pwl mistracking
Haible_Juergen#Tel2743
HJ2743 at denbgm3xm.scnn1.msmgate.m30x.nbg.scn.de
Tue Apr 23 03:35:00 CEST 1996
Hi!
I played around with some mistracking VCO's this weekend,
i.e. with slightly different CV scale, to get a fatter sound.
I found that just a different scale factor is not the optimum,
so I tried to "bend" my own control courve.
I fed the normal 1V/8ve CV to both VCO's (they track perfectly,
as they are 3340's). Then I fed the same 1V/8ve into my
interpolating scanner. This way I could bend the CV courve
in a piecewise-linear way (PWL), i.e. make a transfer courve
of a polygon shape. I fed the output of the Scanner into the
modulation input of one VCO. Then I adjusted the potentiometers
in a way that was pleasnt for my ear:
Almost no detuning in the upper octave, and a certain amount of
detuning at the lower octave, and some different settings in between.
It does make a difference: With just a linear scale error you usually
get too much beating somewhere (for example it's unpleasant
in the 4th octave when you have zero-beating on the top note of
the 5th octave and the desired bass detuning in the 1st octave).
With my method, you I can get very close to (my subjective) optimum.
Then I added a slight LFO modulation to the scanning position, which
brought in additional slight variation. Really nice!
BTW (for Xpander users): Is this also possible with the Xpander's
ramp generators? From what I have heard it should, but I don't know
if the resolution is good enough there for good results. Anyone
tried this?
JH.
More information about the Synth-diy
mailing list