Some ideas...

["Jason N. Joseph" <73311.213@compuserve.com>]

Hey folks, I've *greatly* enjoyed all the discussion of late, and
have decided to get y'alls (sorry but I am in Lubbock, TX) input
on some ideas I've been cooking up.

Pardon if this seems naive in the ways of loopage, but I'm still
very much captivated by one of the simple ideas Brian Eno
cooked up in making that first looping delay system for Fripp,
and that Eno still uses quite a bit (don't know about Fripp):
basically different loops of differing lengths played over 
each other such that the ways they will interact musically
is unpredictable. I've had my Jamman for awhile, but have
only recently began to explore this via recording one loop
to one track on my DA-88, then another to another, etc., and
listening to the end results. The problem with this is that
there is not as much spontaneity as having them loop over
each other live (I'd *love* to see a looping machine do 
that).

Anyhow, on expressing my frustration at the limitations of this
to my "non-musician" but experimentally-minded friend Dave,
he came up with the following, which I'd love to get everyone's
thoughts/reactions to:

>Ever play the game "Life" on
>the computer?  Not much of a game, really.  You have a 2-D grid and in
each
>square of the grid you can place a marker representing a cell.  Place as
many
>cells as you want wherever you want, then hit "go" and watch the results.
>The cells either live, die, or reproduce based on certain rules. 
Something
>like, if in the 8 squares surrouding square(x,y) there are two cells, a
cell
>is born in square(x,y).  Or, if in the 8 squares surrounding square(x,y)
>there are four or more cells, the cell in square(x,y) dies.  Usually the
>population of cells dies off after a few turns.  I guess the goal of the
game
>was to get a population of cells that lived a long time, but people found
out
>that there were ways to place the cells such that the population would
>oscillate between one pattern and another, or some placements would turn
into
>small groupings of cells that would oscillate and fly off the board.  Got
it?
> 
>
>OK, my idea is to do something similar w/ music.  Start out with four (for
>this example) simple drones(or sound loops or whatever) and periodically
>measure some measurable variable of the drones (some examples: 1) number
of
>frequency components in the drone above/below x hertz 2) delay of the
drone
>above or below its starting value 3) reverberation of the drone 4)
distortion
>in the drone 5) volume).  I don't even know if some of these variables can
be
>measured but you get the idea.  Anyway have the drones react to one
another
>based on some set rules, like if drone1 has x number of frequency
components
>above 5kHz then adjust the distortion of drone2 by some function
represented
>by f1(x).  Or if the delay of drone3 is y seconds, then add f2(y)
frequency
>components to drone 4 whose values are given by f3(y, f2(y)).  Maybe in
>addition to rules to alter the drones you have rules governing the death
and
>birth of other drones.
>
>What you'd have instead of a few sound loops with different periods where
the
>interest would be to see how they recombine to sound different at
different
>times would be sound that actually react to other sounds and can give
birth
>to a sound or hook up with another sound to kill a third sound off.  Sure
>it's pretty much impossible or at least super-hyper daunting but it sure
does
>make ya wet yer pants with the possibilities.  
>

Anybody think this is remotely possible with the devices we have today?
Any other thoughts on how to put some fresh, (relatively) unpredictable
or at least systemic experimentation into looping?

Thanks in advance for your thoughts.
jj
jj1@compuserve.com