[sdiy] Fun things to do with a half SSM2164?

[Tom Wiltshire]

I like Aaron's HPF idea. Alternatively, you could add a second VCA  
and then you'd have a pannable, stereo output. Or could you have some  
kind of amplitude or ring-mod stage before the filter? Another  
possibility would be some type of of pole-mixing to allow morphable  
filter responses. Paul Maddox has done some nice things like this by  
(I think? Paul?) mixing in some of the first filter stage's output  
(For example: http://www.vacoloco.net/synths/defender/specs.shtml).

I don't know that I agree with Aaron about not linearising the  
SSM2164. If the VCA is exponential and the envelope is linear, you  
get a rather strange attack curve. This is different from the effect  
of a classic exponential ADSR into a linear VCA, which is the other  
way of getting a 'natural' exponential response.
Of course, you can compensate for the SSM2164's exponential response  
with a software lookup table or by tweaking the envelope curves  
directly - perhaps this is what Aaron had in mind. In theory this  
needs decent DAC resolution (12-bit at least). My own experiments  
suggest that in practice you can often wing it with less.

T.


> On Jun 12, 2010, at 1:41 PM, Olivier Gillet wrote:
>
> > The board I'm working on at the moment uses a SSM2044 as a  
> filter, and
> > the final VCA is a "linearized" SSM2164. So I have two log gain  
> cells
> > available for an additional CV-controlled processing to be  
> determined.
> > I don't want to add a lot of parts, at most a dual op-amp. What  
> would
> > you suggest?
>
> Why are you using the linearized SSM2164? If you're driving it with  
> an envelope from your microcontroller it would be really natural to  
> drive it with a linear ramp and then get exponential VCA behavior.
>
> In any case, to answer your actual question, a simple single-pole  
> *highpass* filter, like on the Jupiter-8, might be fun; that would  
> take up one gain cell.
>
> - Aaron