Casio CZ/ VZ/ FZ - Pro Series

--- In CZsynth@yahoogroups.com, "ezra buchla" <ezra.buchla@...> wrote:
>
> > I think I'm working toward something similar to your work. The key
> > thing I'm looking at is an algorithmic replacement for what you term
> > "linear piecewise phase distortions". I think!
> 
> hm, not sure... i'm using that term quite specifically.
> 
Hm, to you, too.  Could you give me a reference for your specific
term?  I'm eager to learn!
I'm an autodidact in this area, so my terminology may be personal. 
"linear piecewise phase distortions" puts me in mind of the
Karplus-Strong algorithm, and the CZ's Phase Distortion methods.  I'm
focusing my efforts at the point of generating waveforms, which is
what I see the CZs, VZ's, etc., as doing.  I still think in analog
terms, so I think of filtering as something that happens after the
generation of the waveform.  I realize that, in the digital domain,
this may be an artificial distinction, but I think it is a
conceptually useful one.
> > I'm avoiding the emulation of Butterworth or Chebyshev filters (no
> > prejudice against them) for what I see as the more fundamental
> > waveshaping functions of the CZs, VZ's, Yamaha DX's, TX's, and Synergy
> > DXII's.
> 
> yeah, i'm also talkin about memoryless transfer functions
> (waveshaping). very simple. a chebyshev polynomial is just an
> interesting set of orthogonal polynomial bases which have many useful
> characteristics, especially when used as waveshaping functions (they
> produce specific harmonics of the input sinusoidal components.)
> 
> (if you're on mathematica there's even a reserved function:
> ChebyshevT[n,x], to evaluate the cheby poly of order n given input x.
> i use matlab or scilab, personally, but there you are...)
> 

I aspire to using 10% of the functionality of Mathematica!

> chebyshev filters are something else entirely... simple analog
> circuits, pretty complicated and expensive to implement digitally. the
> name comes from the fact that the chebyshev polynomials can be used to
> calculate the amplitude response, but the actual toplogy is pretty
> arcane. as far as filtering goes, i tend to think that the butterworth
> topology is more musical (less ripply), unless you really need a sharp
> transition to the stopband for some particular purpose. anyway yeah,
> i'm not making a filter.
> 
> indeed, the cz architecture's emphasis on memoryless waveshaping to
> produce filtering effects was genius, way effective use of computing
> capacity before 'virtual analog' was a gleam in anyone's eye...
> 
> here's an interesting patent describing (i think) the cz synth
> architecture and implementation, including the phase-reset and
> smoothing stage which i haven't mentioned yet. pretty cool.
> 
> http://music.calarts.edu/~ebuchla/phase_distortion_patent.pdf

Huge thanks for this link!!!
20 years ago (or so) I called Casio to see if they could tell me any
way in which I could determine by what method the audio spectrum of
the CZ'z waveforms were being generated (I may have referred to it as
a "transfer function").  I made it known to them that I just wanted to
 be able to predict the spectra of waveforms which were generated by
the available parameters.  They said it was a secret!  It didn't
remain so.
> 
> anyways, back to work...
As you said, back to work!

- synergeezer