[CM] leviathan etc.

[Anders Vinjar]

    > Bill's granular-based time stretching used in Leviathan is one of
    > the first and the most elegant solutions to changing the time
    > scale of sampled sounds in the time domain without changing the
    > pitch.

Perhaps worth mentioning also "9 Beet Stretch", a 24 hr. long piece made
together with artist Leif Inge in 2002, based on a recording of
Beethovens 9th as source:

  https://en.wikipedia.org/wiki/9_Beet_Stretch

The original version used a CLM-instrument, "AGS-gran.ins", inspired by
Curtis Roads article "Asynchronous granular synthesis".  When re-made in
2002 i based it on the regular CLM granulator-ugen.

The piece took a full night to render, probably on a Next Cube?  Or
perhaps an early SGI?

While experimenting various techniques besides granulation were tested,
FFT-based, tools from other packages, none came close to what i achieved
using CLM.  Csound was faster, but with very bad results.  Probably the
division between synth and score, and buffer based processing? - made it
difficult to get the necessary low-level control from grain to grain?

I think the lower sampling rate at the time was a limit to very small
hop-sizes, could possibly be used for an even smoother stretch?

Would be fun test once more with other tools - perhaps librubberband,
some phasevocoder approaches?  I rather like the kind of chorus effect
coming free with the granulation approach.  The others i tested at the
time made either a much too synthetic/comb-like sound, or the
'granulation'-quality was too present.

-anders