Company response to Generator 1.5 query

[Sean Costello]

Chris (the Hex) wrote:

> That may be true, but I don't think Buero was asserting that computers
> don't age as quickly as synths.  What I interpreted this statement to mean
> was that because a software synth isn't tied directly to a specific hardware
> configuration, it's possible for a software synth to grow as your computer
> hardware grows instead of becoming obsolete along with it!

This may be true, but I really doubt that most of the software synths of
today will work on the computers of ten years from now.  Can modern
Powermacs run Turbosynth, for example?  Maybe they can - OK, I really
don't know what I am talking about here. :) Still, it seems that the
commercial synth programs have less of a chance of running in the
future, on future hardware, than academic programs like Csound or Common
Lisp Music (these are closer to programming languages than simple
programs). As processors change, programs that are designed to exploit
the strengths of various processors will probably have to change in
order to be able to use the new technology.  The state of the art
computer music systems from 1990 ran on NeXT computers, and used the
built in Motorola 56001.  These programs can't be easily ported over to
run on general purpose processors without a whole lot of rewriting
(although Common Lisp Music seems to have solved this problem nicely).

It is frustrating when you realize that even the powerful
microprocessors of today can easily be topped by a digital synth from 15
years ago.  When running a real-time FM orchestra in Direct Csound on a
Pentium II 300, my computer starts coughing at around 6 voices.  The DX7
can output 16 far more complex FM voices, no problem, and you can buy
one for $300 at any given pawn shop.

Where is the $1000 music supercomputer?  I could use one right now!  I'm
compiling my end-of-quarter projects for my computer music class.  The
one-minute composition I was working on today took over 10 minutes to
compile on a tricked-out Pentium II 400 Linux box.  Mind you, this is a
drastic improvement over the mainframe computers originally used for
Csound, and the techniques I am using (granular timestretching, digital
reverb, digital filtering) are computationally expensive.  Still, it
will be a long time before I can get these sounds live in realtime.

Blather blah blah,

Sean Costello