Random Music Generation

I’ve been interested in procedural music since forever, and recently created this tool that allows you to pick your favourite snippets of random music and concatenate them into a “song”.

My thinking: I am not any good at composing music, but I know what I like when I hear it. Since the computer is more creative random than humans, let the computer do the composing, and me do the picking.

An example of what that sounds like is here. two older examples if you prefer it without drums.

Implementation: since I was too lazy to figure out how to interface with an existing synth and I wanted instant feedback, I decided it was faster to write my own synth, out of nothing but raw C++. The synth features an organ-alike and a drum. The composition part then layers these using pleasant ratios, and plays them differently in stereo inspired by Knorkator’s song “Makelos” (listen on separated speakers). At the end I filter all thru freeverb.

You can try the tool yourself here (unpolished, no support, no warranties, will crash, yada yada). To use, simply let it play, and if you want to hear the next random snippet, try “new pattern”. If you find one you like, “copy to song” will remember this pattern for later. “play song” plays the song you have accumulated sofar, and “save WAV” saves the current sound. Speed affects the whole song, measures only the next pattern. In the view menu you can zoom in and out.

Here is some older procedural music by me (with different software). Here some notes on how my non-musical brain tries to understands music.