Workaround for late MIDI pattern change issue?

It was fixed in 1.07 but if it still exist for you i Would first update to the latest firmware and then check the usual suspects of ch.len etc.

1 Like

just a quick check: are you who have problems with this connecting the box with issue via midi out or through? i had this problem with my AR and A4 when these were connected via midi out. Because the change message comes to one machine at the end of the pattern, and it doesnt look ahead for this message; it sends it just as it comes, and therefore, the delay will make the message come down the line on the next bar. Solution; connect via midi thru on all devices. Example: OT, Midi out -> Midi In, A4, Midi Thru -> Midi In, DT, Midi Thru -> Midi In, AR, and so on. Hope this helps someone :slight_smile:

1 Like

mmmm So Question, I just aquired a AR to pair with DT and im using the DT as master as i can use the LFOs on the DT MIDI channels to edit extra parameters of the AR and also Parameter lock the MIDI CCs of the Stored Quick Performances in the KIT (so i need the MIDI out of the DT and not the MIDI thru). I have program changes in the DT pattern on the first note of the bar sent to the AR but alas they end up a bar late?

So am i to understand it that the solution is to shift all the program changes to the DT pattern before i want the change to occur ? so the data gets to the AR before the end of the bar? but it wont change until the next bar starts? In sequential mode if that helps any one helps me.

its not fixed. filed a bug, proved with video, was admitted by elektron.

in my case its octa -> ditigone but it also happens with octa -> analog 4

thats why i believe its the octas fault, its sending its prgm ch message to late. the simplest solution would be jsut make the octa sending the command once you press the button, not on the pattern end. this way, provided the ch. lengh param are set correctly on all machines, it would work like a charm

1 Like

Awesome for you.

It works fine for me, and has since the bug fix.

in which setup? awesome is really something else. its the one problem which makes it impossible for me to play just with the elektrons together.

Awesome that they recognized the bug. No disrespect to you. Which machine did you file it for?

Currently I’m using the monomachine and digitakt together, the mono is master. I’ve tested it with the octa as master both before and after the bug fix. The digitakt was changing late slaved to all this machines and it does not since 1.07. It was the one show stopping bug I was facing.

1 Like

weird that it works for you and not for me…
but its been weird with me too as it did work some time… then suddenly stopped working … and then occasionally again.
at first i always thought i made some wrong setting somewhere but last time it worked perfectly for an an hour or so… then i did nothing but loading another part on the octa… stopped working. went back to the old part (which was saved), didnt work. went back to the last backuped state (where it worked), did not work anymore. tried fresh projects on all machines, did not work …

makes no sense to me.

The late pattern change, seem To be something that happens intermittently. Before it was apparently fixed (as you say its not), there was discussion if the late pattern change was a bug or bad implementation as not every user was experiencing it. Hopefully elektron will get it figured out.

1 Like

I recall someone at elektron admitting they had done all they could, and that having prgm ch trigger pattern change on a downbeat was, and I’m paraphrasing here - like ‘reading the future’ or something to that effect. Which we all know is hogwash, as many MIDI capable devices can change program’s with very little latency. not all, but certainly many. i don’t think the DT’s processor is capable of it reliably, and it needs to buffer the change. it just has more inherent internal latency than other, perhaps simpler, MIDI implementations. as one of the more vocal complainers about the issue initially i’ve frankly stopped worrying about it - if it can’t do it, it can’t do it - and moved on to using the DT in a different, less interconnected manner.

If I recall correctly That was the dt receiving program changes from Ableton or an mpc, not between elektron boxes. Elektron boxes should be able to stay in sync at this point. I’m personally having no issue using my dt connect to other boxes.

But in another thread someone did mention that program changes plocked to the midi sequencer have a two step latency or something like that.

if you get all the digitakt settings right so it “instant changes”, 2 steps does seem to be the latency from ableton from my initial experiments.

I need to do some more investigating, I was kinda hoping you could sketch out a “song” in ableton with program change clips and then trigger the different clips to send PGM Chng and rearrange on the fly. The latter is fine as you can press “change” whenever before the end of a pattern and have the digitakt set up so it changes at pattern end (I think anyway!)
but then the former is quite hard as you have to dummy it up to send the program changes before the clip/pattern end and then the clips will be the wrong length as it only sends PGM Chng on clip start…(I think?)

there must be a way to do this without it being totally unworkable though… i hope.

On my AR on the last OS before dual vco, it won’t change pattern from external pc from OT on time if in advanced scale mode and master length is greater than chng parameter. Chng parameter doesn’t work in this case, nor in song or chain mode, chng parameter only works when using manual sequential change. (Again if in advanced mode and master len is greater than chng)

Don’t have my gear for testing, but I recall tempo multiplier(OT)/master scale(AR) not being the same across devices can through things off. Might even want to check swing settings on devices to see if that throws off timing…

the way im doing it now is sending PC commands from ableton with a negative 300ms track delay. this way its enough time for the machines to react.
the same would be possible to program inside the octatrack, has nothing to do with reading the future. its just that the pc is sent some steps before the actual change is happening.
i just hate to use the computer just for that (ok it has other advantages in terms of arranging of course), still i would like my setup to be as simple as possible, and this adds a unnecessary complication.

1 Like

I’m having the same issue, but sending program changes from Octa to Digitone. Sending from Octa to Rytm works fine.

DT to ARmkii for me is a fail. I have to generate 2 versions of any pattern that is just before i need the AR to transition. One without the plock for pattern change then one with it that starts one bar before i need the AR to change patterns.

Elektron devices are litteraly the only outboard i have that have this anomaly. Drum machines especially need tight snappy changes. BSP changes on que and my crappy akai xr20 changes bang on?

They will work it out…and its not a deal breaker
… but for now its doing my head in “working arround”

How come you are plocking pattern changes instead of using the sync options?

1 Like

Well the theory is to change AR patterns mid DT pattern or any step of a sequence. So for example the last bar of a phrase changes patterns on the AR with instant jump…i can get interesting trasitions and fills on my old AKAI xr20 drum machine this way…so for example the last bar has a different kit/pattern/swing/signal routing…when you send a program change on the first step of the last bar…

Ive only had the AR a fortnight so there may be better ways of doing things? Do you know a more elegant solution with what experiences you have had?

thats an interesting workaround, still seems convoluted. i think personally its really quite embarrassing for elektron to not be able to make their machines play well together. Specially considering how long this problem seems to be around.


Been reading old threads from 2018 but not hearing anything recently about the Digitakt delay 1 bar when switching pattern from the Octatrack…
Setup - OT running master to DT via midi. DT firmware 1.11 , OT MkII firmware 1.31.
Octatrack: Prog ch send on, Ch 14 - Recv on, Ch 14. Transp Send on, Recv on- Clock Send on, Recv on. Under Pattern settings: Scale mode normal (or per track, doens´t really matter), Use pat set. 16/16. Under Master (Scale) set to Inf (but should reset after 16 steps hense the use pat set 16/16)
Digitakt: Sync: Clock recv on , send on, progr ch recv on, progr ch send on. Under Channel settings: Progr chg in ch 14 (same as OT) , Progr chg out 14 (same as OT). Pattern length (doesn´t matter if set to per track or pattern, I´ve got it set up as 16 ch length)

When I change pattern on the OT, most of the times it takes one more bar for the DT to change to that same pattern! I say most of the times cause once or twice it changes at the correct bar. But only rare occasions, sometimes…
Is the bug still there or am I missing something?


1 Like