I sent the DT/DN video to support as well.
I mean, that this issue hasn’t been caught yet. Is nobody using synched patterns maybe? I know a lot of live set for example use manual pattern switching.
2 Likes
It has been talked about, my guess is people just muddle through/make do, rather than pursuing it and finding a minimal case, reporting it etc (as you have done!)
2 Likes
I can’t work out what’s happening in the video… maybe renaming the Patterns to ‘Ptn 1’ ‘Ptn 2’ etc would make it easier to see.
I’m now also curious to know if it’s possible to have one machine working in Adv mode and one in Std mode and have them change together.
It is something Elektron should prioritise fixing IMO too, when you consider how many people work with 2< Elektron units… does that mean everyone is avoiding Adv mode?
@DaveMech do you have a take on this? Asking you as I’ve often seen you using more than 1 unit and I know you’ve just been doing your instructional videos…
They both look in sync in this video is that what you are trying to portray? Or am i missing something obvious?
They are supposed to change patterns at the same. time. In this video it’s clear that the DN changes pattern one cycle later than the DT. I entered trigs on all steps on both machines so it’s better visible.
1 Like
In the video I have two new initialised projects. I switch between pattern A1 and A2 on the Digitakt. Pattern A1 has trigs on all steps on both machines. Pattern A2 does not. So you can clearly see when patterns switch.
As you can see the Digitone lags behind. They should switch patterns at the same time. As a matter of fact they do when you have Pattern Scale set to global. If you set pattern scale to “per track” pattern sync breaks.
1 Like
So what happens if you have CH.LEN set to 0/Off on the Digitone and the M.LEN at a regular length (say 32 or 64)?
I will try some of this myself later on my AR and A4 too…
They will switch patterns together at 32 or 64. But then you cap all your tracks a 32 or 64 steps. Well, why not increase that number? Because then you have to wait a long time before patterns switch. Hence the CH.LEN paramater.
Typically I would set M.LEN to INF and CH.LEN to 16. Meaning I have endless patterns playing but when I cue a new pattern it will play at the next 16 cycle. And it is how the machines behave. But not together 
2 Likes
I don’t have these particular machines but the patterns look like they both start at the right time and run in unison the whole length of the video. Each step lights up on the right step on both machines. Guess I am dumb I just cant see it. Maybe it would help with audio.
It’s just that the Digitone (right) is 1 Bar late.
I think I have watched it 50 times haha
It’s not about the steps being in sync. They are in sync. That’s not the issue. Pattern switching is not in sync. The DN lags behind one cycle. You can clearly see that the DN switches to pattern A2 later than the DT. On both machines the pattern with all the trigs lit should start at the same time.
My OT and AR sometimes exhibit this problem too!
Inspired by @martin01 I filmed and photographed a bit of testing. You can see all the images and videos I took here:
I tested:
AR leading OT with basic and advanced scales
OT leading AR with basic and advanced scales
The folder linked above is divided up into settings, the leading/following configuration and the basic/advanced pattern config.
During my first pass, it looked like I was going to agree with @martin01, that the sync is consistently OK with basic scales, and consistently 1 “bar” behind with advanced ones. It also looked like when the AR leads, the OT always followed correctly. However, subsequent attempts showed it’s actually not that simple in my case. It didn’t matter which box lead. It also didn’t matter whether the patterns used basic or advanced scales (although basic did seem to work correctly more often). It seems to happen more often when I restart playback without first double-stopping. But I couldn’t prove to myself that was a consistent workaround. Dig into the videos and decide for yourselves. The long one in Sync/AR_leading_OT/1 AR leading OT - basic scale is the most revealing/confusing.
On both boxes I set pattern A01 up with 1 group of four trigs in the middle, and pattern A02 with 2 groups, so you can see which pattern is playing.
I’ve also contacted Elektron support, linking this thread and my Dropbox of evidence.
1 Like
Just tested with rytm mk1 into DN.
The change length works as expected if it’s used on the slave machine and set to off on the master but then the master box is stuck to the length of time that you want for pattern changes.
When the master uses the change length setting it seems to be sending the PC message as it’s changing - I think that’s what happens when you try and daisy chain without using midi thru.
Never noticed this before…I tend to jam it out in one pattern a lot.
Have you had a response back from support yet?
I tested with DT/DN combo and AR/A4 combo. Both have the same problem.
Response from support;
Hi Martin,
We have confirmed that the issue affects all our current products (a little bit differently on the Model series that don’t have a master length but it is there as well). I have traced the issue back many years, so it is not new. We now have a consistent way to reproduce so we should have a solution soon and the issue has the highest priority to be fixed. Thanks again for reporting!
10 Likes
Great work for being giving them a way to repro, and therefore fix. Thanks for keeping us updated.
1 Like
It’s always been like this, chng length doesn’t work when chained the only one that can change properly when being chained is the octatrack.
I spoke to support and they basically said it’s designed that way, I would be surprised if it changes tbh but I hope so
The current way makes no sense at all. In their response they say they will fix it with high priority
1 Like
So I stumbled across this thread, and I think I know what’s causing this. I just put in a ticket and the team got back to me saying this:
A known bug can cause patterns to change too late if the receiving device is in advanced scale mode and have a master length that is higher than 16. We are working on a fix for this problem.
So yeah, as they said before it’s an outstanding issue, but that’s the exact cause. As others have mentioned in this thread, using simple pattern length is one solution, but it looks like it’s also fine to use advanced but to have the master length set to 16.
1 Like