Frequency controlled "auto-wah"

[raulgrell]

An interesting idea I was thinking about the other day, what if we could have a pedal that, depending on the frequency of the note played, sounded as if there was a wah pedal after it.

If that didn't make sense:
A pedal that when a low note is played, the sound comes out like the "ooh" bit of a wah pedal; and a high note comes out as the "aah" bit of the wah pedal...

Does anyone have any idea what I'm talking about and if there are any existing designs for it?
If not, how could that be accomplished? Envelope follower of some sort, perhaps or does that only work with amplitude?

[Processaurus]

You want a frequency to voltage converter, that could drive a voltage controlled filter.  Pitch extraction on guitars is tricky, but this wouldn't have to be exact to sound neat, and you could lp filter the control voltage so that momentary glitches and errors would be ignored/minimized.  What would be craazy would be to calibrate the filter so that it emphasized the exact frequency of the note you are on, like you can with analog synths.

Here's a cheesey way to do it.  Split off the guitar signal, and send one copy to a Tim escobedo PWM to make a fixed pulse width pulse from your guitar signal.  When you play higher on the neck, the rectangle wave pulse is on for more of the cycle than when it was lower on the neck.  Low pass filter the heck out of that, so that it makes a slowly changing DC voltage, based on the guitar frequency.  Send that to the VCF.

[raulgrell]

Ouch...

To the drawing board!

[StephenGiles]

Front end of EH Guitar synth yet again, nothing less my friend will provide a decent pitch to voltage short of a dedicated chip.

[Mark Hammer]

Doesn't need to track the actual pitch.  All it has to do is track relative pitch (i.e., higher or lower).  I suggested something like this a couple months back, where the control voltage was derived from the difference between the rectified outputs of one highpass and one lowpass filter.   The filters are not too selective such that a fairly wide range of frequency content is subsumed under each.  However, the closer to the one filter's passband the fundamental is, the greater the rectified output would be from that filter relative to the other.  My original intent was to have a setup where you could control things with your voice by simply talking/singing/humming higher or lower.

The BIG constraint is that this is essentially a monophonic pitch recognition system we're talking about.  Once you play a chord, you're up doo-doo creek without a paddle.  But within those limits it might work nicely.

[gez]

I posted a frequency to voltage converter schematic (though not complete) a few days ago.  Drawback is that it's monophonic.  Useful in analogue synth type circuits though...

[Processaurus]

Doesn't need to track the actual pitch.  All it has to do is track relative pitch (i.e., higher or lower).  I suggested something like this a couple months back, where the control voltage was derived from the difference between the rectified outputs of one highpass and one lowpass filter.   The filters are not too selective such that a fairly wide range of frequency content is subsumed under each. 

Nice idea, Mark.  It would be better if it didn't, less touchy/liable to glitch.  I wonder if it would work to make a shallow lowpass filter (to catch the lower register), and a bandpass filter rather than a highpass filter to catch the higher register, to try and lessen its sensitivity to harmonics and various high pitched string noises, that are present on both high and low notes.

Or, no filter, but less gain on the signals to track the lower register, and more gain on a highpass filtered version of the signal, to catch the high notes.  Calibrate it so when there are bass notes present, the unfiltered signal would have more amplitude than the hi passed, but when there were no bass notes, the hi pass filter would default to being louder, from its extra gain.  Figure out the middle note on the guitar to set the cutoff frequency of the hi pass filter.  Or better yet, get it on a pot and make it adjustable!

I think it would work for chords, the CV would sort of stay somewhere in the middle if you strum a big chord, by having frequencies above and below the cutoff frequency.  Or wah upward as you strum it...

[tcobretti]

This seems like a pretty interesting idea.  If I understood it a little better, I'd cobble one together myself.