[sdiy] Wavetable info

[Grant Richter]

Here is my crack pot theory.

The two effects sound different, but I think the math is the same.

> 1) Aliasing due to frequencies beyond the sample rate (always nasty) and,
> 2) aliasing due to lower sample rate waveforms and hence added harmonics
> (sometimes good).

In the first case, with a fixed sample rate, the harmonics have gone above
the Nyquist cutoff (1/2 Fsample). This mirrors or "folds back" the
frequencies above the Nyquist point, and produces and effect like AM
modulation. Which it kind of is like.

In the second case, with a variable sample rate, the sample rate has gone
below the harmonic content of the waveform. This also mirrors the harmonics
above the sample rate, but because alias tones are moving with the sample
rate, the ear seems to take the aliasing as part of the harmonic content of
the overall waveform.

To demonstrate the idea on an ARP 2600, run one audio oscillator to the
sample and hold voltage input, and another audio oscillator to the S/H clock
input. Adjust for a timbre, then play it on the keyboard. Now put a "dummy"
plug into the sample clock oscillator keyboard CV, so it no longer tracks
the keyboard. Play the keyboard again and compare the sounds.