Wavetables, wavetables

[Don Tillman]

   From: Roman Sowa <rsowa at WizjaTV.pl>
   Date: Fri, 30 Oct 1998 17:54:29 +0100

   I was thinking about simple waveforms like sine, sqr and triangle
   which wouldn't suffer from aliassing.

Aliasing has to do with the harmonic content, and a square have has
lots.  So your square wave would alias all over.  

But really, what's the point of going to the trouble of looking up the
wavefrom from a table in memory if you're only going to use "simple
waveforms"?  

----------------

And since the topic of wavetable oscillators was brought up, I'd like
to know why folks think wavetable oscillators are a good thing.  Yeah,
it's theoretically nice to have any waveform you'd like, but the human
ear can't really hear waveforms, and any waveform with static harmonic
content is just a waveform with static harmonic content.  The exciting
stuff happens when the harmonics are non-integral or the waveform get
manipulated or modulated.  For instance, a pulse wave isn't very
interesting by itself, even though it's harmonic spectra is pretty
complex.  But PWM is a blast.

So I'm thinking if you want a wavetable VCO, it would only be worth it
if you had some hooks to do something interesting with it.  For
instance... 

Memory is cheap, perhaps a three or four dimensional table loaded with
waveforms that morph to each other.  Or a second table that provides a
parameter to modulate the pitch of the oscillator.  

Or a "sliding waveform", where starting point of the waveform can be
modulated and the waveform is designed to do something interesting
with that sort of change.

I'm unfamiliar with wavetable VCOs in commercially available synths;
are there any cool manipulation or modulation features?

  -- Don