Tuesday, August 22, 2006

Jamming with the slicerzoid

I am lazy. Instead writing a new patch or whatever from scratch, I have pulled out an old one of mine: the slicerzoid. It is designed to work in live situations. I am going to try it out with Luke Digance (see his blog Eclectic I) on Thursday's improvisation class. He wanted to have a more rhythmic / beats based approach than the patch I used last week (called Jeckyl). Hopefully this will suffice.

The idea is pretty simple:
  • There are 9 slicerzoid instances
  • Each slicerzoid has a ratio assigned to it. This controls the tempo.
  • These tempo ratios (multipliers) are 1, 2, 3, 4, 6 and 8. There are 4 instances of the 4x ones. There are only one of each of the others.
  • Inside of each instance, there are a number of controls.
  • First, there is a simple step sequencer with a maximum of 16 steps.
  • Each step can be 1 of 8 variants, or can be a 0 (for silence)
  • Each variant has an independent point for sample playback start
  • Each variant has an independent pitch ratio for sample playback
  • Each step in the sequence has an independent volume
  • An amount of clipping distortion can be set for each slicerzoid instance independently
  • Make sense? Now here are the interesting things...
    • Each slicerzoid shares a single sample for all of its step sequence variants
    • Variety in sound is achieved through user selection of a playback point for each variant, and stringing variants together using the step sequencer
    • Each slicerzoid can be set into 'selfer' mode, where the current step sequence is slowly morphed or evolved, step by step and variant by variant, through a series of probability tables. Thus, the computer is easily able to create an ongoing change in sound, rhythm and texture, to some degree. But not so quick or drastically that the listener hears these changes sticking out.
    • Furthermore, all 9 slicerzoids share the same sample from a single buffer.
    • Quick changes in timbre and sound can quickly be achieved by recording something new into this buffer. Thus, the material feeding all 9 slicerzoids changes instantly. However, this change is not heard as a harsh one, due to a delay line with a length of exactly one bar and slight feedback.
Have a listen. Here is a short example (661KB) from me jamming with it this morning. The whole thing goes for about 10 minutes; this file is made up of a number of short snippets spanning those 10 minutes. The audio material feeding the 9 slizerzoids in this sample was me whistling and clicking. Although, this could be anything. Hopefully it will sound good with the jupiter synth that Luke has been playing.

Yeah, I know, the patch is pretty messy. It's not going to win any beauty pageants.