I’d like to call attention to the fact that Elektron sequencers have a serious and extremely non-musical flaw in that when changing track division or length, the sequence for that track is no longer aligned to the “1”. It’s just off, plain and simple. You have to completely stop the box entirely and start it again for things to line up correctly.
In a live performance this is pretty much unacceptable and can’t really be done with intention. I just want to half or double time a track, or loop in 3/16 forever until I decide to go back to 16/16 and have it all lined up perfectly. Nevermind the fact that I might have a pattern where one track is in 5/16, another is in 7/16, and another is in 16/16 and I’d LOVE to be able to maintain this (to use an Ableton term) “legato” flow when changing patterns - which to me is a must have in every hardware sequencer yet hardly any do - that’s worthy of its own topic, but first can we at least get this one foundational thing fixed across all of your boxes? I know this is an issue for the Octatrack, Digitakt, and Digitone, so I assume it is consistent across the whole range.
I really do think this should be treated like a critical bug and am a bit surprised there isn’t more demand for this to be ironed out.
Am I the only one making polymetric music? Of course not, and I’d really love for people to chime in here so Elektron has some idea of how important this is.
So again, my request:
the result of changes to length or division should be calculated in such a way that the sequence is aligned to the grid as if the sequencer had been at that division or length since the pattern started all along.
Can I get a +1?
==========================================================
The rest of this post is just a hypothesis. I’m trying to explain and understand what is going on in code/math terms, but again, it’s just a hypothesis.
The algorithm for determining the next step in a track, in ultra-simplified terms and ignoring track divisions and just looking at track lengths, in pseudo-code, should be:
nextStep = (globalStep + 1) MODULO trackLength
Where globalStep is the count of steps that have elapsed since the sequencer started (or since the pattern started, but that’s inferior re: legato pattern changes, but I digress). Not taking division into account, it really is that simple.
For example if globalStep = 8, when calculating the next step:
# TRACK 1 (length = 16/16)
nextStep = (8 + 1) MODULO 16 = 9 # CORRECT
# TRACK 2 (length = 5/16)
nextStep = (8 + 1) MODULO 5 = 4 # CORRECT
Using this very simple algorithm, it doesn’t matter when or how many times you change track length. It will always be aligned to the grid as if it was that length all along since you started the sequencer. This appears to be exactly how the sequencer appears to work currently when you press PLAY for the first time, but once you’ve changed the length or division, it falls apart. I believe the way it actually works, is that instead using the modulus of a global step count, Elektron does something (but not exactly) like:
nextStep =
if (currentStep + 1 >= trackLength)
0
else
currentStep + 1
Because it’s not using to a global counter, and using individial counters per track, it has to do something silly like reset currentStep to 0 whenever you change length or division and it is at risk of being incremented to an invalid number. Using the same example where globalStep = 8 and you change from 16/16 to 5/16, comparing the two algorithms:
# changing from 16/16 to 5/16 while on step 8
nextStep = (8 + 1) MODULO 5 = 4 // CORRECT
# changing from 16/16 to 5/16 while on step 8)
nextStep =
if (8 + 1 >= trackLength)
0 # INCORRECT!
else
8 + 1
But then, why does Elektron fall off the grid when you increase the track length? This is where my hypothesis falls apart. I really have no idea what the hell is going on - I just know it’s wrong. Curious to know what the actual logic is, because trying to nail it down in practice is pretty difficult - it just seems to randomly pick a step.


Apparently this is much requested feature so I’m keeping my eye on it.
And your advice on this has always been very relevant in the past 