Korg Forums

[OpenWindowUnit]

I was playing around with modulation of the OSC parameter on a one bar loop and came up with a really crude form of time stretching. Apologies if this is old news, but I found it rather interesting and hadn't seen it mentioned before (except for comments that the sampler is incapable of time stretching). See the link below for an example.

I took a one bar sample, set a loop point of 0 and switched the part to Mono2. I then selected modulation '39 SawUpB OSC' with a speed of 100 and an initial depth of 0. Oscillator edit knob was set to 0% FWD. Holding down the pad plays back the loop at normal speed, but increasing the mod. depth gradually slows the sample while preserving pitch. With the depth at 127, the speed can further be reduced by increasing the oscillator edit parameter from 0% FWD to 98% FWD. At the highest percentage you end up with a virtually infinite sustain.

The mod. speed parameter acts like a grain size. At 127 it sounds pretty good, but I found the sample didn't always loop reliably at that high a setting. I think the loop point was occasionally being skipped so the sample would play out and stop.

Haven't figured out how to speed up the sample, only slow it down. Well, you can speed it up, but I forgot how I managed it. Basically played the sample while gradually increasing the oscillator edit parameter throughout the bar. The looping grains progressed through the sample faster than the original speed, but it didn't sound pretty.

Thoughts?

https://soundcloud.com/openwindowunit/es2-crude-timestretch

[OpenWindowUnit]

Here's a second example. Mod. speed set to 127 and oscillator edit stays untouched at 0% FWD. First the sample is played unaffected, then I retrigger with mod. depth at 127 and slowly rotate the knob to speed up to regular playback, and down again. Then up...

You can hear that re-triggering initially played a small portion of the sample in reverse before flipping to the forwards direction. This can be avoided by setting the oscillator edit parameter to 1% FWD before triggering.

Seems a bit glitchy, but I'm still trying to wrap my head around what the LFO and modulation is actually doing! Definitely usable though.

https://soundcloud.com/openwindowunit/es2-crude-timestretch-2

[colulizard]

that looks very interesting. And the sounds was quite weird when slowed right down.
_________________
Korg MonoPoly, APC 20, Ableton Live, NI Massive, Bass Station rack.

[colulizard]

I just gave it a go.. you get some very odd sounds. I pushed it through the grain delay mfx as well (at a low setting) for extra weirdenss
_________________
Korg MonoPoly, APC 20, Ableton Live, NI Massive, Bass Station rack.

[Redmond Barry]

This is awesome! Thanks for posting. I need to dust off my ES2 and get back to learning it.

[volcahelper]

Yeah this is great, thanks so much for sharing! Can't wait to try it out.

[apapdop]

Excellent find, nice one!! More weirdness potential...
_________________
If I'm not listening to music, or if I'm not making music, then I'm probably thinking about music.

Volca Sample, FM, Beats, Kick. OP-1, Monologue, Pocket Operators. And an ipad.

[musicmagus]

Nice find! Now I've got to try it

[DiscoDevil]

I figured there had to be some kind of timestretching. Some of the default patterns allow you to change the tempo and the sample loops seem elastic on some level.

[gizmoismogwai]

Ah, man. I knew it was only a matter of time before this community starting figure things like this out. This is a seriously awesome and useful trick and I'm going to put it to good use tonight

[colulizard]

One "happy accident" from last night.
I was looking to programme a kick drum, and tweaked the oscilator button in error. And quite wildly too I think, and I ended up with a breakbeat instead, which started on a nice sounding kick
I adjusted the release setting so that instead of the entire break i just got a nice syncopated ghost snare and hat sneak in as well, which ended up sitting quite nicely in the pattern i was playing with.
I suppose it's doing the same thing as "time slicing" and then picking a certain slice, but by playing with the start time of the sample I can also get different sounds without auditioning loads of different samples.
I'm assuming this is actually what other people are already doing, but like I said... for me a happy accident.
_________________
Korg MonoPoly, APC 20, Ableton Live, NI Massive, Bass Station rack.

[OpenWindowUnit]

Thanks guys, I'm glad to be giving back to the community!

@colulizard That's a great technique. In fact, I was trying to achieve something similar when exploring modulation of the oscillator edit parameter.

My ultimate goal was to trigger different start points of a breakbeat sample throughout a sequence while only using a single part. In theory this is achieved using motion sequencing, and there's even a section of the Step Edit function specifically for OSC edit. In practice however, I've found it very difficult to get a start time value that perfectly coincides with the beat of the sample. Has anyone noticed that there's no FWD 50% value? It's either 49% or 51% so something is a bit wonky there. Oh, and whoever mentioned that you HAVE to have some recorded motion sequencing on the OSC Edit knob first before tweaking in Step Edit is absolutely right! Tweaking the values first will have no effect.

That leads me to the Time Slice function. I think it's really good, and contrary to my initial thoughts it isn't simply doing a division by time. It's detecting transients for slice points so I think it should be called 'Transient Slice' instead. I also like the ability to add/remove individual slices by holding shift and hitting a pad. Found this out by accident although it is documented in the Parameter Guide. I love being able to play slices on the pads in any order while time slicing, but it's AGONIZING that this cannot be programmed into a sequence on a single part. There should definitely be the option to map slices to the keyboard.

[colulizard]

[apapdop]

[sqz]

I figured out 3 other interesting ones.
The examples below are using the same technique as the original post (except I was using triggertype mono1)

RANDOM RYTHMIC PINGPONGLOOPS
modulation type: S&H BPM Osc
moddepth: 100
modspeed:81

result: beat plays but occasionally pingpongloops different parts of the loop, and so on. Increasing moddepth increases the probability of pingpongloops. Sometimes there's an occasional silence (pingponglooping end of sample?).

NEGATIVE STARTPOINT / REVERSED PRE-ROLL
modulation type: EG+ Osc
moddepth: 127
modspeed: 0-127 (= 0-100% negative startpoint)
osc edit: 0% forward

result: for modspeed 0, sample plays as expected.
However, when set higher (20 e.g.) the samplestart is shifted further, but plays back in reverse until it reaches the osc edit-value, and starts playing as expected from there.

Example: "one two three four" becomes "eno-one two three four"
Could be very nice to offset a percussionloop on the fly, or to turn 4/4 loops into polyrythms.

OSC EDIT WITHIN OSC EDIT
modulation type: EG+ BPM Osc
moddepth: 0-127 (=OSC edit)
modspeed: 0
osc edit: any

result: ability to scrub within a sample which startpoint has already been manipulated by osc edit.

Example: a sample "one two three" which (after setting osc edit) was trimmed to "two three" can still be trimmed to "three" by setting moddepth.
_________________
Korg Electribe SX Upgrade v3:
https://www.youtube.com/watch?v=iYAPfE6z0MI
A naysayer sees the difficulty in every opportunity, an artist sees the opportunity in every difficulty