Octatrack midi cc delay( send priority )

How do I reproduce this bug? When Starting from a blank new project?

Some bugs only start creeping up after something else happening within a project, so we need to be able to reproduce this from a blank, fresh project. Being able to do so will help the tickets getting the point across.

To reproduce this behavior:

  1. Lay down a note trig on a MIDI track (e.g., note F3 with a velocity of 127, on step 2 of MIDI track 1)
  2. Parameter-lock a CC message (e.g., CC#1 / modwheel to value 70) on that same step (step 2 of MIDI track 1)
  3. Observe the output on a MIDI monitoring tool (hereā€™s a free one for OSX)

The log of sequential MIDI data will look like:

Note On 1 F3 127
Controller 1 Modulation Wheel 70
Note Off 1 F3 0

Say youā€™re sequencing a synth whose filter and amp envelopes have a really quick attack, and say CC#1 controls the FM index of that synth. When step 2 triggers, youā€™ll hear the filter and amp open up, and then youā€™ll very quickly but often perceptibly hear the FM index jump to the value you p-locked it to. The amount of delay between Note On and the CC message, and so the perceptibility of this, will depend on how much MIDI data youā€™re chugging out and where the receiving synth is in your MIDI chain. It sounds glitchy and itā€™s not controllable.

If you were to do this on the :elmd: or :elmm:, youā€™d see:

Controller 1 Modulation Wheel 70    
Note On 1 F3 127
Note Off 1 F3 0

And so right before the envelope even opens, the FM index is locked to the level you want it, and you donā€™t hear it step to that value mid-note. This is also how it works when you p-lock timbral parameters on the OTā€™s audio tracks or any of the other Elektron machinesā€™ internal sequencers.

Iā€™ve seen this happen for entirely new project as well as for projects Iā€™ve worked on for months, and several owners have written about this, here and on Muffwiggler. Some examples:

Aside: Itā€™s unclear whether this MIDI behavior was intentionally designed (maybe to address some tradeoff) and thereā€™s just an apparent consensus that this is not how CC and Note On data should be sequentially ordered, or whether this is classically a bug in the sense of unintended behavior. Either way, I see this as something more than an idiosyncratic feature request.

2 Likes

Would it be possible to get around this problem by using a second midi track for the CCs, and offsetting the events via microtiming?

1 Like

:thup: Yep, each midi track can be any MIDI chan

00000A74 3 ā€“ 9F 3C 64 16 C 4 Note On
00000A75 3 ā€“ BF 0A 04 16 ā€” CC: PAN
00000B18 3 ā€“ 9F 3C 00 16 C 4 Note Off

As of OS 1.25H this is the order Octatrack is sending MIDI instructions. The only way I can see to workaround this is double programming of the steps. (earlier step for CC or worse, eat another track)

Any feedback from Elektron?

i would be really interested in this as well. cant manage to map velocity to cc in the eventprocessor :frowning:

Any news on this? Is it fixed?

Donā€™t think so. I use Event Processor Plus to control slices and I can reorder midi data.

There has been no new software revision since 1.25H, so no change. I imagine it is planned behaviour, rather than a bug, strictly speaking.

1 Like

Oops I meant I donā€™t think so.

1 Like

So is this just a feature we have to live with on the ot? Why would they plan such a behaviour?

All my midi tracks are slightly lagged which is annoying when wanting to sequence faster patterns. Are there any decent workarounds? Nudging the trig back means its unable to trigger patterns starting on the 1st step, so that isnā€™t ideal :confused:

I donā€™t understand what you mean by this. Can you explain more? What are your MIDI tracks lagging behind? What MIDI messages are you sending to what instrument?

1 Like

Iā€™m going to bump this one. I never use the midi sequencer much, but i just bought a piece of gear with the OT sequencer and CC capability in mind, and they were going to be best friends. Then i found this out the hard way.

Why is this even a thing? Iā€™d like to know the reason behind it since It makes no sense at all in my head.

What are your problems concerning CC/notes order? Some synths donā€™t work properly receiving CCs before notes?
Can you hear a lag?

As described in the original post. CC arrives after notes on each step, making CC have no effect on the step itā€™s set on.

@Paland Just curious with what midi gear and what parameters. It seems to be ok with my Micromonsta. Good control.

Btw I made a test :

  • midi track 1 with trigless control changes, channel 1
  • midi track 2 with corresponding note trigs, channel 1

That way notes are after control changes.

[MIDI-In1] - B0 00 27 (Control Change)
[MIDI-In1] - 90 30 64 (Note On)
[MIDI-In1] - 90 30 00 (Note On)
[MIDI-In1] - B0 00 41 (Control Change)
[MIDI-In1] - 90 30 64 (Note On)
[MIDI-In1] - 90 30 00 (Note On)

I guess nothing ever happened with this? Iā€™m sequencing my Rytm with my OT, and want to modulate the sample start time on arpeggiated drum hits. But since the first CC gets sent after the note on, the sample start time is never where I want it on the first drum hit. BUMMER

Would you use another midi track for CCs?
They can be sent before that way.
Lower tracks midi messages are sent before.
T1: CC
T2: Notes

Btw I had problems with AR only, with sample start and a sample chain, hold random lfo. Wrong start position, not corresponding to the beginning sample of a 120 samples chain.

1 Like

I read you mention that before. In my case Iā€™m trying to do this on multiple parts, so using 2 tracks for it would take me from an 8-track sequencer to a 4-track sequencer. :frowning:

Iā€™m not doing sample chain stuff at the moment, just a single sound (doing kind of granular stuff to it).

I did sample chain stuff in the past with the Rytm though, and the problem with the LFO is that it will actually trigger on ā€œin-betweenā€ numbers! Like itā€™ll pick start times that you canā€™t choose with the knob itself (e.g. 56.3, not just 56 or 57). Makes it impossible to do fun sample chain tricks with the LFO (or velocity) on the Rytm.

On 1 track itā€™s also possible to use trigless and microtiming.
Possible to invert notes/CC order with a midi processor.

Thanks to confirm, I also experimented this with a friendā€™s Model Samples. Probably the same problem with DT?

1 Like