envelope-followed overdrive: does it exist?

[barret77]

Hello all

I have this question for pure curiosity - theoretical  sake:
Is there an overdrive pedal where the drive is guided by an envelope follower? If not, why? Isnt it a good idea to capture the dynamics of the pickup?

[Mark Hammer]

That would be the E-H Graphic Fuzz.  There is a thread about it and a schematic (not sure of accuracy, though) within the last few days.

There was also a "smart fuzz" project that appeared in ETI some 20 odd years ago, using a 570/571 compander chip, where the input signal dynamics were restored after the fuzz was produced.

The "Fuzztain" project from Bill Berardi that appeared in Electronic Musician in the mid-to-late 80's used a pair of reversed LEDs from two CLM6000's to do clipping, and used the LDRs from those optoisolators to produce compression earlier in the circuit.  There is no reason why the LDR half of an optoisolator could not be applied to producing dynamic expansion so as to maintain the dynamics of the input, instead of using it to reduce/compress the dynamics.

How?

You will note that the gain of a non-inverting op-amp stage, depends on how much feedback makes it from the output to the non-inverting pin.  Take a look at the MXR Distortion+ and/or DOD 250 for examples.  In each case you will see that the gain depends on making the resistance from the "-" pin to ground lower in value.  Since the feedback resistor in that op-amp remains constant, that shrinking resistance to ground means that the two resistances (from output to "-", from "-" to gnd) behave like a volume control that turns down how much output feedback comes back to "-".

With me so far?

If a glowing LED makes the resistance of an LDR get lower, then that LDR can be placed between "-" and ground (or Vref, as the case may be) so as to increase the gain (by reducing that resistance) whenever the person picks harder.  I've already suggested this in the conversion of a Flatline compressor to a "Punchline" expander ( http://ampage.org/hammer/files/Punchline.PDF ).  That suggestion was merely an idea, not a true experiment, but Rick Lawrence, bless his leather-hatted soul, DID test it out and confirmed that it works reasonably well.  What remains, of course is to:

• Stick an LED-based clipper stage before any expansion stage (The Fuzztain has the dynamics-altering stage before the LED-based clipper).
• Work out (experiment) with how much gain change the LDR produces in response to picking.  It may be necessary to put both parallel and series resistances in combination with the LDR to achieve the right amount of expansion so that it feels natural.

You will note that it takes two reverse-biased LEDs to clip, but it only takes one LDR to produce the expansion effect.  That second available LDR could be allocated to a number of places.  Alternatively, it might be easier (and cheaper!) to use a single optoisolator and a separate LED for the other half of the back-to-back LED pair.  If I didn't have so many other things to finish, I'd offer to do it myself, but....

[SolderBoy]

There's also a circuit in one of the R.A Penfold books (Published by Babani) called "Twangy Fuzz", if memory serves. 

Its a opamp fuzz with a LM13700 VCA.

I breadboarded it several years ago to run before an envelope filter for nasty synth tones - worked quite well.

I've always thought that if I ever find a bass fuzz that I really like, I would have a switchable VCA in the same box - so I could run a mutron or something after it...

[gez]

If not, why? Isnt it a good idea to capture the dynamics of the pickup?

Penfold has done a number of envelope controlled fuzzes in various books.  Personally, I don't like these circuits.

One of the nice things about overdrives is the compression which allows you to sustain notes.  A lot of this tends to get lost if the overdrive tracks the natural decay of a note/chord.  As for hard fuzz, it sounds like turning the volume down on a hi-fi, or the fade-out at the end of a CD track...a tad artificial.

[barret77]

nice, thanks for all the feedback.
I'll have to read carefully M. Hammer's post, I'm not that familiar with all the concepts on these circuits.

And by the way, I noticed that the guitar rig 2 from Native Instruments allows you to attach a Envelope follower or LFO to anything - any parameter on an amp, effect, speaker... worth experimenting on the virtual world.

[Tim Escobedo]

I've tried a few envelope controlled distortion circuits. My goal is is a circuit that really follows dynamics, allowing me to go from clean when playing soft to full distortion when playing hard. The trickiest part I've found is the transition window is surprisingly small. And subjective. One of these days, I may clean up one of the circuits and post it.

[Paul Perry (Frostwave)]

Maybe an envelope follower that controlled (via a pair of voltage controlled amplifiers) the blend between an unprocessed and a fully overdriven signal would be useful?

[smccusker]

i might be wrong, but dont overdrives follow the dynamics of your playing anyway? the softer you play the cleaner your signal, and the harder you play the more overdriven it becomes.

[mac]

I am working on a overdrive-wah-¿env-foll? pedal. Very similar to what Mark Hammer wrote above.
Basically I use a TL071 with a 100k pot in the feedback loop and a pair of back to back leds to do the clipping. I added a Twin-T network in the feedback loop with a ldr going to ground. And a 1K resistor, a 224 capacitor and a ldr in series going to ground from the opamp pin 2.
Both ldr's take the light from the leds to vary the tone response of the notch network and to change the gain of the opamp.
It is a very preliminar prototype so it can have many bugs. Use at your own risk! No money back garanteed. ;-)

mac

[Tim Escobedo]

i might be wrong, but dont overdrives follow the dynamics of your playing anyway? the softer you play the cleaner your signal, and the harder you play the more overdriven it becomes.

They're often marketed that way. And perhaps they are somewhat responsive to dynamics. But I've never heard any circuit that went from true clean to distortion purely by playing succesively harder.

[Eb7+9]

an interesting design problem - I think what you want is an amplifier that increases gain with signal envelope to drive high threshold clippers and then an attenuating amplifier following that to keep the output at a reasonable level prior to diodes fully turning on ...

I'd start with a side-chain circuit that detects the input signal directly ... preferably a circuit that doesn't overshoot too much, and one that can drive two opto diodes (in series or whatever) ... they need to be somewhat matched - at least same type and close enough resistance at one test current value ...

if you use a clipping circuit that has grounded LED's, like in the Rockman X100 with a 10k in series with the diodes (in the real circuit not the one in the patent) you can adapt the gain altering resistors easily ... the idea of using LED's is that until the diodes fully turn on they act as pseudo-linearish high valued resistors - so within +/-1.8v or whatever you'd get a cleanish signal at that diode node where the signal moves on to the next gain stage and output ... using LED's should yield enough "clean" room ...

if you stick one of the photo-resistors in series with the input resistance of a first inverting op-amp stage driving the LED's  you'd get an increasing-gain amplifier there, and if you use the other photo-resistor in series with the feedback resistance of a second (post LED) inverting op-amp stage you'd get an decreasing-gain amplifier keeping the output more-or-less steady ... resistance that's in series with the cells can be used to limit the extent of the drive/reduction - and the gain of the side-chain circuit would determine the dynamic transition range between clean and dirt ... the second series resistance (setting minimum gain in second amplifier) would have to be tunable to prevent output squashing when LED's are on hard - a key requirement ...

[GFR]

i might be wrong, but dont overdrives follow the dynamics of your playing anyway? the softer you play the cleaner your signal, and the harder you play the more overdriven it becomes.

They're often marketed that way. And perhaps they are somewhat responsive to dynamics. But I've never heard any circuit that went from true clean to distortion purely by playing succesively harder.

Going from clean to distortion by playing harder is not that difficult. The difficult part is, we want this to happen while the apparent volume level remains constant (well, at least that's what I'd like). You can try something like first you distort, then you compress the distorted sound. This way the compressor will try to even out volume variations.

[birt]

if you put an ldr in a distortion pedal it would start to turn the volume down the longer you wait because the inputvolume that triggers the ldr goes down too right?
what about feedback from after the ldr to the triggersignal? wouldn't that stretch the sustain?

i think it woul also be nice to turn everything around and instead of pushing the sound when there is a lot of input doing the opposite. in that way you lose the attack of every note. The recognition of an instrument by the human ear is very dependant on that attack tough and if you take away the attack of instrument sounds they sound synthy instead of natural. i'd like that in a smooth violin like distortion.

[JimRayden]

i might be wrong, but dont overdrives follow the dynamics of your playing anyway? the softer you play the cleaner your signal, and the harder you play the more overdriven it becomes.

They're often marketed that way. And perhaps they are somewhat responsive to dynamics. But I've never heard any circuit that went from true clean to distortion purely by playing succesively harder.

I think that's what the Trainwreck amps are known for. I remember reading about the approach they use. Instead of the several amplification stages and the output transformer starting to distort one by one, they actually are designed to saturate at the same time. This results in a more dramatic clean/OD transformation, and thus the dynamicness. Though that's only what I've read, I can't prove it, as I find the amp's schematic rather ordinary, except for those humongous cathode resistors.

Logically taken, a fuzz or overdrive box with no sustain would sound pretty bad and misbiased. (as tube amps do have natural compression even when played clean, this effect is not so dramatic)

Any objections and nags welcome.

---------
Jimbo

[Mark Hammer]

i might be wrong, but dont overdrives follow the dynamics of your playing anyway? the softer you play the cleaner your signal, and the harder you play the more overdriven it becomes.

They're often marketed that way. And perhaps they are somewhat responsive to dynamics. But I've never heard any circuit that went from true clean to distortion purely by playing succesively harder.

Well that's the thing.   Rather than being miles away from clipping with all the clean headroom a person could want, all distortion pedals are designed to be "primed" to clip, with the extra amplitude from stronger picking bringing the brunt of the signal in closer proximity to the clipping threshold (as opposed to only having some of the peaks lopped off in the midst of all the rest of the signal).  Like Tim, I've never seen anything that remained a very safe distance from clipping during light picking, and moved completely into fuzz territory when pick attack increased.

The Enveloope from Toadworks takes an interesting approach to that particular problem, bringing an effect loop in proportional to envelope amplitude.  I'm not sure if the actual panning is directly propotional to envelope ampllitude, though, or whether it simply turns on/off in response to envelope (as opposed to needing a switch-press to actuate).  If it DID fade the loop in, then it would be possible to start with a lighter picking not passing through any clipping circuitry whatsoever, and heavier picking passing through the clipping circuitry, with even stronger picking not only shifting the mix balance between completely clean and distortion, but the distortion being more distorted as well.

The "dynamics" of distortion circuits fall into two broad themes.  One is the extent to which different shades of distortion can be produced with pick attack strength, independent of output volume changes, and the other is the extent output volume changes track the input volume changes, independent of clipping amount.  Because of the properties of the typical design, most distortin pedals will provide more clipping in response to input signal amplitude, but this is not reflected in corresponding changes in outout volume.  The classic case is the Tube Screamer that produces significant compression of dynamic range (even though it is not a "compressor").

My sense of what is generally being addressed in this thread is the design of a distortion device that produces BOTH sorts of dynamic shifts.  That is, the output is always louder as you pick harder, and the output is always fuzzier when you pick harder.  In general, because of how that clipping is normally produced, one needs to do something "extra" to restore the input dynamics, post-clip.  The approach taken by the Graphic Fuzz is to use a VCA/OTA after the clipping whose gain is given by the signal amplitude BEFORE the clipping.