Aaron Isaksen wrote:
> 
> I was in the lab the other day creating some music by randomly moving a
> resonance filter.  I liked the fact that every time I played a note, it
> moved the filter in the *same* "random" pattern.  I also implemented

	Sorry to sound like a broken record (broken in a random (but not, it
seems, unrepeatable) pattern) but what you are talking about is NOT
randomness as it should be defined inside an orchestra!  You are in
effect using rand to create a SCORE effect - i.e. to create a random set
of p fields.  This is what the score is for.  If you want this
"randomness" to stay fixed for all time you should immortalize it in an
ftable and read the ftable.  Why?  Because, for instance, if you add
another instrument to this orchestra which uses rand, you will break the
pattern you are happy with.
	We are bumping up against the question of what is "fixed" in a csound
score and orchestra (and betraying some intellectual biases, at least on
my part).  If I ask for random I want to NOT know what is going to come
out, other than the fact that the number should be within the range I
asked for and every possible number should be equally likely (if I'm
asking for flat randomness, like with rand).  But this is certainly
coupled with the fact that I don't like my music to sound the same each
time I play it.  Which, I suppose, puts me in a csound minority.
	So the question is: is csound a tool for creating a single audio
rendering or is it a compression scheme for music.  In the first, it is
conceivable to design orchestras that render differently each time.  In
the second, it is not.  And I realize that I am likely to be in the
minority for desiring the first and not the second.

	Rant over (perhaps briefly)...
-- 
Mike Berry
mikeb@nmol.com
http://www.nmol.com/users/mikeb