Octatrack MIDI note quirks, help needed!

Elektron Gear Octatrack
#1

So i recently bought a DMX lighting controller (the ADJ DMX Operator 384) and my first light fixture to mess with and program with the octatrack as it responds to MIDI notes and translates them to “scene” changes on the lighting controller (MIDI note C0 (number 12) corresponds to bank 1 scene 1 on the controller for instance).

When tested with my digitone keys, analog four and rytm this works like gangbusters. A MIDI sequence can even be 300 BPM firing a different midi note every single step and the lighting controller will instantaneously change scenes without a hiccup.
HOWEVER
When setting up the exact same MIDI sequence on the Octatrack it will respond intermittently or sort of choke up for 2-3 seconds after each note sent. Like the first note will register, but then the next couple won’t and it’ll start tripping over itself. I have sent both this and the digitone through a midi monitor and the results are identical. Confusingly, when the octa midi signal is sent to both an external synth and the lighting controller, the external synth will of course play back the notes perfectly in time whereas the lights will still be choking. Whats more, if i send the digitone midi into the octa in then back out through octa out on the same midi channel (which acts as a soft thru for those who don’t know) it exhibits the same problem. Swap the cable from out to thru however? gone.

Any ideas? if this is just a quirk of the Octa midi sequencer behaviour can somebody help me understand it? Since this is one of its primary functions as a £1200 8 track midi sequencer, its a little disconcerting that it would be struggling to reliably send out a one page quarter note sequence at 30BPM without constant errors.

So the primary confusion stems from: Why does the lighting controller react so poorly to the Octatrack but exactly as expected to the digitone midi notes when, as far as I can tell, both are sending the exact same MIDI commands on the same channels. Same timing, note lengths, note numbers, and according to MIDI monitors: same exact hex codes. No visible difference.

Any insight would be wonderful, thanks gang

UPDATES

  • See the most recent comment from myself for the solution, and the other posts for the progress of the very unusual problem itself, for anybody reading this in the future
    Octatrack MIDI note quirks, help needed! - #28 by chrl_nrys

  • 01/03 ACTUAL FIX: Sending the octatrack’s midi through a processor box like the blokas midihub or even the rk-002 cable seems to filter and clean up the signal and the lighting control behaves exactly as expected. huh. the OT is sending a lot of extra confusing stuff that standard midi monitors dont seem to be able to pick up on.

#2

Does turning the OT MIDI Clock off make a difference?

1 Like
#3

Are you aware that these Elektrons use different numbers for notes ?
Their C0 correspond to C-1 on OT/MD/MnM. Note 0.

2 Likes
#4

Missed that. I second @ericd’s idea to check with and without clock, (check with OT and Digitone).

There must be a difference, maybe clock is filtered by the midi monitor. I read the way clock is sent between notes can change behavior on certain gear.

Clock makes my guitar fx buggy if I send CCs at the same time.

#5

It was off by default, i’m testing this using Demo Mode to try and reduce variables from the way i might have things set up

Clock send on or off changes nothing however

1 Like
#6

Yeah i am, found this out with a midi monitor, but i’m sending the right note numbers as it’s triggering the right scenes on the lighting controller albeit intermittently on the octa

1 Like
#7

Clock send set to on or off changes nothing either. This was also elektron support’s first question

1 Like
#8

Yes I saw that after. Did you check if proper Notes Off are sent by the Digitone ?
OT sends Notes On Velocity 0 as notes off, so I’d check if it can influence your DMX behaviour.
Happen with some gear not accepting Velocity 0 as note off.

1 Like
#9

I’ve had this problem with a couple of pieces of gear being sequenced by the OT- but they have triggered properly when the OT MIDI clock is off.

If you happen to have something like a MIDI events processor, mapping MIDI On Velocity 0 messages to MIDI Note OFF should help. My only other suggestion I’m afraid would be to contact the lighting controller manufacturer- I know that some other manufacturers have released firmware updates to address this exact problem with the MIDI Note off issue.

2 Likes
#10

Yeah, both Octa and DNK send identically:

  • Note on 12 velocity 100 (90 0C 64)
  • Note on 12 velocity 0 (90 0C 00)

For every step
No clock is being sent

Confusingly, one online MIDI monitor i tried has both machines sending note on messages for both velocity 100 and 0 messages, and midiview for windows has both sending the above note on 100 and note off 0. Annoying, but both are consistently the same.

#11

Ideally we’d have to analyse bits !

Just in case : Did you check Turbo midi on OT ? New project ?

#12

And, as @ericd also mentioned, a midi processor would probably solve this issue.

1 Like
#13

Checked turbo, it isnt activated and makes no difference however i set it
New project also does nothing, i’m using demo mode to test it in order to remove variables

#14

But since it’s apparently outputting the exact same midi notes and information as the digitone, what exactly would i be processing it to?

#15

Apparently there is a difference !

Clean the midi data with filters, change Note On Vel 0 to Notes Off.

#16

@chrl_nrys you didn’t mention note lengths. I’d try with shorter note lengths on OT.

1 Like
#17

Well apparently there is, but not in anything a MIDI monitor seems to be able to pick up as, again, the results from the digitone and the octatrack have no visible differences, same hex codes and timing, same channels

But I’m left wondering why this would be necessary when the digitone sends out note on 0 just like the octatrack and the lighting controller responds perfectly. If it works fine coming from the digitone, what would converting it to midi note off accomplish? Is there a way to use an interface with midi in/out and some software or a plug-in to process the midi like an event processor in order to test this just to rule it out?

Updated original post. Note lengths are the same (1/16) across both devices by default. They start to cause even more choking or just overlap when turning them up but no matter where they seem to be set they don’t solve the issue either.

#18

try same without: CONTROL > MIDI SEQUENCER > CC DIRECT CONNECT [ ]
Not 100% sure if that fixes but CC DIRECT CONNECT interferes with the way the sequencer is working compared to devices who don’t have this feature.

Thus is you mainly do not change the midi signal but the behaviour.

Apart from that, if a sequence is given as trigs in a midi pattern then switching the pattern will by default put the next pattern on the stack waiting for the “chain” to end and act (visible in the display written behind the pattern number). Switching patterns can be set to direct (one step) per pattern (is stored in pattern settings [RECmode-OFF, FUNC+BANK]) and this feature is called CHAIN BEHAVIOR.

also to consider: the OT supports notes with Length 0 and many devices react not happy about it specially when they must interpret note-offs. A typical symptom for such devices is that they can not turn a note on as long no note-off is properly received. Also LEGato ON can interfere or ARP in any of the modes (which practically forces a note pattern/length and uses the second note-length parameter meant to control arp note lengths) and of course the LFO can spook any expected note out of range (well you have already a method to check with midi-monitor).
Some time ago (a couple of years) i wrote to the support that note length 0 makes technically not much sense as a the user set a note and expects note-off after note-on no matter how short (shortest would be one tick, but with len==0 the behaviour of many midi interfaces is undefined).

There are many more ways to spook your data in maybe unexpected ways. In example ONE-TRIG shots and like-like

1 Like
#19

Latest revelation that makes 0 sense to me

#20

It didnt make a difference unfortunately

THIS is the only way i can make sense of it, but i don’t understand how the octatracks midi behaviour differs from the other devices in a way that would cause this particular use case to not work

I’m not exactly sure what youre suggesting to set to what here but i fiddles with eeeevery parameter on these pages and yeah, no difference

My first guess was that something about the octatrack’s more full featured midi sequencer was injecting something confusing into the stream or or could be a one shot trig or something but i am doing my darndest to keep it as 1:1 as the others as possible. Note length especially i’m keeping at 1/16 or whatever works on the others

Also see my latest PROGRESS update, can you make sense of THAT?