Karplus-Strong with continuous input?

[neill0]

Hi, has anyone considered implementing Karplus-Strong algorithm with a constant audio-rate input to the synthesis, rather than a simple trigger? I understand that this would allow the "string" to be continually excited and allow different sounds to be achieved through different attacks.

I think the Kalichord Strum instrument uses something like this:
"Signals from a piezoelectric pickup on each tine are used as excitation signals for a digital physical model of a string, creating a nuanced and expressive plucking interface. " - http://blog.bela.io/2017/05/15/kalichord/

and (of its predecessor):

"In the case of the Kalichord, a set of piezoelectric films can be plucked to drive a simple Extended Karplus-Strong model" - https://quod.lib.umich.edu/cgi/p/po...e.pdf?c=icmc;idno=bbp2372.2014.089;format=pdf

Thanks for any info - I'd love to try to build something like this instrument using Teensy!

 

[PaulStoffregen]

That's an interesting idea.

I wonder how it would actually be implemented? I mean, the continuous input is audio rate samples, but the trigger event is a big group of samples suddenly placed into the Karplus Strong circular buffer.

So would the incoming samples just add to the buffer? Or would certain types of changes trigger dumping big chunks of white or shaped noise samples into the buffer?

 

[oddson]

Isn't that just a tuned resonator? It's just audio fed into a delay loop... but the delay loop's length is controlled via midi.

I played with KS extensively with SynthMaker years ago; including tuned inputs of varying length and shape and filters in the delay feedback line but always triggered by MIDI.