[sdiy] State Var Filter with Nonlinear Feedback

[Ian Fritz]

Scott --

Looks like a very interesting experiment.

> It can add harmonic content, but not across a
> wide bandwidth, rather the distortion kicks in at points
> where the input signal's harmonics correspond to
> that of the filter's Fc setting.  When sweeping
> Fc the harmonics seem to pop in, like there are detents.
> At least that's what it sounds like.

Is this because your non-linear elements clip at a fairly low voltage? (Just
guessing.) Maybe rescaling the levels so you get more gradual changes would
be something to look at.

> Maybe I need to play with more complex waveforms??  I
> used square, pulse, saw, tri and sine.  All get very
> well screwed up, but not in a way I would call musically
> consistent or useful.  Ie, you might get a cool effect
> on one note, but play a semi up or down and it's gone.

Well, what if you track the VCO and the VCF together? Then wouldn't you
always have the same effect?

> It will also go chaotic at points as well as into
> self oscillation.

That could be interesting -- can you get chaotic effects at low frequency?

> I tried several different nonlinearities, but they
> all seemed to have similar sonic characteristics.

Sometimes it pays to look closely for more subtle effects. For example, some
people will say that all multistep waveforms sound the same .. but if you
have a sound in mind that you are trying to zero in on, suddenly they don't!

> At the very least, I can convert it into a plain old
> VC-SVF, so all is not lost.  Scoping around in it,
> it appears to be doing what it should, however
> I'm just rather underwhelmed by it's sound altering
> usefulness.

Too bad it hasn't worked out the way you expected. But sometimes if you look
at things the other way around -- what can I do with what I've got -- you
come up with useful results.

  Ian