I just started out with the Rytm last week, and instead of asking tedious questions first (there are some to come…), I thought I’d provide something to play with for you all first. The following link allows you to download 26 chains of single cycle waveforms for the rytm, all sliced divisors of 120. You can select slices and modulate velocity/aftertouch as with other sample chains, but here you can do this with looped single cycles. While this does not yet turn the Rytm into a fully powered wavetable synth, some sounds come pretty close.
Some of these chains are broken at the end, I don’t know why. Serum, which I used as the editor, could not produce perfect wavetables first, but the chains all are usable, for instance to create wildly different basses and drones.
p.s. The waves are now corrected, with some help form the internet and the library of Multiman. Big Like to Adventure Kid whose original waveforms have been used here.
First of all, welcome to the forums, bring on the questions !!
Now, hmm, how to tread here …
Results matter, if you like it, then that’s all that counts
A+ for sharing/intent
But, in all seriousness with a view to your getting better results in the long term, have a look at your editor and check it’s working correctly - every wave chain i looked at was full of glitches and dropouts, not to mention very questionable zero crossings etc … it looked a bit random tbh
Don’t want to sound churlish in saying this, it was intended as a constructive comment, if you put a lot of effort into doing something may as well be confident it will pay off, maybe time to revisit the process, approach or source material … i could have said nothing, but that wouldn’t be in the spirit of progress
if it sounds good, it is good … but does it sound as intended or as good as it could be, besides errors can be the making of some things
Anyway, what do i know and, again, welcome … the AR is astonishing, good call
thanks here for the comment. I used Serum’s editor to build the waveforms and exported them as 16bit waves. When uploading these with that C6 tool failed, I converted them again using Oceanaudio, which created the waves I uploaded.
I had no problem to create nice sounding drones from what I have in my AR, but I saw in OB that the end of the perfect waves chain (= 4 single cycle waved) was corrupted.
I will not let that stay things as they are, but could somebody give me a hint with what tool I can build the chains better? If Serum can not work like this, I currently have no option to repair the glitches.
Which already does a lot of what I was trying to do. Unfortunately, these contain chains with 64 waves, so not usable on the Rytm. I decided to retract my waveforms to spare people the disappointment and try to create new waveforms from these instead.
Will come back later when I have succeeded (or not)
That is not fair on you or at all what was intended. I’m unfamiliar with Serum and OceanAudio, so don’t know what they do or offer, and indeed, i don’t truly know what the chaining utilities shared on here do - there’s certainly an established one for OT and an adapted one for AR i think. When i make chains, i do them myself using my own tools, which is why i guess i was curious and i just so happened to be having a discussion this morning about dropouts from the OT whilst streaming from CF card, so my thoughts were all in this area !
I have made my own chains for fun and the important aspects i recall were getting the zero crossing point at sample start & end, ensuring the sample length equated to the root note i needed, if preparing for the AR i’d also do it in 48kHz
Here’s a snapshot of the start of one of my chains, it’s full of simple minimal harmonic content, but you get the picture, sample accurate crossings (zero level at phase 0)
But , the aspect that was most obvious in the ones i looked at was the frequent dropouts right in the middle of a low freq part there’s be a sample dropout which introduces clicks
I view this on Wave Editor on Mac, but i prepare (and generate) the content in Max/MSP
I’d done a lot of the groundwork on a true wavetable oscillator project whereby the shape is the important part (rather than the implied frequency content from the sample length) so introducing partials for a given frequency was easy, so building a chain became straightforward. It’s a little project i may have to get round to finishing and see if it’s worth my while to do so
I see you’ve retracted your waveforms, that makes me feel bad, but if you can perhaps have a look inside one to see the dropouts then work back from there to see where the issue lies, it may also be much simpler to audition glitches if the samples are (and are played at) lower frequency
first of all, don’t feel bad. I am grateful, I am a beginner, and if you tell a German that something is not 100% ok
But it seems that I have found a good solution. The wavetables I build and uploaded are mostly constructed from AKWF anyway, so I loaded the sample chains mentioned above. With mostly 64 slices, they have a length of 43.200 samples, so 60 slices are 40.500. I can cut down the sample to this length, thus converting the original into one that the Rytm can use.
I have done so with the first chain, and for me this sounds good. We would loose 4 slices per chain, but then there are more waves one could use anyway…
Does this sound ok? I’d be converting at least some 10-20 waves manually this evening, so that the fun can start in earnest…
that’s all that really matters, plus you’re clearly not a beginner
the logic is sound, but to my ‘germanic’ tendencies, those samples look suspiciously flat, at 44.1k 674 samples is closer to a C than 675, i think this is one of the reasons i discounted looking into those samples before, details, details
I really, really hope Elektron come good on a vague unofficial half-promise rumour that a new LFO destination would be added [Samp Start & End] … this would make ALL the difference
FTR I really strongly dislike that setting the same value for start and end does not momentarily silence the sample (at least as an option) -so for now, the best way to simply sequence ‘wavetable’ scanning is via velocity mod
A way to constrain (or auto snap) to a user selected divisor of 120 would also be a boon, but that’s at best a distant pipe dream
hi, today is my first day with my new rytm. (yeah!)
i also got a pack (not this one) of single cycle waves on my rytm (just very tiny pieces of clicks) i understand to use them you have to put on loop mode BUT my problem is that the sample dont stop anymore. even if i parameterlock the last step with “loop 0” it wont stop looping the sound. even if i press [STOP] it will go on. it just stops if i press [STOP] 3 times in a row.
what i am doing wrong?
also if i mute the pad with the SCW the sound wont stop
now contains an update. 26 single wave chains. The samples might not be 100% at C as the original AKWF waves were at originally in D with +2 cents detune - nothing you can’t fix with the sample detune in the AR, but this might an explanation.
plus you’re clearly not a beginner
Started that whole thing a year ago, first ITB in reaper, then iPad, then Elektribe 2, then Ableton and Push, now AR. Head is sometimes dizzy. Love the sound design part of electronic music, though.
That they could do by not using a sample end, but a sample length (that could be negative). For compatibility reasons, the loop switch could have 4 settings: 0: no loop, 1: loop with end, 2: no loop with lenght, 3: loop with length.
You would still have to set the length yourself, but we would get wavetable scanning with that LFO on sample start, right?
The OT works in the position / length way … i think there’s a reason why it makes sense the way it is for the AR, a dual destination is not without its considerations, but it’d be relatively easy to make it robust
Hey Thanks for these I have it up and running with Vel > Samp Start and End doing some modulation.
Accuracy of the wavetable Mod is all over the place but at least it works.
I guess i have to reduce the depth of the mod and also using the pads isnt a precise method.
There are clicks and some buzzing loops sometimes so I’ll have to figure out if its the modulation parameters not hitting zero-crossing points or an issue with the results of the modulation parameters not being great or not working well with specific SCW’s
I am sorry that there are some glitches, but I too haven’t found out yet where these come from. In theory, all single waves should zero cross once you have selected the right length. All I did was to shorten the 64 wave chains to 60, so not much there that could have gone wrong. I also have not the best results from using the velocity or aftertouch wavetable mod, some thing is off there.I guess the best bet would to create chains with exactly 120 waves in them, but I have not managed to get there yet.
Hi, maybe you could help me? I was looking for a solution to stop the loop (and keep FX tails) when stopping the sequencer but haven’t found any other than turning the loop switches off manually. It’s there some setting for loops to stop with sequencer stop?