Shifting veils — generative patch from my Nov 20th livestream

Shifting veils — generative patch from my Nov 20th livestream

This patch uses sample and hold shift registers (explanation in Tips and Tricks) to create slowly evolving sequences. The shift registers randomly sample from a note bank that uses a minor scale, with an emphasis placed on roots and octaves, then pentatonic notes, and finally some options for seconds and sixths to sneak in.

The voices are pretty simple sine waves, with modulated duty cycles. The voices are panned left, center, and right. The left and right voices uses 16-step registers; the middle voice uses an 8-note register. (If you want to play around with how many steps a register uses, the output of the last sample and hold goes to a switch immediately after the sample and hold bank. So, if you wanted to change a register to 13 steps, for instance, you would disconnect the output of the 16th sample and hold from the input of this switch and replace it with the output of the 13th sample and hold.)

I create a control page:

Each voice has a dedicated clock divider (I generally like running all of them at the same speed, but there are interesting options available by trying other ratios). There’s a tap tempo pushbutton in the bottom right corner.

There are controls for root note for each shift register (again, I like to keep these all the same -note- but I do change the octaves). There is also a control for phase shift, so you can use the panning of the voices to create interesting stereo spreads (by default set to -240, 0, and +240 degrees).

The instability control determines how much the sequence evolves: more instability means less repetition. 0 instability will lock a sequence in place. Trigger chance determines how likely a given voice is to have its envelope triggered.

Octave chance determines how likely a voice is to jump up or down an octave on a given beat.

The middle voice has an additional control: hold chance determines how likely it is to hold on a given note… I named this one poorly, though. Succinct names are hard. Basically, with a value of 1.000 it will always advance to the next note, with a value of 0, it will hold the same note forever.

I do these livestreams at ungodly early hours on Saturdays (usually beginning around 8 AM US Eastern), if you ever want to stop in.

I restarted my Patreon, if you want to check that out:

  • Platform:
  • Category: Composition Sequencer Sound Synthesizer
  • Revision: 1.0
  • License: Creative Commons Attribution Share Alike 4.0
  • Views: 211
  • Modified: 2 months ago
Chat about this patch on Discord! Download (119)

Leave a Reply