At 12:48 PM 7/28/2008, Scott Juskiw wrote:
>I'm considering creating a PCB (or two) that would have all the
>additional circuitry needed to make a "modular vocoder". You'd need
>two of the JH string filters to make this work (and a few patch cords
>to hook it all up). You could build a standalone vocoder and eliminate
>all the patching, but then you wouldn't be able to use the JH string
>filter "as is". I'd like to at least try it. I designed a vocoder back
>in the mid 1980s but never got it completed. The JH string filter
>already has the necessary mixer and the filters have 10-turn trimmers
>for calibration, that got me thinking it wouldn't be all that much
>extra work to make a vocoder.
Wow, interesting! Personally, I would not say that 40 VCAs and 40
envelope followers are "not much extra work" :-) although I
understand that at least the 40-band filters are taken care of.
(BTW, I might be interested in PCBs with, say, 4 VCAs or EFs per
board. Are you planning anything like that?)
>Here are some other uses for the string filter separate outputs:
>
>1. extracting vocal formants
>2. spectral processing of individual bands
>3. using specific bands to trigger envelope generators
Indeed. However, these functions can generally be performed by a few
bandpass filters. I'd rather use the 2U to house 3 or 4 manual BP
filters, like a pair of Blacet dual filters
(
http://www.blacet.com/DF.html), instead of a 40-jack array.
But, that's me. As an observer, I'd enjoy seeing someone else build a
40-band analog vocoder!
Oh: I prefer the panel design on which the jacks are staggered like the knobs.
John