Support and discussion for all of Akai’s modern standalone MPCs including the MPC X / X SE, MPC Live 1, 2 & 3, MPC One / One+, MPC Key 37/61.
By clausi Fri Mar 20, 2026 11:44 am
MPC-Tutor wrote:If you join your samples into a single sliced file (sample chain), assign only to layer 1, set layer play to velocity and then modulate slice index in the mod matrix using 'random 1' you will achieve true randomness. If you record a sequence with this, it should play differently each time.

Creating a sample chain within the MPC is however a bit convoluted, but basically create an audio track and import the samples one by one into the audio track (uncheck 'place at start). Arrange the regions efficiently to minimise unnecessary gaps. 'Bounce to sample' and chop this bounced sample. Delete the audio track (no longer needed).


Sounds good, even if as a bit of a workaround. It would be easier if I could just set layer play to true random. But I will try this out, thank you.
User avatar
By MPC-Tutor Fri Mar 20, 2026 12:03 pm
Another option is to put your round robins on different layers (as usual) but set layer play to crossfade type D. Now modulate 'crossfade x' using random 1 (use 'max' if you need to limit the number of layers to be used). Quicker to set up, should be truly random.

The big difference is that you are now randomly 'crossfading' between layers, so as well as sometimes getting 'pure' layer 1, layer 2, etc, you'll get crossfaded points between layer 1 and layer 2 etc (e.g. layer 1.7). This could be great as it could effectively achieve unlimited variations, but won't work if, for example, your layers have different pitches. Potentially some samples might phase. But worth testing it though.
By clausi Fri Mar 20, 2026 12:09 pm
T-Monk wrote:The whole random “seed” thing was added for people who wanted randomness that remained next time they played the sequence. If you want truly random maybe just use automation and randomly put numbers in. Personally, I think you may be too caught up in the intellectual side of it. Anyone listening will either hear it as random (if they are different samples) or perceive the slight different sound as “analog” and not really pay that much attention if it is truly “random” or not.


I don't agree at all. Whether listeners will perceive the hits in a loop as truly random/analog, or as the repeating pattern it really is, entirely depends: on the source material, the arrangement, etc.

The seed feature is great to recreate and 'keep' randomly created stuff. But it seems like a totally unnecessary restriction to me.
By clausi Fri Mar 20, 2026 12:10 pm
MPC-Tutor wrote:Another option is to put your round robins on different layers (as usual) but set layer play to crossfade type D. Now modulate 'crossfade x' using random 1 (use 'max' if you need to limit the number of layers to be used). Quicker to set up, should be truly random.

The big difference is that you are now randomly 'crossfading' between layers, so as well as sometimes getting 'pure' layer 1, layer 2, etc, you'll get crossfaded points between layer 1 and layer 2 etc (e.g. layer 1.7). This could be great as it could effectively achieve unlimited variations, but won't work if, for example, your layers have different pitches. Potentially some samples might phase. But worth testing it though.


Another great idea. Thank you so much for the suggestion!
User avatar
By MPC-Tutor Fri Mar 20, 2026 12:20 pm
The mod matrix is incredibly powerful, basically if you think of a problem that needs solving and spend enough time mulling over it you'll normally found a solution (especially combining mod matrix with Qlink macros). Id love to make a course for it, but it's super geeky stuff, I don't think many people would be interested.
User avatar
By Lampdog Fri Mar 20, 2026 6:20 pm
MPC-Tutor wrote:The mod matrix is incredibly powerful....

It REALLY is and seems to reward deep exploration.