deskthority

I have long had an idea for a device that would allow a great deal of extended-range playing to be possible on a normal six-string electric guitar without adding any strings or otherwise hard-modding the instrument. The only change necessary would be to alter one of the pickups to allow for the following:

Divide the pickup as follows: The three high strings remain tied together as one group and are completely unaffected by the device. (They may be routed separately if the player desires, but that does not affect what I'm trying to do.) Each of the three lower strings (I will call them E, A, and D even if the performer chooses to use another tuning system) has its coil separated and routed individually to a circuit which does the following:

(1) If there is enough happening on the E pickup, route that signal to output 1, and the A and D pickups to output 2 where they will be recombined with the signal from the GBE strings.
(2) If the E pickup does not have enough signal to trigger the circuit (it can't be a pure binary yes/no because there is always bleed from adjacent strings), then the next step is to check for a signal from the A pickup. If found, that is routed to output 1, while the D pickup is routed to output 2 to be recombined with the GBE set. The E pickup is null routed.
(3) If neither of the above criteria are met (the E and A are both silent), then the D pickup is sent to output 1, and the E and A pickups are null routed. There is no need to check whether there is signal on the D pickup.

So what's the point of all this real-time pickup splitting? Output 1 would go through a different processing chain (and quite likely a different amp) than output 2, that's what. This would likely include an octave effect, making it simple to play bass lines on a standard six-string guitar while simultaneously leaving the rest of the instrument available for conventional playing. However, since the pickups are being actively switched, the complication of having multiple notes routed to the octave pedal (which is expecting a single note, and even if it can handle more than one, it's going to sound like mud) is completely avoided.

The whole thing needs to be able to run off a single 9V battery and fit in the space available in a guitar body. (A 7805 is probably going to be necessary.) Why 9V? That's just what guitar pedals use, along with active pickup systems. (Some are 18V but they just use two 9V batteries instead of one.)

So... is a Teensy reasonably well suited for a task like this? It wouldn't have to do any actual processing on the analog signals, just route them.

I know this is about as off-topic as it gets, but some of you know Teensys pretty well, and hopefully I was able to describe the requirements adequately that one need not be a guitarist to understand them. There will be a bit more magic to be done splitting the pickups, but that's not relevant to the circuit.

For further clarification, the frequency of the low E string (assuming it's not drop-tuned) is about 82 Hz, which means a period of about 12 ms. In order not to feel laggy to the player, switching needs to take place within one or two cycles, especially since the processing chain comes with a latency of its own. Again, no analysis is taking place, so the frequency does not have to be identified. No Fourier transforms are required. It's simply a case of detecting whether a signal is present or not.

I'm not an expert at these things, but I don't think that it would be easy to make it sound good without a lot of fine-tuning.
I'm afraid that there could be weird audible clipping effects. A switch from one amp to another would not take effect until after a string has been stroked because of there always being a bit of a delay in the circuitry picking up that you are playing a note, and from residual ringing below the switching threshold. Even if the transition would be analogue and gradual, I think you would get short pings from strings in the wrong octaves.
I think it might be better to use some kind of easily accessible toggle switches on the guitar instead. You could let those switches be routed to a digital amp that suppresses residual ringing in favour of new notes, but then you would get into audio processing again.

Findecanor wrote: I'm not an expert at these things, but I don't think that it would be easy to make it sound good without a lot of fine-tuning.
I'm afraid that there could be weird audible clipping effects. A switch from one amp to another would not take effect until after a string has been stroked because of there always being a bit of a delay in the circuitry picking up that you are playing a note, and from residual ringing below the switching threshold. Even if the transition would be analogue and gradual, I think you would get short pings from strings in the wrong octaves.
I think it might be better to use some kind of easily accessible toggle switches on the guitar instead. You could let those switches be routed to a digital amp that suppresses residual ringing in favour of new notes, but then you would get into audio processing again.

Fine tuning would almost certainly be required. That's exactly why I'm asking about Teensy, because it can be reprogrammed to accommodate those fine tunings. Switches are out because the primary design goal is to add no new responsibilites to either hand. Anyone who can play guitar should just be able to pick it up, noodle on it for five minutes, and grasp how it works without having to read a manual. Actually setting it up to perform at this level would be a job for a tech, or a guitarist who is willing to put in the time and effort required to do it himself.

Ultimately, since there will always be latency involved, the guitarist will have to learn to play the bass line on the "front edge" of the beat if he wants it to sound in time with everything else. I can't get around this without significantly modifying the instrument, and once the instrument changes feel and form, it would be simpler to just pick up an eight-string guitar.

If I were to start modifying the instrument, I'd probably swap out the low E and A for strings an octave lower, and then use circuitry to raise them back up to their normal octave for mixing into the overall sound. This way the latency would be on the upper octave rather than the lower. The problem is that I would essentially be turning a 6-string into an 8-string with the 5th and 6th string missing, and taking up the slack with effects.