Skip to Content


1 reply [Last post]
Joined: 11/17/2015

I'm looking for an enviroment to test an idea of mine.

it's based on the idea that using normal variation is a good idea.

and further that a variation of let's say +/-32 samples is not the same as a variation in the range of +/-32 samples.

so my idea is to playback a sample with a different offset every hit with the help of a good interpolation to have results beyond what is possible samplewise (44.1kHZ)

while looking for source code for a creating normal variation, I found some code.

it's quite complicated and if applied to 32 or 31 or 30 .. or -30 or -31 or -32 samples, it's a waste of resources (it's better look up the value by assigning chances to a equally variied random number), but if applied to a soild interpolation it should yield the best results one can expect from computed randomness.

I hope you'll like the idea.. as I'm german, it might sound strange, but I really thought about this..

feel free to ask.

I'm also curious about your ideas,




in my experiments, using a discrete number of samples to do the humanization might lead to repeating clashes of sounds (some combinations stick out and repeat, if the sounds used are very percussive - even worse if you play the same instrument twice..)

Joined: 04/17/2010

Hi wildjamin,

how have you been with setting up an development environment ? There is some documentation about that in the menu here, I hope your found you way.

Please, tell us.