since both of you have already experienced this bug (which I’ve yet to run into, though it is early days setting up this new live rig); if either of you do endeavor to employ my suggestion above (i.e. to strictly assign voices), I’d love to know if it solves the problem.
I really hope it does; else, much to my dismay, I’ve got some rather urgent maneuvering to do.
It’s on my to do list for the day, we’re facing the same problem! I’ll try to recreate the bug, and also check if your voice allocation suggestion prevents it.
So, I tried reproducing the bug but it didn’t work. Same type of set up (hapax to DN2 directly), with the same type of midi information flowing out (different melodic sequences + midi LFOs).
I couldn’t spend a lot of time testing. When the bug happened, it was during rehearsals, so the DN2 had been sequenced over several hours.
I’ll give it another go this week. I really hope I can find a way to reproduce it, I hate the randomness even more!
For what it’s worth, I spoke with a colleague of mine who uses a Digitone OG live, sequencing it exclusively from his OT and A4 (he doesn’t use the DN’s sequencer at all), and asked if he assigns voices…
He said that he does assign voices, indeed out of necessity, and that he experienced a whole host of problems before he started doing that (i.e. hung notes, voices dropping out completely); though no hard crashes per se.
According to him, everything has been stable since defeating the dynamic voice allocation.
So, I’m just going to do that from the outset here, in hopes that I never see this problem. But if you do find a repeatable scenario, I’m all ears. I agree that there’s nothing worse than a random critical issue.
Cheers!
P.S. I’ve neither owned nor have I tested a single piece of kit that presumes to use dynamic voice allocation that doesn’t suffer from problems because of it.
And with that, despite taking said precautionary measures…
I managed to crash the DN2 within seconds of pushing play tonight, sequencing it from the Hapax, drums only.
However, the firmware update that Elektron just released may have something to do with this issue. Specifically…
Sending Note On/Off messages to the FX control channel could in some circumstances freeze the device.
The default for the FX control channel is MIDI channel 9 (and therefore Track 9); which is where I, for one, tend to keep my main kick.
Before I even did the update, if only to be thorough, I went into the System Settings and simply turned the FX control channel ‘OFF’ altogether. I couldn’t crash the DN2 for trying after that.
I haven’t had a crash since updating either, and I’ve been taking it to task, so…
Here’s hoping that’s the last we see of that problem.
I experienced very sluggish encoder / values today while trying to set all trigs to shorter length using the new trick, holding trig+track. 64 trigs in 128 steps. Never noticed this before so it seemed weird.
currently I’m on 1.10 and 1.10A as well as 1.10B are available. Shall I update from 1.10 to 1.10A first, wait a sec and then update from there to 1.10B or it’s okay if I jump directly from 1.10 to 1.10B because all the awesomeness of A will be included in B anyway?
I am very sorry but intellectually I have been experiencing a massive and quickly accelerating IT regress recently (not just with synths) and don’t wanna mess my little box up.
Thanks for tips! Shaken and stirred, yours truly,
the DN II idiot
