Pattern change one bar late issue

Elektron Gear
108

This has been hurting my head for a long time also, I usually use my Squarp Pyramid as to dish out program changes and usually have a dedicated track for doing this, I leave it muted during pattern one for example and unmute it just before moving to pattern two so it sends out the program change. I have to be pretty “on it” to execute this and it means I’m so focussed that I can’t be tweaking other things live during these moments.

I’m now working on a new set and thought I’d use Octatrack, Digitone and Digitakt together, OT as master. And yes, what a headache with the late changes!!!

I’ve tried allsorts of ways of doing this and on reading your suggestion of changing the M.LEN to 64 and CH.LEN as 16, everything changes bang on time and not a bar late. This completely throws all my lovely poly rhythms out which is a huge pain as I use them a lot!

I tried setting the M.LEN to 1024 and it doesn’t work, I’d usually in my other setup have M.LEN as INF. So why oh why will M.LEN 64 be the magic number? No idea here!

Another strange strange thing is that if my patterns are all on for example A02 and I stop the machines playing and change OT to pattern A01. DT and DN are still on A02, and if I play from their own play buttons, that pattern continues. But, if I then press play on OT with A01 active, everything instantly jumps to A01 with no prior cue.

I swear this is a bug that has been swept aside, or perhaps it is so weird, it’s hard to describe to Electron.

Try it out and see if you get the same, sure there’s no reason M.LEN has to be 64 for instant changes, my understanding is CH.LEN is the length that occurs before a "CH"ange and M.LEN is the Master LENgth of the pattern, ie, with M.LEN as 64, everything is in sync, but still changes according to CH.LEN.

Does that all make sense?

109

Could be good, though I find Octatrack really doesn’t like to be a slave to anyone. If I have it as a slave, I have to say goodbye to overdubs. I’m in a band that use two Octatracks and my mate is gutted that I’m master as he can’t do any fancy overdubbing. (sorry, maybe for another thread but kind of relevant)

1 Like
110

Yeah… you probably read the manual.

The manual lies about this. It’s only true for manual pattern changes. Chains and syncing via MIDI is still based on M.Len. On all the boxes.

Electron were planning to update the manual… but that was two years ago… their Customer Service agent at the time told me there was ongoing debate within the product+dev team about the right way to solve this. I guess that debate hasn’t resolved yet.

1 Like
111

I wonder where that leaves the two models, which don’t have m.len, but they exhibit the same ‘one bar late’ problem.

112

Ha ha, I think with the Octatrack, it’s one of those devices that when you have read the manual you come away less knowledgable. Or perhaps we need a manual dedicated to how to read the manual.

That’s a testament to the sheer scope that this device covers, took me ages to figure out what it does before I bought the thing.

Anyhow, this issue is very strange indeed, in that with what I explained above, there are still inconsistencies. I.e, for instant changes to work, M.LEN is set to 64 on both slave boxes (DN and DT) and master on Octatrack is set to INF with pattern CH set to 16. All boxes will change after 16 steps, even though slaves are set to M.LEN and CH.LEN as 16. So ultimately they are cueing from CH.LEN but it’s having M.LEN set to 64 that makes that work, as if it’s somehow linked.

Does anyone know if Elektron read the forum when looking at bugs? There’s tons of useful info here that might make some sense to programmers there.

113

Maybe a little harsh on the manual here, it does give a lot of info, it’s just that the beast is so vast it’s gard putting into words or even knowing what you’re trying to do in the first place.

115

This looks like it’s getting even worse. Now I have managed to get everything to change in sync, it now jumps to pattern C02 after one bar of whatever pattern change has occurred. I’m working on a set to play in a club thinking it would be great to not be a laptop performer but after all these hurdles I feel like selling all this gear as the amount of workarounds to make the thing do what it is supposed to is just killing all my creativity. Has anyone else had this happen? It’s an extremely strange and annoying thing and not intuitive what so ever.

If I press stop and start on the Octatrack it then plays the pattern it had jumped away from

116

i was having this issue and the way i got it fixed is i used midi b/thru after the 1st machine.
for example: digitakt midi a ---- digitone, midi b/thru-----a4mk2 ,midi b/thru—ar mk2, etc…
i asked elektron support why and they say thats how they recommend hooking it up…otherwise the slave box(es) have to wait for its master to get the PC message before it sends it thru so that makes the 1 bar offset…its kinda dumb tbh but theres a work around…get something like the midihub or rk006…having a digitakt as the brain makes this work pretty well tho since i have 8 ch of midi i can use to automate things on the other boxes

2 Likes
118

I am facing this issue & I can’t figure how to solve it

I have been using M:S for some years as master device with some other gear (Novation, some Elektrons & some Rolands) with zero issues until I got a Roland SH4D

I have tried multiple MIDI combinations but I can’t fix it 🥲

Any ideas?

119

Which is currently sending program change and which receiving ? (There’s more than one issue in this thread).

1 Like
120

M:S send program change
SH4D receives it

121

So, the m:s sends its PC within the last step of the sequence, I worked it out once, it’s either 0.4 or 0.6 of a step away from the end. Maybe that works for your other devices but not for the new one, perhaps too close for the new Roland to be able to switch in time?

Might be worth running your sequence at a slower tempo as an experiment, to see if that’s the case ?

EDIT: Even something silly low, like 30 bpm, just to identify the issue.

EDIT 2: Found my previous post:

2 Likes
122

Not implying that this isn’t the correct place to ask about the specific issue, but did you already try the sh4d thread for any hints someone in a similar elektron/roland pc situation might have left for future users?

123

I will try & report it back here :+1:t3:

1 Like
124

Yeah, coming here from there 🥲

2 Likes
125

Tempo 60 did the trick
:sweat_smile:

1 Like
126

So … could you run your track at 60 or below and double up (or 4x up) the real speed by scaling on the m:s (and roland ?)

1 Like
127

I guess I could try, is that the only way to make this work? 🥲

Edit: tempo 80 is the limit

1 Like
128

Damn! Tried at tempo 60 + scaling X2 both machines
It stopped working again
What a nightmare :sob:

129

If my arithmetic is right, that would seem to imply that the Roland sh4d needs at least 450ms notice to swap patterns.

And so far, it looks like you can’t persuade the M:S to do that with a 120 bpm track because, no matter the scaling, its sending PC at 300ms before.