Engineer Thumb: is it normal behavior?

[lrgaraujo]

Hello everyone! I've built an engineers thumb according to ValveWizard's schematic (schematic here) and it compresses as it should.
However, depending on the settings (mostly higher ratio and lower threshold combined) I get this weird artifact where it feels like the volume (and the noise) is increasing after the initial decay.
I thought there was something wrong with my build, until my band mate's CS3 did just the same in our last rehearsal, so my question is: is it normal? Is this what people call "breathing"?

 

[Betty Wont]

That's how compressors work. I can't even hear the actual effect, all I hear is the noise floor going up and down.

 

[lrgaraujo]

That's how compressors work. I can't even hear the actual effect, all I hear is the noise floor going up and down.

That's good to know! I can definitely hear compression (the engineer thumb has more compression available than I find useful, actually), but this behavior was throwing me off. It should get a little better noise wise when I box it up

 

[Chuck D. Bones]

Depends on what you call "noise."
If it's hiss, then boxing it up won't help.
If it's hum, then boxing it up might help.
If it's RF pickup, then boxing it up will definitely help.

 

[lrgaraujo]

Depends on what you call "noise."

I'm guessing, it's mostly hum and RF, but to be honest, after I heard my bandmate's CS3, I'm not that worried about the noise floor (my ET is probably less noisier even outside the box).

I might tinker a bit with the release time (built with the suggested 470k), though, to see if I can tame the 'breathing' a little bit

 

[Chuck D. Bones]

What's to guess? Hum is a constant buzz at the power line freq and it's harmonics. RF sounds like voices or music over the radio.

Breathing can be caused by a slow release.

The Engineer's Thumb works differently from most compressors. Most compressors measure the volume after the variable gain stage and adjust the gain to obtain constant volume. This is feedback control.

The ET measures the volume before the variable gain stage and adjusts the gain to compensate for the input level. This is feed-forward control. For all of its advantages, the disadvantage of feed-forward control is you have to measure the signal level very accurately. The peak detector in the ET (and most other compressors) measures the instantaneous peak signal strength. What the ear hears is more of an averaged (RMS) signal strength. When we play a chord, the phase relationship between the notes is constantly changing. This causes the peak amplitude to vary widely. The average amplitude decays more-or-less uniformly. My conclusion is the ET would work better if it measured the RMS signal amplitude. Doing that is possible, but takes a lot more parts.

Bottom line: you can try speeding up the Release, but you will probably still get some breathing.

 

[lrgaraujo]

I'll see what I'm able to do without destroying the board, might even try a switch or a potentiometer there. And I can live with some breathing, specially if I manage to reduce it a bit

 

[lrgaraujo]

What's to guess? Hum is a constant buzz at the power line freq and it's harmonics. RF sounds like voices or music over the radio.

Just played with it again. RF makes for most of the noise and breathing did not even bother me all that much (now that I know it happens). I'm boxing this one up

 

[Laundryroom David]

Just played with it again. RF makes for most of the noise and breathing did not even bother me all that much (now that I know it happens). I'm boxing this one up

I built one a few weeks ago. FWIW, boxed up I get no RF and it’s no noisier than any other comp I’ve ever used. Helluva squish to it. Fun stuff.