How about a vcf rating scale?

Grant Richter grichter at execpc.com
Sat Sep 25 21:49:20 CEST 1999


The 24 DB octave resonant lowpass filter

Sounds pretty good and has had lots of implementations.

The Minimoog filter is on example which used transistors for
transconductors. It sounds pretty good depending on how you program it.
The EMS Synthi filter is an interesting variation.
which used diodes for the transconductors. It also
sounds pretty good depending on how you program it.
People say it sounds more "squelchy" than a transistor
filter, whatever that means.

There were also a lot of variations on this filter design using
integrated transconductors with discrete components.
Examples would be the four CA3080s in the Synare drums.
The Wiard implementation using LM13700s.
I am sure someone built one using CA3280s.
They all sound basically the same.

Latter on the architecture was adopted for highly integrated
circuitry where all the transconductors were put on one chip.
Examples are CEM3320 and SSM2044 as used in the Oberheim
OB series and the Korg Polysix and Monopoly respectively.
They also sound pretty good if you program them that way.

State Variable Filter

Probably the first filter made by accident on an analogue computer
doing math. Since it is so easy to implement, it has been, over,
and over, and over again. Current examples are the Sherman filter bank,
which took the trick from the EML 101 and put the mode output on
a tapped pot. Thats the best trick anybody ever came up with for
that filter.

Another good trick that was introduced briefly for the state variable
is adding an AGC circuit to the Q control. The Aries state variable
was the only on that had this and would automatically back off
the Q if the filter stated to oscillate. This let you use the maximum Q
for sweeping the whole range.

It can also make some pretty good patches if programmed that way.

Buchla Lowpass Gates

A completely different filter topology using Vactec opto-couplers
as transconductors. The relatively slow large step response
gives an acoustic "roundness" (whatever that means) to any
thing processed through them. The three operational modes
allow a wide range of expressions.

It sounds pretty good if you program it that way.






More information about the Synth-diy mailing list