Interest Jurgen's Scanning Interpolator PCB?

[jh]

 >IMHO this
 >version seems much more attractive, as I have been told that with 
 >moderate matching of the trannies, one can achieve satisfying results.

I really appreciate Don's variation of the scanner, but "much more attractive"
hurts. The original version does not require transistor matching at all, btw.

> There's no reason why the VCAs wouldn't be DC-coupled. 

They *are* DC coupled. The schematics clearly show this, don't they ?

>Indeed, this is
> required for the type of use I'm looking forward to. You see, by feeding
> offset voltages to the VCA inputs (most conviniently done with vertical
> sliders) and scanning them with a VCO output, we have total control of
< the resulting wave. However, the VCA inputs can be used for any type of
> signals (not just static voltages) so there are actually no limits to the
> amount of animation we want to apply to the output. Further more, the
> scanning signal doesn't have to be a simple repetitive wave either.
>
> Actually by feeding in an audio signal we have, just by the sliders, full
> control over the "transfer curve" of the dynamics in the signal. I would
> certainly like to try, what kind of distortions are you able to do with
> something like this? 

Yes, these are some possible applications. I've written a long description
of this stuff, which you can find at my homepage.
As for the sound in "waveshaping" mode, I have not made direct comparisons,
but the "folding back of corners" should cover much of the Serge wave multiplier
stuff as MC has displayed them on his page.

> Then again, with much more inputs, 

Consider a design with *less* inputs as well. IMO even a 4-stage version would
be quite useful. With more stages, it's easier to set up completely different
"ranges" (see description at homepage for details), but 4 stages with dynamic
CVs will already offer more possibilities than you possibly can explore.

> It's a pity that in the traditional design, there doesn't seem to be
> any simple way to gain control over the fading curve of each step, this
> would make things even more interesting.

Don't overestimate this effect. The transition from abrupt steps (comparators)
to picewise linear (PWL; my circuit) means one integration, or 1-pole
filtering if this analogy is allowed. IMO a similar change in sound would only
be introduced if we went from PWL to cubic splines. But these are completely
out of reach, and you might be better off with a tracking filter after the scanner.
Don's variable transitions should have advantages for slow scanning of audio
sources, though. 

 JH.