I could not reproduce the error either way (using tunk or cpsmidib);
when testing, I added these lines to the orchestra header to
initialize the controllers for the lowpass filter:
  ctrlinit 1, 3, 120
  ctrlinit 1, 2, 60
Note that you have an "always on" note with zero key number and
velocity, due to the i1 0 3600 in the score; while it may not be
related to the problem, is this really intended ? You could also
use f 0 3600 instead.

On Sunday 14 May 2006 01:07, Jonathan Murphy wrote:
> 
> Can anyone explain why the following .csd is broken (Csound tidy up:
> Floating point exception), but works with cpsmidib (see commented
> region, the only other change being the replacement of "kvel" with
> "ivel". Please note that it has nothing to do with i/kvel, k-rate
> control of amplitude is working with cpsmidib)? Additionally, why does
> cpsmidib still bend notes, even with an irange value of 0. Finally,
> what are sane ranges for the values of kmass, kstif, kcentr, kdamp,
> and why (taking by way of an example the tables used below (snatched
> from Dr Boulanger's toot, if memory serves))?