Re: AW: fractal loops (was: keeping loops interesting)

[Brian Good <bsgood@adelphia.net>]

Rainer Thelonius Balthasar Straschill wrote:
>>No, I don't think Rainer's examples are fractal either.
> 
> 
> And you're right about that. It's not fractal, and Kris' example with the
> eight tracks of mobius is not fractal, either.


I work with fractals a bit in my day job.  One way to look at them is to 
consider that you can generate them by repeated rescale/replace cycles. 
It's pretty obvious how to use this rhythmically--generate e.g. a drum 
pattern, make a sped-up copy of the pattern, and replace each note of 
the original one with one of the sped-up copies.  Lather, rinse, repeat.

You could make things even "fractaler" by applying the same method to 
pitch, if you don't mind jettisoning the twelve-note-per-octave 
chromatic scale. Play a chord.  Pick one note as the root.  Note the 
ratios of the frequencies of all the other notes to that of the root. 
Scale the ratios by a number much less than one, generating a new 
"chord" whose pitches are likely much closer together than a half step. 
Replace each note in the original chord with a copy of the "chord," 
pitch-shifted to the original root. Repeat until traumatized.

Something like this would be much easier to do with, say, csound than in 
real time with with a fretted instrument. And, like a lot of other 
applications of mathematics to music, I'm guessing that in practice this 
is a lot less musically useful than it sounds.

Brian