ls srcpre reqs

audio fro input or playback
wave + fft + other



create objects from the audio wavs or fft
manipulate them with forces - the audio itself + transforms + fractals
artifically create a low quality transform algorithm
give them a lifetime - maybe auto to preserve frame rate
animate them - line changing length, or growing into a geometic shape


control the number and persistence of the (dots/sparks/chips)

1st step: collect live audio data into 2 point arrays and draw it
how do we deal with the screen size??

we draw to the output windows size: points are in output window coords
not resolution independednt:
will probably need to fix this some time


1st example: make a colourpolyline of centre, 256 points
frame budget= 2k points, so make it have a life of 8 frames max
ways to extend this- less lines, just points, 

a connected series of lines
a constellation of dots
a series of shapes that grow or change or interact

are these necessarily different things?
what do they have in common?
- audio data that is stored over a few frames

start with the dots one.

how to make sh!t look good.

how to draw the dots.. will this ever really look good, are there realistically gonna be many


we like drawing lines.

we could draw waveforms in loops rather than across?

we could draw fft bars

we could graduate the colours

make a realistic breakdown between technicalities (ie audio, mapping) and content