[no subject]

Notes:
	Stage 1:
	Original recording was set up using the tempo of a sine wave
oscillating at 1/2 pi, aprox. 94.2475 bpm. Starting with the first 2 digits
of pi I would vary the time signature. So the first bar was played in '3'/4
for '1' bar, the next setting was '4'/4 for '1' bar, next being '5'/4 for
'9'  bars, until I hit the 3.14159 second mark in the song. Once the
structure was set up I used the circle of Fifths to derive the melody. Each
time signature change was accompanied by moving around the circle based on
the next number in the sequence of Pi. So for the first Bar I move 3 Fifths
from C clockwise on the circle, to A, the next moved 1 Fifth to E, the next
4 fifths to A flat, etc. All of the instrumentation was created using FM
synthesis on Sine waves. Ratios and modulation were all set up based on
3.14159. Envelopes used initial 1/2 pi tempo.

	After all of this, It was determined the original piece sucked :)
(Boring, structure was not apparent, melodies were weak etc).

	Stage 2:
	I took the original piece and loaded into a custom built sampler
(using reaktor in this case, for speed, and UI). The start point in the
sample was derived from a table loaded with the first 128 digits of PI (a
value of 9 would play the sample 90% of the way through the song). The table
was scanned with an oscillator running at divisions of PI. Based on the
initial songs grid I would trigger 3 different versions of the song, with
each tables oscillators scanning at different speeds. I could 'freeze' the
table so that the original song would just loop on its current position, I
could also change the phase of the Oscillator scanning the lookup table so
that I would not get the same 10 start points over and over again. Loop
lengths on the sample were just divisions of the internal clock oscillator
running at 3.14159 Hz. I spent an afternoon tweaking the instrument and
learning to play it, and ended up with 8 3.14159 minute variations. I picked
the best of the bunch in my mind, normalized it and converted it to mp3.

------------------------------