[sdiy] Polymoog resonator question

David G Dixon dixon at mail.ubc.ca
Sat Oct 6 22:56:17 CEST 2018

> Thanks for publishing your findings, David.

You are most welcome, Brian.

> I was about to ask whether it might be better to design a 
> single module with an SVF, so that folks could chain a 
> triple, quadruple, or whatever they like.

That was my original inclination, before I delved into the Moog rabbit-hole.

> It's somewhat limiting that the Polymoog resonator has bands 
> that don't overlap, forcing users to work within the three, 
> pre-determined bands. In addition, the LP/BP/HP switch (and 
> hidden BP with inverted phase) forces all three resonators to 
> the same mode, rather than allowing independent selection.
> I suppose that the limited range allows more precision 
> through finer control of frequency with the same slider 
> travel. I assume that the shared LP/BP/HP switch is more 
> likely due to limited panel space and the cost of additional 
> switches that might rarely be used.

It would be easy enough to put a three-way switch on each resonator, and an
additional two-way switch for inverting one of the bandpasses.  This would
take up very little panel space, and would certainly be easier and more
intuitive than a bunch of routing options on a rotary switch or something.

> In light of what you're describing, three independent SVF 
> modules would perform vastly differently than the Moog version.
> I released an AudioUnit plugin that implements a triple 
> resonator, but it does not mimic the gain disparity between 
> bands (and it certainly doesn't model clipping or other 
> distortion). It might be interesting to add those features 
> and see what sorts of different sounds it leads to.

Well, I don't really want the distortion, but my simulation went very
square-wavey when I put signals through the LP mode at high resonance.  I'm
probably feeding too hot of an input relative to what the Moog unit expects.
Does anyone know what the input signal amplitude is for the Polymoog?

In any case, I believe that a more uniform gain structure would lead to a
more reliable behaviour.  Since the Polymoog never mixed the modes, the
different gain levels would never have interacted anyway, so other than
changing the overall volume, I don't see that converting the circuit to a
unity-gain structure could have any deleterious effect on the actual timbre.
I also think that blending different modes from the three resonators in the
Moog unit might not have given very good results, because the LP mode would
simply have swamped the other two.  In any case, this is all very easy to
find out by trial.  Another thing that can be done is to put each resonator
through a VCA and adjust the levels independently for mixing.  Maybe that
would be a better use for the two spare VCAs than adding VC-Resonance.
Decisions, decisions!

More information about the Synth-diy mailing list