Compressor before Distortion vs Piecewise function generator before Distortion

[Vivek]

Dear Teemu K,

I had meant that if we set a soft clipper to have same envelope as a compressor for all steady signals

And then clip the signals from each very hard


Will any dynamic movement of the compressor show after the hard clipper when the static envelopes are the same ?



Dear Mark, I found many brutal pedals that make square waves out of any guitar signal over let's say 5mvp.

Re Rockman X100

After a compressor with an overall gain of let's say 5 for large signals, The Rockman has a stage that has gain around 120 and that rail saturates, which is followed by LED hard clipping to ground , so yes, in the DIST mode, the Rockman does make square waves of all input signals above a very small level. And indeed, Tom almost makes the compressor become a transfer function generator in DIST mode, by changing the configuration by clipping.


Hence I request you gurus to help me understand, can I replace a compressor with a transfer function generator if it will be followed by extreme distortion ?

[teemuk]

I had meant that if we set a soft clipper to have same envelope as a compressor for all steady signals

How will you do that?

Instantenous compression will clip the peak portions of the waveform, gradual compression will attenuate at gradually increasing rate the whole waveform evenly.

Say, we have an input signal with 1V peak transient fading down to 10mV. Instantenous compression will limit amplitudes exceeding, say 100mV, by clipping and leave lower amplitudes unaltered.

Gradual compression will start decreasing gain at certain threshold so that, say 1V signal is attenuated down to 100mV, by tenfold ratio. Simultaneously the 10mV smaller signal will also be attenuated by tenfold ratio until the compression "recovers".

Now run these two to your clipper that cuts everything above 20mV and note the difference.

[ElectricDruid]

I had meant that if we set a soft clipper to have same envelope as a compressor for all steady signals

A soft clipper *is not* a compressor. One deals with instantaneous input level and output level, and the other deals with *average* input level and output level. You can't make a compressor instantaneous so that it's like a soft clipper because if you did it wouldn't be a compressor any more, and you can't make a soft clipper act like a compressor because it doesn't know what the average level is.

I think you've been staring at graphs of input level versus output level for too long, and you're seeing the similarities between the graphs but losing sight of the (much more) important differences.

[Mark Hammer]

Correct.  Clipping of any kind, whether hard, soft, or other, reduces the dynamic range by imposing a ceiling on amplitude.  But that's not the same as altering dynamics, under control of a sidechain.

[Vivek]

Thank you respected Mark, Teemu, Tom,

Each of you explained that a compressor dynamically changes an envelope while a clipper statically limits the envelope. Obviously these are different mechanisms and lead to different end results. I agree

Yet I have a gut feel that when both of these are followed by an extreme clipper, not too much of their differences show through.

The analogy that comes to mind is:

We take 3 fresh army recruits of different heights

We give each of them a different approved army haircut

then we remove their army boots and look at them and try to decipher what happened to that recruit's hair and what was his height by looking at his boots.


I will try to hand draw some diagrams so that I understand what I am trying to learn about / achieve 

and I will also breadboard both the options to listen to them.

The goal is to develop, if technically possible, a distortion device that has some of the benefits of a compressor in front, but at lower cost and complexity than a 13 USD FET which needs adjustments. I wish that the signals that reach the hard clipper section are massaged to be a bit more uniform in amplitude than a direct guitar signal reaching that hard clipper.


Unless there are other alternate methods to achieve the same thing ? or no other method to attain similar results than use of a real compressor.

Thank you dear sirs for your inputs !

[ElectricDruid]

Yet I have a gut feel that when both of these are followed by an extreme clipper, not too much of their differences show through.

I agree. I've built pedals with such extreme gain and clipping that it hardly mattered what *note* you played, let alone what type of guitar you used!! If the clipping is harsh enough, you can remove just about everything that identifies the original signal, whatever it was. After all, one square wave sounds much like another...

In terms of other ways to achieve the result without needing a 13USD FET, how about looking at alternative technology to build a compressor? You could look at optical compressors or OTA-based compressors - vactrols and OTAs aren't as expensive as 13USD, although they're not cheap either. FETs were often used in the 1970's because at the time, they were *cheaper* than OTAs or vactrols, but that's not necessarily the case any longer.

[puretube]

Here`s how they did it (compress and/or clip) 81 (!) years ago:
https://worldwide.espacenet.com/publicationDetails/originalDocument?CC=US&NR=2342238A&KC=A&FT=D&ND=8&date=19440222&DB=EPODOC&locale=en_EP

And here`s my view of the diff`rence between compressing & clip/distorting:

[iainpunk]

Hence I request you gurus to help me understand, can I replace a compressor with a transfer function generator if it will be followed by extreme distortion ?

no

cheers

[iainpunk]

i might add that it has to do with the harmonic content that youre getting out of both.

the soft clipper will generate harmonic content, intermodulation and it favours keeping the fundamental in tact and adding its related overtones, where the compressor does not create such harmonics, intermodulation and it favours natural (chaotic/imperfect/naturally modulating) overtones. also, having the piece wise or smooth soft clipper is in front of the clipping circuit without filtering in-between, it becomes exactly equivalent as multiplying the two gains at the center before distortion. just multiply the linear gains or add the dB's together, and build a single gain stage with that combined gain. heck, even a set of 2 hard clippers without filtering in-between with the same base gains of the soft clipper and the other hard clipper.

if the extreme distortion is an opamp comparator fuzz, this will make very little to no difference, but if there is some frequency favouritism in the gain structure, or filter in in between the two stages, the resulting harmonic spectrum can or will be quite different from one another.

to get the differences between the two under control we would put the filtering before the compressor or soft clipper, but i don't think that is what the original rockman does anyway. what is the actual gain range of the distortion section of the x100 anyway?

cheers