At 16:49 20/03/97 -0500, you wrote:

>Now, if you believe that timestretching a sample on an Akai sampler
>gives you a better result than using a little imagination with table
>procedures or any number of frequency-domain procedures, then you should
>do just that. Csound (or CLM or SuperCollider or CMIX or other
>non-commercial software) is not meant for the 5-minute attention span.

Jean,

For your information I was attempting to switch between a large number of
l-o-n-g samples, and then time-stretch and/or transpose the resulting chunks.

GEN 01 wouldn't do it because there wasn't room for all the samples in memory.

PVOC was even less suitable for the same reason.

Soundin might have done it, but it didn't have the neccessary features.
(Until very recently.)

In the end I used a copy of Steinberg's Recycle, a MIDI sequencer and a
MIDI file produced with some custom code. That combination, after a fair
bit of tweaking, worked quite nicely.

>Please do not consider your personal views as the only ones that matter.
>If you are so passionate on improving Csound, lets see the work you have
>done towards that. Whinning does not count.

I'll be contributing some new spectral processing routines to the CDP
system later this year. And have been helping them design an elegant and
powerful (at least we all hope it is) user interface for the next version
of the system.

In the meantime it obviously isn't just my own personal opinion that there
are serious shortcomings.  Kudos to Eric and Barry and MIT who are
obviously familiar with the standard criticisms, and have come up with some
very exciting proposals to deal with them, and do a lot of fun new stuff
besides. 

Perhaps we can move on to getting son-of-Csound right now?

Regards,

R.