DIY Variax

[jishnudg]

Okay, this isn't as ambitious as the title may suggest...just been toying with the idea of using a magnetic hex pickups (Ubertar makes one, I think)... and use the 6 independent signals and tweak them accordingly...say, run them through six different octave down circuits, or six different acoustic sims, each tuned to the frequency range of the string it handles...and a mixer with EQ at the end of it. Conceptually,shouldn't it work similar to a variax?

[Fender3D]

in the 90' Avalon - Paradis guitars had octave dividers for the three bass strings.
Interesting analog stuff...

[Mark Hammer]

Okay, this isn't as ambitious as the title may suggest...just been toying with the idea of using a magnetic hex pickups (Ubertar makes one, I think)... and use the 6 independent signals and tweak them accordingly...say, run them through six different octave down circuits, or six different acoustic sims, each tuned to the frequency range of the string it handles...and a mixer with EQ at the end of it. Conceptually,shouldn't it work similar to a variax?

Basically, an analog guitar synth using a hex pickup.  I don't say that in any dismissive way.  Rather, that means that all the accumulated knowledge about analog guitar synths becomes relevant to you.

Although the Variax uses piezo saddles as a hex signal source, where it diverges even more from what you are imagining is that it applies digital models to process those 6 string signals to output a mono audio signal (or maybe stereo, I forget).

The chief challenge in any hex-processing system is string separation.  Any mag pickup has to be jammed right up against the bridge in order to be where the string motion is most localized with the least bleedthrough to adjacent sensing areas.  That has several implications.  One is that you don't get all that much fundamental (or string motion at all, for that matter), and the other is that the coils have to be pretty small for individual ones to be snuggled up together there, and that also challenges the output level.

Octave division in the analog world, in general, is a matter of meeting triggering threshold requirements, and the picking style that requires.  It is certainly do-able, but requires careful planning of the circuits.  I would certainly not begn by plugging a hex pickup's output to a half dozen MXR Blue Box units.

More general processing of the individual strng signals, for example by separate fuzzing, and EQ-ing, is certainly feasible, and sounds nice.  getting into emulation of different sorts of instruments, as the Variax does, will be several orders of magnitude harder, though.

[jishnudg]

This is opening up a new Hornet's nest, I think You're right, I should go several steps back first, look at analog synths in greater detail...thanks a ton for the input,though!
J.

[Mark Hammer]

I had the opportunity to try out Harry Bissell's guitar synth a decade back, and it was an Ibanez with a G-Vox divided/hex pickup snuggled up against the bridge.  Each string feeds its own semi-independently channel where it is fuzzed up but good, and filters, VCAs, and envelope generators work their magic on a hexaphonic signal.

Sounds fantastic, and is eminently responsive and playable.  You can see it in its 2004 state here, and decide if its something you want to attempt.  My guess is that you'll probably decide against it.:  http://www.wiseguysynth.com/larry/guitar.htm

[Ben N]

I wonder (hypothetically speaking--NOT trying this at home  ) if the string isolation and harmonics problems couldn't be mostly solved by a combination of mixing a bit of signal from adjoining strings, phase reversed, in with each string, along with some steep filtering.

[Mark Hammer]

I don't know what companies like Roland do with the signal, once it's "in the box", but what leaves the guitar when a hex pickup is used, is simply a boosted, and potentially squared-off signal from the isolated portion of the pickup sitting under the string, by the bridge.  The harmonic content of the string is going to be much lower amplitude than the fundamental, so if the polepiece/sensing-area is right up near the string, you probably wouldn't pick up very much adjacent content.  Where you might is in the fundamental.  But since any guitar synth unit that does some P-2-V conversion is likely going to do some strategic passband filtering of each string (nothing lower than X, nothing higher than Y) to optimize for the notes found on that string, there would appear to not be any requirement for active cancellation.

At least that's how it seems to me.  I might be wrong.