strange idea....

[mbartkow at ET.PUT.Poznan.PL]

> Anyway, you see this "beating" effect of the 2 grids cancelling
> and reinforcing difference patterns (if i remember my yr 11 
> physics, Newton's rings?!).
> 
> Well, I figured I was looking at a really complex filter. The closest filter 
> i could think of that might have a similar characteristic (if u could even 
> compare light and audio filters!)
> to a comb filter. Depending on the angle you look at the screen, and how 
> close you are, all manor of crazy patterns are created before your 
> eyes!

Well, this whole Moire stuff is a very clever way to obtain 
a variety of complex patterns (be it harmonic pattern like in FM synthesis,
or formant pattern like in subtractive synthesis) using very few parameters.

What is attractive to our ears is if these patterns change over time in
some (rather ordered than random) way. That was Chowning's major achievement-
he gave us a way to produce complex evolving spectra controlled by one or
few parameters, which in turn was easily controlled by envelope generators.
I think Bud's inspired idea may be realised in a simple way. Take two comb
filters and set them in series. Now, if the base frequency (the distance
between consecutive peaks in the frequency domain) in one filter is different
to the other one, interesting "beating" between the bands occur.
In case the comb filters are realised with BBDs one has to take care to 
precisely damp the clock bleedthrough as well as the spurious output spectra
components, especially at the output of the first filter in series. Otherwise,
nasty intermodulation of clock signals will destroy the subleties of filtering.
Of course, slightly changing the clock rates would most probably drastically
influence the beating pattern between the comb bands. Expect nice effects,
similar to those known from KORG/JH triple resonator.

regards,

MB