How fuzzy can something be/get?

[Paul Marossy]

aren't some fuzz circuits designed around an intentionally mis-biased transistor?

Yes, I think a few of them are. But, those are "splatty" and do start gating. I'm talking about something a little different but I really don't know how to put it into words....

[Herr Masel]

Even a misbiased distortion will sound "gatey" and a bit fuzzy. What I've never thought of is exactly HOW we percieve this modulation of the waveform and what makes it sound as it is. Of course I know about square waves and hard clipping etc. but I wonder how the receptors in our ears respond to these unnatural (I assume?) sounds, or wether they are carried any differently through the air...

[Paul Marossy]

I wonder how the receptors in our ears respond to these unnatural (I assume?) sounds, or wether they are carried any differently through the air...

I would submit that it's a little bit of both. Enter psychoacoustics. 

[Paul Perry (Frostwave)]

I don't think there is going to be an answer to this one, because a lot of the character of fuzzes (I suspect) is tied up in hte way that the 'fuzziness' changes with dynamics.
The human ear & brain is particularly sensitive to rapid changes in harmonic mixes (hence the popularity f the wah) & the changes in shape as drive level changes triggers this perception. More so with chords.

[Mark Hammer]

Yes and no.  Everything you say is true, but it IS possible to hear a single note, held and "finger-wiggled", that strikes one as seriously sizzling.  Heck, a number of cult hits over the years have virtually depended on the strikingness of such single notes at a crucial junction in the song.  I'm curious about what the properties of such a note might be.

At the same time, the properties of that single note are one thing, and maybe not to be confused with the properties of a chord, or how something sounds with a series of rapid-fire notes, and so on.

[R.G.]

The amount of fuzz correlates pretty well with the amount of higher harmonics present. A square wave is not the fuzziest.

You can make a square wave fuzzier by adding a sine to it which cancels the fundamental. What is left is entirely harmonics. The fundamental is reconstructed in your ear from the harmonics, so you can't get less than that.

Waveforms which sound fuzzier tend to have strong high pass characteristics, which amounts to much the same thing as cancelling the fundamental, but also includes cutting the size of the low-order harmonics as well.

The more a waveform approaches narrow pulses, the fizzier it sounds. A one shot which fires at each positive zero crossing sounds really, really fizzy.

One way to make such waveforms which track guitar volume is to extract the volume information by doing an envelope detector, then extracting frequency information in a sidechain by amplify/limit stages to give you the zero crossing info. With that, you can gate the envelope with logic signals into a switch that switches the output either to the envelope or ground, giving the same fizziness but the normal volume trail off.

You can use the envelope to do different things at different envelope heights, too, so a fuzz might change from more to less fizzy, or switch in different filters at different levels. It could change the modulation you apply to the envelope to do, for instance a change between one pulse per zero crossing (a fizzy octave) to one pulse every two crossings (fizzy fundamental) to square waves (really hard normal fuzz), things like that.

[DDD]

Pls see my opinion and impressions below:
1. I've tried to play through very thoroughly designed and high-sensitive comparator to achieve strictly 50% duty cycle of the output signal. Also the input signal had been seriously filtered then compressed and filtered once more to cut off overtones. Result: the output square waveform with the 50% duty cycle sounds "sweet", something like "Starless and Bible Black" solo party by R.Fripp ;-)
2. Heavy asymmetrical clipping sounds roughly and produces much more intermodulation than symmetrical distortion
3. Gated distortion combined with the asymmetrical clipping produces very impressive roaring and bassy sound with power chords and especially with quarts on every two "neighbour" strings beginning with the 1st and 2nd ones.
4. For me, the fuzziest sound oscilloscope picture is square form fundamental (maybe slightly asymmetrical) with the high and narrow peaks that are 2 to 5 times higher than the square wave itself. Good for single notes and 2-string power chords, but produces too much chaos from chords. As far as I remember, Colorsound Tonebender is a good example if playing on 4-5-6 strings, but sounds "too sandy" on the 1-2-3 strings.

[Paul Marossy]

The amount of fuzz correlates pretty well with the amount of higher harmonics present. A square wave is not the fuzziest.

It probably looks that way to me on a scope, but I can't see all of the harmonics going on because I don't have a spectrum analyzer. OK, so, of those higher harmonics, would they be more even or odd order? Or would it matter?

[Quackzed]

ever notice how light colors look brighter against a dark background? maybee a signal would seem fuzzier if it started out cleen then swamped into a fizzy mess! by virtue of the relative contrast of the clean tone against the fizz of tons of harmonic content...
might not be the theoretical 'fuzziest' but might seem more fuzzy by virtue of the comparison to its clean state...
another thought ... is there any way to give white noise - frequency??! like by turning it on and off at regular intervals?? i guess not... would the white noise at high on off frequency just sound like .. well, white noise?   
 

[MikeH]

This question immediately made me think of the into to "Territorial Pissings" by Nirvana.

[scaesic]

i think "square wave" is the wrong route to go regarding the original question. A square wave is pretty easy to make, and doesnt, to my ears, sound interchangable with a fuzz effect. If you wanted to play a squarewave you could just stick in a midi pickup and play a set of perfect square waves off your newest fanciest software sampler.

to me fuzz is fundamentally different, it is suggested that clipping the hell out of a guitar signal and "squaring it off" to remove all original guitar tones is the "ideal fuzz", but i feel this is wrong. i think to start with the ideal fuzz is somewhere between a perfect square/pulse wave but still has at least some guitar harmonic content not present in a squarewave, but still exsists (although out of context, so it doesnt sound like a clean guitar tone). I think the ideal fuzz starts to come close to modular synthesis in the ways you need to modify the signal. As well as "squaring off" i think its important to modify the attack and decay times, i feel strongly that this is where the "sick to the stomach" feeling comes from.

so my two cents is, squaring off, but not squaring off to form a square wave, and attack and decay curves. after all a guitar signal isnt just steady state, the transients introduce the important harmonics which determine the quality of a sound to humans.

[Steben]

I have to refer to my switching tranny topic. In Bipolars, fuzziness can also mean the sticking of electrons, giving the wave form a different shape at the clipping edges. This can be found too in clipping opamps. Switching transistors are faster, giving rather clipping without the edge stickiness.
    --
  /    \
/        \   switching

    ---
  /     |
/        \   sticky / opamp

[AdamB]

I think it would be much easier to define fuzz in terms of the design of the device causing the distortion (transistor distortion vs. op-amp distortion etc.) rather than as a waveshape, because the waveshape a fuzz device will make is always different with different signals. Fuzz is a process applied to a signal, not a signal itself.

You couldn't ever say 'this is what a sine wave looks like, this is what a square wave looks like, oh, and this is what a fuzz wave looks like...' because the fuzz shape is not something elemental like the two previous waves.

With that in mind, the ammount of fuzzness you could apply to a signal would be determined by the naunces of the components making up the device. Not just how they react individually (transistor frequency response etc.) but how they interact with each other. There are far too many variables involved in measuring a concept as mickey mouse as 'fuzziness'. But that's the beauty of analog electronics.

[Processaurus]

the 8 bit Nintendo with its limited sound synthesis would use very asymetrical rectangle waves for edgier more obnoxious sounds (the closest they could get to distorted guitar), and squarewaves for the smoother sounds.  Lately too I put that rectangle wave idea to use with a cheesy synth guitar, playing with someone employing a "sick fuzz", the two sounds blend together really well. 

The Big Cheese is interesting to me because (in addition to sounding sick) it varies pulse width depending on how hot the signal is going in.  If I remember right it starts out 50/50, and gets narrower as it gets louder, and at a point starts to get more even again.

is there any way to give white noise - frequency??! like by turning it on and off at regular intervals?? i guess not... would the white noise at high on off frequency just sound like .. well, white noise?   
 

cool idea, you could filter it with an adaptive filter, tracking the guitar.

[brett]

Hi
Are we searching for far too simplistic explanations for what are quite complex mental models of sound? (ie our concepts of great fuzz).  These choices are very subjective.

In the end, the question seems analagous to asking "What is a really great form of the colour green?".  People who blend colours can talk about that stuff for hours, but the question is clearly unanswerable in absolute terms.  

Are the answers to Mark's questions influenced by the type of connection between the people and the fuzz (stompbox maker, fuzz user, music listener)?  Their musical preferences (ie. Jazz vs Acid Rock)?  Their previous experiences (dare I say that using a FuzzRite or Fuzzface *will* change your perceptions of fuzz)?  What other factors are there, and combinations of factors?

Maybe we should first come to grips with selecting a "high quality/ultimate" fuzz tone, then discuss the technical aspects of achieving that tone.  Contenders might include "Spirit in the Sky", "Satisfaction",- my favbourite- "Thickfreakness", something by Jimi, and a few others??

FWIW concerning:
1 scope traces:  The slope (dV/dt) and overshoot of the leading edge seem important in those "edgy" fuzzes like "Spirit in the Sky".  The FuzzRite and ETI fuzz have plenty of overshoot (ie extra length) on the waveform.  A little overshoot usually seems good, while a lot is often bad.  These two fuzzes, and probably other fuzzes, achieve the overshoot by cancelling out the fundamental and some harmonics.
2 equalisation: the FZ-1A/B, Fuzzrite and others have high rolloff frequencies (I think I once calculated 160 Hz for the FuzzRite).  No wonder they sound brassy and lacking in bass. e.g. Satisfaction
3 trailing edge/PWM.  IIRC the MFZ-1, "starved" BJTs, and a few other circuits manage to spread out the signal so that it gives a pulse width of 70 to 80% of the period.  Joe Davisson''s Blackfire was quite hot on this (IF I'm thinking of the right circuit).  And yeah, the Blackfire is right up there with mighty mean effects.

[Mark Hammer]

In social science research, surveys are a commonly used tool.  One often hears complaints from those receiving them that the questions are too ambiguous.  Of course, the ambiguity is what makes the questions so useful.  If they were fully articulated, with tons of contextual information, the vast majority of people would answer the same way.  E.g., "If an election were called in 3 months, and the economy hasn't picked up, the degree of fractiousness in the legislature hasn't improved, my quality and timeliness of services received from government hasn't improved in any noticeable way, I would vote for ________".  Such a question would elicit more uniform response than a mere "Next election I plan to vote for_____".  Typically, it is the mental schemas the responder holds that fill in the missing information in responding to ambiguous questions, and it is those very mental schemas that the researcher is truly interested in.

So, in that spirit, Brett is absolutely right that the question leaves a lot of assumptions unclarified.  I deliberately asked it that way because I realized that people hold different mental models of what "fuzzy" should be for them in their context or experience.  That's the data that I think will prove ultimately the most useful to the group and to any designer: What do people think a fuzzy sound ought to be?

[R.G.]

Good point Mark and Brett.

I like distortions that have some kind of subtle modulation in them. It doesn't seem to matter a whole lot what the modulation is - voicing that changes during the note, interfering oscillations, cross modulation "growling", pitch shifts, or just a degree of distortion that changes. Too much of this gets in the way, of course. But a little adds an interest that grabs your attention.

I also like distortions that are either not too fizzy or are really synthy. In between just sounds bad to my ears.

[Steben]

Too much of this gets in the way, of course. But a little adds an interest that grabs your attention.

Listen to Clapton's solo in "Sunshine of your Love" (and actualy the lot of guitar tone in the song).
That's really "beautiful crap" to me.  It's very broken speaker sound, yet it retains the nuances of the "neck position" guitar. Very questy sound.

Another one: eternal lasting dirty fuzz lead on "Cowgirl in the Sand" by Neil Young.