[sdiy] x0xb0x, yay

[Jeff Farr]

Oops, some bad memory in those earlier posts.  It's not Q8 I'm talking
about, but instead Q9 and Q10.  I'm also not trying to provide a
completely tecnical description of what is going on, I'm just saying
what I can hear and why I think it makes for a better sounding clone.
You can check the results of subatomicglue's tests here:

http://www.subatomicglue.com/x0xl0g/733k-analysis/index.html

I know that the VCF doesn't track frequency, but if the resonant peak
is occuring at a higher frequency than the interval between any given
fundamental from the VCO and the peak is going to be wider, given the
same cutoff setting (yea... this is a bad way of explaining it..).  To
my ears the effect of the low beta is different than simply increasing
the cutoff frequency either by boosting a control signal or turning
the knob up.  It has occured to me that it would be unusual for the
resonant peak to occur above the cutoff frequency, but this is
essentially what I hear.

My impression from all this is that the amplitude of the rezonance
isn't as important to the 303 sound as the pitch of the resonance.

>the "filter ringing out" is probably filter feedback oscillation

Close, before a filter goes into self oscillation it will oscilate
(ring out) whenever there is sound at the input, but quickly decay
when there is no input.  You can make nice kick drum noises by tuning
a filter to just below self oscillation and feeding a control pulse to
the audio input.... the 808 essentially works this way.