Really easy time stretching on the Digitakt

[avantronica]

lots of discussions on this technique for DT and AR before that, it’s always encouraged to search before creating new topics, but here linked are some of the others for info anyway

Tutorial: Timestretching of Amen Breaks on Digitakt

… and the Digitakt:

[dialectrics]

Ive searched quite a bit and it seems the only person doing the same thing as this is @garf (what’s up dude??) Most other methods I’ve seen have harsh discontinuity artifacts.

This is the best way I’ve found to just play back samples and keep them in sync with time stretching so I made a separate thread. Cheers

[DEEPMOSES]

I don’t really hear how this method improves on others TBH to merit it’s own topic, plenty of audible artifacts here also

[dialectrics]

I’m not claiming this is artifact free - rather that the artifacts you get are the same ones you get with “real” timestretching vs the harsh discontinuity artifacts from the “short looping time window” method I’ve seen most commonly used.

[DOEOP]

This is awesome! I’m glad you posted this in a new thread. Otherwise I wouldn’t have noticed.

[blaize]

Cool, thanks for the tips!
I used to another method working on long samples

Digitakt Unofficial Guide

Time Stretch To stretch the time on the Digitakt, I use the reverse and forward play. It’s a bit tricky and don’t work on every sound, but for long notes or pads it’s work well. Situation : You have a good sample, nicely tuned to catch this vibes you want. You set the trig but the sound end too early before the next note. You have to make it playing longer. First, set the TRIG Lenght to the corresponding lenght you want, and adjust AMP enveloppe to make sure it will play enough time. Then, i…

But I will check your asap

[teknokarhu]

On the contrary, I find the step-by-step tutorial posted by the OP to be vastly different from all of the links you mentioned. They’re all about extreme glitchy stuff and don’t go into detail of what is actually happening.

[teknokarhu]

I assume that if you set the amp envelope to have a faster decay, there will be fewer artifacts on those speeds that otherwise sound glitchy. This is also how Ableton’s warp method works in the beat mode.

[Phillip]

For whatever reason this does not work for me

It just machine guns the sample loop…

Edit: got it working when setting the BPM multiplier at 4 instead of the default 16.

[dialectrics]

Heres how to use the method to chain longer clips. https://youtu.be/rCusXxrftK4

[Padevil]

THANK YOU!

[Boxymusic]

That’s a nice trick! I’d be interested to understand the maths of the -8 and particularly 30 in the LFO.

[garf]

LFO Depth of 30 is a division of the sample’s length which is always 120. The Speed of 8 is a division of the step numbers.

The length of a sample is always 120 on the Digitakt, and the LFO is bipolar, so you have to set the sample start in the middle at 60. An LFO depth to 30 sets the sample start to 0 (seems like it should be 60 - half of 120 - but that doesn’t work). You can set the SPD to 8, 16, 32… whatever works. OP is using -8 SPD with +30 DEP. I use +16 SPD with -30 DEP and adjust the MULT BPM to double/half the speeds.

[Boxymusic]

@garf

seems like it should be 60 - half of 120 - but that doesn’t work

This is the bit I didn’t get! Does the LFO oscillate by 200% of the target?

[garf]

heh, good question! i still don’t know why.

[Jack_69]

Experienced exactly the same thing. Had a 4 on the floor Housebeat Loop. With the Multiplier at 16 it got played back 4 times as fast. With MULT at 4, tempo was correct.
I did not understand it in Theory, because the OP used the same Setup with Multiplier set at 16 and it worked out.
Does someone know why?

Edit: Because i‘ve loaded a 4 Bar Sample.
If you load a 1 Bar Sample, set MULT at 16. With a 4-Bar set MULT at 4 when using a 4 Bar Sample

[DOEOP]

That constant flashing of the trig buttons

[JPM]

We need ping-pong loop mode on DT to give us some smoother time stretch capabilities

[dialectrics]

Exactly right. Intuitively it makes sense the LFO depth would be 60 but for some reason it’s half that. Also correct that there’s many combinations of SPD and MULT that will work, and that it changes depending on the length of your source material. Like @Jack_69 figured out if your sample is 4 bars you need to divide the MULT by 4.

A quick tip if you’re doing that - you can still use just a single page of trigs by putting a “1:4” trig condition on the first trig so it only resets the LFO once every 4 bars. Also putting long retriggers on each step to up the resolution helps with playback “coherence” at extreme values of stretching, although in my experience more retrigs doesn’t always mean better timestretching.

[dialectrics]

I’m not sure what you mean by that, but there’s lots of ways to get smoother stretching with the existing functionality. One thing I want to explore more is using multiple tracks to stretch the same sample. Layer up a couple tracks with unique p-locks / retrig resolution then resample the mix and play it back using the same method. Rinse and repeat to your heart’s desire! Does make you wish for stereo sampling but we can’t have it all…