[sdiy] Additive VCO

[Glen]

At 11:58 AM 3/8/04 , john mahoney wrote:

>As a point of reference for additive synthesis, the keyboard museum has a
>description of Kawai's architecture.
>http://www.keyboardmuseum.com/ar/k/kawai/add.html
>There are a whopping 64 sines per "Wave" with a 5-stage amplitude envelope
>for each partial. Custom digital chips made this possible.

For the best results, it's very useful to also have a *frequency envelope*
for each partial, as well as the amplitude envelope. More than 5 stages in
the envelopes would be nice as well.

If the idea of 64 partials seems intimidating, limiting yourself to 32
partials can also give some nice sounds as well. Many sounds don't even use
the whole 64-partial "spectrum" anyway.


>Terminology note: If you are mixing complex waves (i.e. not sine waves) that
>are not precisely tuned (and phase locked?) relative to each other, I don't
>think it's accurate to call it additive synthesis. It's just subtractive
>synthesis using multiple oscillators. But I could easily be wrong...

It only becomes subtractive synthesis when you start filtering the bank of
oscillators. There's nothing that says the oscillator amplitudes and phases
have to be locked, to "qualify" as additive synthesis. As for using
waveforms other than sines, that doesn't matter either. It's possible to
build an additive from square waves, if you want.

The combination of the additive oscillator bank and filtering would
actually be a hybrid. Now if we just replace those several oscillators with
several FM synth modules, we could have a really interesting hybrid system
-- especially if we add granular synthesis to the output, via several BBD
devices.    

That last sentence was partially for Harry. (Pun intended.)    ;)


later,
Glen Berry