[sdiy] Dithering 8-bit waveforms (was Re: PPG or Prophet VS waves for class)

[Olivier Gillet]

> I'm not sure how adding noise to a short wavetable of say 256 samples can
> de-correlate the quantisation noise.

I have never claimed that what I was doing was truly decorrelating the
noise. I just aim at changing its spectral shape. Actually my mistake
is to call this "dithering" while the technical term is more
"noise-shaping" (though both often go hands in hands). What is
happening here is that the quantization noise stays cyclic of course
(it's in the wavetable), but its shape changes from something with
high lower harmonics, to something "buzzy" and "thin" with a flatter
spectrum.