Parallel distortions with all pass filter on one side

[hans h]

Hi all,

Wanted to run an idea by you to see if I'm making any sense:
This was inspired by Billy Corgan's (smashing pumpkins guitarist/ singer) story on how they would blend two big muff guitar tracks in the studio. One big muff would have its tone pot turned to where it goes from neutral to bassy, and the other would have its tone pot turned to where it goes from neutral to trebly. The bassy and trebly tracks would be mixed together and apparently help in creating the wall of sound type of guitar tones.

So I was thinking, what if I put two big muffesque distortion paths in parallel, both with a big muff tonestack and one with an allpass filter to create some phase shifting of one path relative to the other. This last idea is from mark hammer's "woody" (https://www.aronnelson.com/DIYFiles/up/M._Hammer_s_Woody.pdf).

Does that make any sense, or will this not work the way I intend it to work?
Once I get the green light i'll breadboard it

Edit: muff a and b both get the triple back to back diodes in the feedback loop (not shown in schematic)

Schematic below:

[Fancy Lime]

The Big Muff tone stack is already a kind of HP-LP- mixer. I think what Corgan meant was what is often done to create that Wall of Sound: record one guitar track with one sound and then another guitar track, where the same player tries to play the same exact thing as before but with a different. Some people did that with dozens of tracks. The trick is that each sound is a different track and a different performance. It's not one guitar signal split, processed, and then recombined. You can sort of fake this effect with a stereo chorus or, more convincingly with digital processing for a life performance but usually this is unnecessary or counterproductive because life you are battling room acoustics at high volumes, meaning the Wall of Sound is baked into the venue anything and simulating more of it just makes it all mushy like.

HTH,
Andy

[idy]

Blending fuzz is one of those things: if you want to try it get two muffs and a mixer blender. Add an all pass to the mix or anything else. If you love it build a pedal.

[hans h]

I am aware that double tracking in the studio means playing twice which cannot be replicated exactly using a pedal.

So what I am trying to recreate is two things:

1. We can create two different tones and blend them, much like people blend different amps. Blended big muff stacks may result in an interesting notched frequency response since the notches of the two tone stacks will be at different frequencies. The result should also be less scooped due to the notches not being at the same frequency.

2. Faux 'two guitars'. Normally one might use a short delay or chorus on one side to make it somewhat similar to two guitars. However I'd like to keep the parts count low so including a delay for just this aspect is a bit overkill. Therefore I wondered whether the delay introduced by an allpass filter might make for a bigger sound. But I am quite unsure about that last bit

And indeed you may be right that it would be easier to just build another opamp muff and use a blend pedal to mix it with the delay and opamp muff I already have, but I like to tinker with new circuits

Big edit: the upper big muff should be inverting because the all-pass in the lower path is inverting. Whoops

[Fancy Lime]

1. It would be easier and more effective to tune the tone stack and maybe add a mid control, to get this effect.

2. Due to the way our auditory brain-bits process sound and interpret signals as "same" or "different", a simple phase delay via all-pass filter is not going to sound like two guitars at all. To fake double tracking you simply need some kind of modulated delay, ideally in addition to a bunch of all-pass filter with different frequencies in the delayed path, and, if at all feasible, stereo processing. This can actually be done fairly cheaply with some commercial components plus a bit of DIY but if it's worth the effort depends on the use case. Life it is likely to cause more problems than it solves in most band contexts (exception: three piece or less band open air), and in the studio, using real overdubbing is easier and more convincing than faking it.

That being said, please don't let this discourage you from experimenting. Sometimes we find interesting sounds while searching for completely different sounds.

Cheers,
Andy

[anotherjim]

Some time ago, a regular here produced an adjustable "fixed" phase shifter for tone adjustment of a parallel fx mix. Can't remember what/who it was but I've seen the basic premise come up a few times.
Guitar sound in the studio is often crafted from a blend of close and distant cab mics.
Supposedly, the coqued wah sound of the "Money for Nothing" intro is nothing more than mic blend.

[hans h]

Thanks Andy, idy and another Jim for the responses. Since the allpass isn't going to do anything for the 'two guitars theme' I'll abandon the allpass.

I think I'll do muff side a without coupling cap and muff side b with an undersized coupling cap that cuts bass. That way it will be a distortion in parallel with a muff. I'll keep the parallel tone stacks for now and just see how it goes on the breadboard. Does any of you know what the allpass is used for in Mark hammer's woody if it is not for a 'bigger sound'?

Thanks in advance, Hans

[idy]

If you are in to this idea and want to set it up with a splitter/blender, a phase shifter with a manual (disengage LFO. like bad stone) allows you to tune one (or both!) sides, not unlike mic tricks.

At some point the various split band fuzzes may begin to call your name: quadrafuzz. triumvirate... there are some cool commercial and diy designs to there.

[marcelomd]

I'll suggest a chorus, or chorus-like effect on the second loop.

[hans h]

Might be a good idea to insert an effects loop on one side then I guess the send level would be a bit high though in the circuit I drew due to the preamp gain...

[Mark Hammer]

The single allpass stage on the Woody was intended to align the harmonic content with the fundamental.  That was based on my understanding (or perhaps misunderstanding) of how the Aphex units worked.  Had nothing to do with sounding "bigger".  A single allpass stage doesn't really add anything that might be perceived as "delay"

There are a few things that may well have been "under-described" by the Corgan discussion.  The use of two Muffs in parallel with different tone-knob settings is perhaps more than the mere filtering.  Indeed, nothing seems to have been said about how the gain/Sustain controls were set on the two units.  Was there the same clipping/drive for lower and higher sections?  And once again, I'll note Mike Matthews' comment in an interview, some 15 years back, that if you took any four consecutively-produced Big Muffs off the line in the 70s, they would all sound different from each other.

Another thing is the trick used by many during recording of using two different amps, or perhaps one amp, mic'd from up close and from farther away in the room.

Forum contributor Thomeque had a circuit and layout some years back that was intended for producing through-zero flanging by adding a very short fixed delay to the "dry" signal of an existing flanger circuit.  As I recall, there was an MN3007 and MN3207 version, and the result yielded an adjustable, but fixed, delay in the single digits (1-9msec if memory serves).  I haven't used it yet, but I was curious about mixing two drives with a slight "stagger" between them, to sort of mimic what you get by sticking one mic right in front of the speaker, and another 15 feet away.

[GibsonGM]

Right. If you mess around with this stuff in a DAW, you can see the different potentials and shortcomings of various 'big guitar' techniques.  Sometimes I just use a single track, copied, and put a different amp sim/IR on each (and pan L and R).  The sheer difference in tonality creates a full 'big mono' sound without the need to delay or double track (which sometimes I find muddy and 'indistinct' on certain things).  Sometimes I'll 'advance' a track a few ms to create that stereo feel.

Other times, I will do the full double track..always with VERY different tonalities.  I still can't believe that wall of sound (Pumpkins) was many Muffs + settings, all different takes..but have no reason to think Corgan lied ha ha.  I just personally can't get that kind of thing without mud.  2, 3 tracks, ok...more than that, no.

Point being...I'm of the school of thought that splitting a guitar signal like this, you get the best sound (subjective) by making each one radically different...subtle variations won't make much difference.  Just something based on 'doing', YMMV. It's worth messing around with; it's extremely educational! 

[hans h]

I think at this point I'd better whip out the breadboard and start tinkering then. Small coupling cap pre-clipping on side 'A' might make for a good starting point since it wouldn't hurt mixing some tighter bass with the big muff sound. I'll keep the chorus/delay options in mind but will start without since I do not want the build to become untenably large.

I'm hoping to do a two pot build (tone a and tone b) with fixed volume and gain/sustain settings. But I invariably seem to end up with more pots

[Strategy]

It's maybe not exactly the same thing as what you're planning, but try Deadendfx LEO X, their clone of the ultra rare Schumann Lion X. I built one but haven't gotten to do a deep dive with it yet. It's like two Red Llama channels, EDIT: it's one clean channel and one Red Llama ish channel! including you can split them across two outputs to amp separately (can't remember right now if the series/parallel of the two channels is switchable, I think it is), and each Llama channel goes through its own almost-synth-like bandpass filter, which you can turn on and off on each channel. I may be not conveying all the technical aspects correctly but you could definitely rig it up to achieve get two parallel, filtered dirt channels with it. Not a muffy sound, though, definitely that CMOS overdrive tone that can get very distorted but not fuzzy really.
What a great project I haven't seen it discussed much on DIYSB. I haven't found the perfect track/project for it yet.
STrategy

[hans h]

Thanks for the suggestion strategy. Will keep that one in mind as well. I tried the double muff idea and as some of you suggested the two paths are a bit too similar. The end result is interesting but not interesting enough to warrant the extra parts count.

So to keep in line with the pumpkins vibe I switched to a jmp1/jcm800 ish clipper on side a and the muff on side b. The mids pot of the classic marshall tone stack will serve somewhat as a level control as well since the mids is sort of 'tone stack bypass'. Schematic:

[amptramp]

One of the things FM stations must do is limit the FM modulation frequency deviation and one method used to do that for speech is to pass speech signals through an all-pass filter like a phaser but with fixed values.  Under most conditions, the harmonic waveform peaks in speech tend to line up so the signal is a bit larger than it needs to be.  Since the ear is not all that sensitive to phase shifts in various harmonics, using an all-pass phase shifter makes little difference to the perceived sound but reduces the peak signal amplitude and therefore the peak modulation.  Using an all-pass, you can carry the average modulation level a bit higher before overmodulation occurs and signal clipping is only needed as a last resort.

If you use an all-pass filter like a phaser, you will find that signals that would be clipped or distorted at a certain clipping level without one can be too low in amplitude to be clipped or distorted at the same level once they go through the filter.

[Mark Hammer]

While, in practice, a wee bit of phase shift can be helpful for some things, the basic perceptual instinct/reflex is to unconsciously scan to determine which sources harmonic content should be "assigned" to, and assign them based on relevance to the fundamental, as well as "time coherence".  That's what we do when perceiving this and that instrument as having their timbre.  Separating harmonic content from fundamental "enough" in time, and that harmonic content doesn't get connected to the fundamental.

The point is that, unless the two 'versions' are separated enough in time, that harmonic content being produced by the two circuits will be assigned to their common fundamental/s.  Not unmusical, but unlikely to be heard as two different sounds.

Some delay-based pedals will provide dual outputs, wet and dry.  Consider running those two outputs into the two tone-differentiated Muffs.

[bartimaeus]

this is a very cool idea, so i made a model of this using max/msp.

it definitely doesn't sound anything like two guitars. but depending how you set the allpass, the phase cancellations can add some really nice warmth to the sound, kind of like when you move a mic further away from an amp.

[hans h]

Hi all, thanks for all the answers.

I might insert a wet only switch in my perf-deep blue delay and an effects loop in the parallel distortion pedal to run one side slightly delay. Or look out for a chorus. I'm not going to make it built-in: too much circuit for my perf skills

Bartimaeus, i' d be interested to see those max/msp results. Do not know that software, I use ltspice.

I updated the schematic for distortion on one side and muff on the other. I reverted the last stages to inverting to get harder clipping. Then i was reading about full wave precision rectifiers and wanted to try the octave circuit from the omnidrive by John hollis. Will this implementation work? He has a Non-inverting opamp following the octaver. I used one of his 10k resistors as the input resistor to my inverting opamp but not sure if it will work....

[hans h]

I'd also be interested in your thoughts on delaying/ chorusing the distortion or the muff side (fx loop). I'm currently leaning towards delaying the muff side so I get the sharper distortion attack up front followed by the fuzz.