Rings/Elements changing modal resonators harmonic distributions


#1

Hey guys, so if i understand correctly and from listening, the geometry/structure parameter in the modal resonators moves from plate responses (counterclockwise) to string (11 o’ clock) to metal and glass (clockwise).
I have a few questions regarding this:

  1. what are the actual responses on each position of the knob? I couldnt find any information as to precisely what knob position corresponds to which real world structure precisely
  2. Olivier, how did you obtain these structure informations? did you record some impulse responses and tried to mach them by hand with the bandpass filter distribution in the resonators?
  3. Would it be possible to change them via an alt firmware? Im thinking to use rings as a primitive convolution-ish device, not for time-based reverbs but for instrument bodies. Would it be possible to simulate the harmonics of say a violin body, adjusting the “absorbance” via brightness and the “size” via frequency?

Since many early physical modelling attempts relied on band pass filters (eg polymoog resonator) I am fairly confident it would be doable in a modal resonator which is essentially the same but with more bandpasses, no?

cheers
Don


#2

1 & 2.

I originally selected up mode frequency tables from several textbooks and from STK’s source code. Laid them out nicely by similarity, interpolated between them. Wasn’t great.

So currently it’s a simple parametric model that changes the spacing of the harmonics.

It wouldn’t be hard to write code that manually sets the F and Q for each bandpass filter (somewhere here https://github.com/pichenettes/modules/blob/master/rings/dsp/resonator.cc#L55). A little modification will be necessary if you want something different from the cosine amplitude modulation controlled by the position knob.

Approximating an impulse response by a sum of second-order filter is a well-studied problem. You can use Prony’s method to get the frequency/damping coefficients/amplitude of each filter, and from there convert to F and Q.


#3

Brilliant, thank you very much.
So just for clarification: the structure knob actually just shifts around the harmonics without really emulating anything concrete?
Could you elaborate upon the position knob amplitude modulation?
It just applies a cosine shape curve to the harmonics amplitudes?


#4

Correct. Though the sequence of mode frequencies of the materials or structures mentioned in the manual resemble some of those obtained by changing this parameter.

Yes. This is equivalent to the way the different harmonics are modulated when changing the excitation position of a string.