I’m new to the octatrack and wondered how i would go about sequencing polyrthyms if its even possible? I tried to research online but i came across a lot of different information some of which was contradicting and potently out of date due to os updates.
each track can run for a different number of steps, between 2 and 64 (the master can be as long as you want including infinite, in terms of repeats), you also have varying tempo dividers per track, so there’s a lot of polyrhythmic potential
You can not only do polyrythms, but also polymetric “stuff” using tempo multiplification and division. Or record and chop, you can do polywhatever in an MDUW using that ol’ trick!
1 Like
Ah thank you. Not quite sure what ho to do what you said but will use the manual to work it out!
Research resampling for yourself! When you get bored with that try ‘feedback’ in a less traditional sense, whole new planets full of sounds can come from exploring it.
Function+Page to open your step count/scale multipler window. Adjust to taste.
But but, first setup your pattern for individual track lengths! By default every pattern is set to have a global steplength.
1 Like
But but, first setup your pattern for individual track lengths! By default every pattern is set to have a global steplength.[/quote]
Good call. Mine be default is set that way. I tend to forget that step these days.
I haven´t explored this much:
What are you preference polyrhythm setup (in a “musical” sense)? I e, if and when making drumbeats.
4/4 against (insert time signature)?
Do you adjust microtiming alot too or just leave evrything as it is?
I kinda gets confused about how to think about the grid, whenever I´ve tried other settings.
If I´m having one track 4/4, it´s easy to “see” the rhythm on the trigger buttons (1-16). But if having 3/4 instead, it gets harder “visually” to feel where I should set the trigs. Which means quite some trial and error, but I haven´t been fully satisfied with the results.
Note to self:
What happens with the retrig time parameter (musically) if track is in another time signature?
Nothing changes, as the RTIM parameter is related to tempo.
really hope better support for poly/odd time sigs comes in an update some day. arbitrary number of steps per sequencer page, arbitrary number of steps for all quantized parameters (QREC, manual track trig, etc), arbitrary number of slices in a slice grid, and so on. actually I think the only odd time-friendly features right now are RLEN and the metronome…
so much potential that is frustratingly just out of reach, especially if you’ve become accustomed to using the machine in 4/4.
1 Like
Nothing changes, as the RTIM parameter is related to tempo.[/quote]
OK. Am I thinking correct here that if you would set a tracks time signature to, i e, 3/4. The RTIM, i e, to 1/2. That the sound affected by the RTIM would alter in being in syncope every second trig? Hopefully I´ve got the example number/time signature right in above, I suck at thinking in this way…
Let me explain my thoughts further, (think of this as a drumkit example):
Kick track going 4/4
Snare track going 3/4, but only one trig hit (on beat) which together with the RTIM set would have it “expanded” into syncope whenever an random LFO gets the re-trig activated?
F**k, that was hard to explain. Hope it made enough sense…
Don’t overcomplicate things, I do that all the time and it kills creativity! 
Hahaha. No it’s alright. I’m trying to create an method to be able to get an drumbeat programmed with certain randomness in the rhythm, but which still will have some kind of “logic” in the random selections. Random LFO’s with very modest depth on re-trig seems to be a part of the “solution”. The reason for all this is that I want to get an jamming drummachine that I can play guitar with. Where an fixed pre-made beat/rhythm with no variations (or them at the same place in the same way everytime), certainly would kill my creativity flow.
That was an quite interesting experiment. Will upload an example, perhaps tomorrow or something.
Hahaha. No it’s alright. I’m trying to create an method to be able to get an drumbeat programmed with certain randomness in the rhythm, but which still will have some kind of “logic” in the random selections. Random LFO’s with very modest depth on re-trig seems to be a part of the “solution”. The reason for all this is that I want to get an jamming drummachine that I can play guitar with. Where an fixed pre-made beat/rhythm with no variations (or them at the same place in the same way everytime), certainly would kill my creativity flow.
That was an quite interesting experiment. Will upload an example, perhaps tomorrow or something.[/quote]
There we go:
[url=“https://soundcloud.com/mikael-johansson-6/timesigretrigbeat”]
https://soundcloud.com/mikael-johansson-6/timesigretrigbeat
Still some tweaks to be done, I think I should make the hat re-trigs being on the snare track (instead of the kick track) to “tie” the arm to alter between the hat and the snare (but not both the re-trig and snare “being hit” at the same time).
nice, haven’t taken time to play with polyrhythms yet, thanks for posting the example. I hope to try this out at somepoint.
very nice, gotta love all the potential with the OT
It depends on what you’re after.
Start off by simply using standard timing and different step lengths per track.
Your better off understanding the programming of a straight polyrhythm before you get into polymeric programming.
Just remember that every 4 16th steps are a crotchet (quarter note)
So by setting one track to say 16 steps and another to 25 steps you have 4/4 and 6/4 combined.
When dealing with poly metric timing you’re usually best to start of working in easily divisible multiples 1/2 or 1/4 or x 2 for starters so that you can get something that is half the speed, double the speed and a quarter the sped of the master tempo of another track.
A simple example of this is using D’n’B breaks at x2 on a 90BPM track so that your bass and pads for example clock in at 90BPM but your drums scuttle over the top at 180bpm - then you could create a melody playing an odd number of steps say 29 for 7/4 and slow that down to 1/4 for 45BPM riff.
These will loop around each other and essentially have different points of rotation so they will sound similar but not always the same on each repeat.
The 3/4 and 3/2 divisors act like triplet 8th and triplet 16th sub divisions per quarter note and don’t always gel well with a straight up time signature.
Generally on a unit like the OT I’ll keep working on the basis that each for steals in a quarter note and work out how many 16th steps I need for 4/4, 5/4,6/4 etc I may need. By halving it each step lasts for a 8th note (quaver) and doubling means each step his a 32nd note (hemisemiquaver).
By using my 2 Monomachines and my A4 in this way I cam quite easily get pretty intricate results very quickly.
On something like an RS7000 you could define your phrase lengths in measures and add a time signature for each phrase as well as a metric divisor which get s a whole lot more complex.
Using a more simplified method first (and technically a poly rhythm will usually be in the same tempo with a different rhythmic count) you can get very complex sounding material when you don’t place notes on obvious accent, down or up beats.
Generally I love the combo of 7/4 and 4/4 or 15/16 and 4/4 on a more complex sequencer but as stated start of simple and go from there.
This track of mine is a prime example of what the OT can do just using variable step counts per track over both the internal sampling engine and midi. It can get quite wonky pretty fast.
6 Likes
very nice, gotta love all the potential with the OT ;)[/quote]
Yup, agree on that. LOTS of potential indeed. I have also been thinking about patterns as well as the LFO designer and Arranger regarding this “method” too.
Thinking about the differencies between humans/machines. An (good) human drummer have his/hers taste in their selection of fills, beats etc. That could be seen as something “random” for anyone else (except the drummer who “knows” what he/she is about to play).
But it still limited down to 2 arms and 2 legs. Which in a way would/should define which parts of the drumkit that belongs together (as in: never hit at the same time due to the humans restrictions) in the OT. Then we have the risk of the “randomness” of anything set to random, which in a sense is counteractive to a drummers “taste”. If they´re playing a beat as similar, i e, to my example. They wouldn´t fire off anything equal to the OT´s re-trig parameter maxed out or RTIM at minimum. Unless they have an total bad taste of course ( 
). Thus, we need to narrow down the worse randomness in the OT.
This is where I´ve also been thinking about the LFO designer, perhaps it should be tied to the RTIM. Where I set a few “random” but still choosen values which should be the “taste” selections (tom/snare fills).
And the Arranger of course, to set the song structure as a whole. If grid size is big enough (i e 64 steps), I could have the first 32 steps being the normal beats without much/any random variations. And the last 32 steps pretty much the same but with the variations included. Then arrange up the structure as needed. Easy enough to build short/long passages of any of these two “halves” thanks to the repeat and position parameter in the arranger. And the patterns could be used for different kind of beats, which you really can´t get the machine come up fully with by itself. But letting it taking care of the variations, in any of these patterns.
It depends on what you’re after.
Start off by simply using standard timing and different step lengths per track.
Your better off understanding the programming of a straight polyrhythm before you get into polymeric programming.
Just remember that every 4 16th steps are a crotchet (quarter note)
So by setting one track to say 16 steps and another to 25 steps you have 4/4 and 6/4 combined.
When dealing with poly metric timing you’re usually best to start of working in easily divisible multiples 1/2 or 1/4 or x 2 for starters so that you can get something that is half the speed, double the speed and a quarter the sped of the master tempo of another track.
A simple example of this is using D’n’B breaks at x2 on a 90BPM track so that your bass and pads for example clock in at 90BPM but your drums scuttle over the top at 180bpm - then you could create a melody playing an odd number of steps say 29 for 7/4 and slow that down to 1/4 for 45BPM riff.
These will loop around each other and essentially have different points of rotation so they will sound similar but not always the same on each repeat.
The 3/4 and 3/2 divisors act like triplet 8th and triplet 16th sub divisions per quarter note and don’t always gel well with a straight up time signature.
Generally on a unit like the OT I’ll keep working on the basis that each for steals in a quarter note and work out how many 16th steps I need for 4/4, 5/4,6/4 etc I may need. By halving it each step lasts for a 8th note (quaver) and doubling means each step his a 32nd note (hemisemiquaver).
By using ny 2 Monomachines and my A4 in this way I cam quite easily get pretty intricate results very quickly.
On something like an RS7000 you could define your phrase lengths in measures and add a time signature for each phrase as well as a metric divisor which get s a whole lot more complex.
Using a more simplified method first (and technically a poly rhythm will usually be in the same tempo with a different rhythmic count) you can get very complex sounding material when you don’t place notes on obvious accent, down or up beats.
Generally I love the combo of 7/4 and 4/4 or 15/16 and 4/4 on a more complex sequencer but as stated start of simple and go from there.
This track of mine is a prime example of what the OT can do just using variable step counts per track over both the internal sampling engine and midi. It can get quite wonky pretty fast.
[/quote]
WOW!
Thanks for your tips here, gonna go all in to understand as much as possible of this. Never heard of anything called polymeric programming. Gonna look into that too.