[CM] 2nd International Linux Audio Conference

Rick Taube taube@uiuc.edu
Wed, 28 Apr 2004 13:02:52 -0500


i think for the term "real-time composition" to mean anything then the 
composer that follows this model should require exactly 60- minutes to 
write a 60 minute piece and never go back and change their mind about 
any note (or whatever) that they claim they are composing durring that 
time.  Not even Hydan did that! And if they are not doing this then 
there should be another name for what they are doing. I think 
composition is about makeing something as good as it can be --  and so 
i dont see how it is possible to remove reflection, interation, 
revisitation and refinement from what composition is. These all involve 
going back and changing something such that -- if it ever happens -- it 
will be better. That is not real time, or if it is then the term means 
nothing. I think its only random chance when composition is in real 
time. Sometimes the composition process is "faster than real time", 
like the sudden moment in which you understand what the next minute of 
the pieve will be. Usually is slower, like when you actually make that 
minute be as good as it can be. For me this process is much more likely 
to involve hours or days than it is to take a minute. It once took me 9 
months to write a 22 minute piece, that's a 388800/22  compute ratio!