Any interest in guitar synths?

[amptramp]

If you have a sequential enabling of each string and multiplex the signal so it can come out on one line, you may need a diode at each contact to prevent sneak paths through the switching matrix.  I never said it would be easy to get the "gesture capture" information to the synthesizer, just possible.  You can capture the information on a multiplexed line like a shift register so you can have a single wire or some I²C type of transfer to the synthesizer.  Then use the six hex pickup outputs to determine the amplitude of the signal.

SMT devices in the guitar neck would eliminate the need for bringing all the six fret contacts for each fret out to the synthesizer - they could be embedded in the neck.  This was something the Thomas organ synthesizer did not have available.

[Mick Bailey]

I already capture amplitude, and velocity is easily calculated from amplitude. It's straightforward to modify my existing setup to apply the guitar envelope to the synth output, as I already apply a variable envelope that uses timings from tables. Thinking about it, it wouldn't take very long to add another menu option to select the guitar envelope.

My most recent modular interface has a continuous tracking mode that does bends, finger vibrato and slides, though it is a little bit glitchy - down to me not being a programmer. Someone with coding skills could possibly do a better job of signal analysis to get it spot-on.

[Mark Hammer]

One of the tricks used in "the old days" was to employ unwound strings on all 6, tuned high, in order to encode pitch quickly, from shorter wavelengths, and then add a pitch CV offset to what the VCO got, to restore desired pitch.

[ElectricDruid]

If you have a sequential enabling of each string and multiplex the signal so it can come out on one line, you may need a diode at each contact to prevent sneak paths through the switching matrix.

Absolutely! You can't do a diode matrix without diodes! I was taking that as a given.

I never said it would be easy to get the "gesture capture" information to the synthesizer, just possible.  You can capture the information on a multiplexed line like a shift register so you can have a single wire or some I2C type of transfer to the synthesizer.

Yeah, that's a good idea too. It wouldn't be hard to use some tiny SMD shift register some way up the neck to reduce the number of lines required markedly.

Then use the six hex pickup outputs to determine the amplitude of the signal.

Hey, I think you've thought about this, haven't you?!?

[ElectricDruid]

One of the tricks used in "the old days" was to employ unwound strings on all 6, tuned high, in order to encode pitch quickly, from shorter wavelengths, and then add a pitch CV offset to what the VCO got, to restore desired pitch.

Oooh, yeah! That's cunning as a fox too!
If you're not actually deriving sound directly from the strings, there's no reason why they need to be pitched at the actual frequency and s..o...d..a..m..n...s..l..o..w..so, yeah, makes a lot of sense to pitch them up and then deal with that at the synthesis stage.

[Mark Hammer]

I thought it was pretty clever-ish, too.  The one drawback is that if one had any intention of using the same instrument to produce guitar sounds, as well as synthesized ones, you're kind of out of luck...or out of pitch.  But for an instrument dedicated to synthesized sounds with a string mentality, it IS a clever idea.