[sdiy] ODP: Re: Schmitz ADSR replace pots with CV control?

[Roman Sowa]

Let's assume a freak like me drives this EG in shortest times, and looped, to treat it as VCO.
So about 3ms one cycle. At 100 kHz resolution, tuning such "vco" will result in steps of about 5-6 cents according to my rough calculation. Still very ok for digital implementation of abused envelope.
But it's theory, because PIC's 10-bit resolution will create much bigger steps.
But who does that with ADSR anyway...

Roman 

> 
> A cursory glance shows 4 cycles for every instruction, 8 for
> conditionals, a lot more for maths functions. So those tiny cycles
> pile up real quick. We're left with probably ~ 10-15 instructions in
> the loop. And that's still not going to ensure we won't have sidebands
> leaking into the signal.
> 
> Yeah, navel-grazing, but there's a good reason to consider worst case
> scenarios: people will always experiment with synths and push them to
> their limit.
> 
> D.