> Your simplification works, and so does your second solution (the one with
> k_have_trigger == 0 && ktr != 0) with no noticeable difference. However,
> replacing
> 
> > k_have_trigger == 0 && ktr != 0
> 
> with
> 
> > k_have_trigger = k_have_trigger | ktr
> 
> will constantly retrigger whatever MIDI note is being sent out, and the
> velocity will correspond to the amplitude, like the aftertouch controller.
> However, there also seems to be constant velocity put out simultaneously.

The replacement is not in the test of `if' but

  k_have_trigger = k_have_trigger | ktr

instead of

  if (k_have_trigger == 0 && ktr != 0) then
      k_have_trigger = 1
  endif

it is only for brevity because it seems insignificant for the performance.

I think that the combination `iupdte = 0.001' and `ifrqs = 120' is not good
for your case. You can also try `mediank' with koct and kamp. Another test
is to use `spectrum' + [`specfilt' and/or `specscal'] + `specptrk' (perhaps
kvar is useful, see the example in `specptrk' man page).

tito