Moonwalk - a compressor to pump your kicks

It was suggested to me to make this patch it’s own thread, so here it is!

A compressor with a side chain input AND a clock input. This compressor will automatically adjust the release time to be in sync with your patch.

I wanted to call it “Pumped Up Kicks” but I’m sticking with my space theme.

Enjoy

Moon Walk by SteveX.audulus (42.7 KB)

6 Likes

Nice. I have been drooling over the WMD MSCL for my rack, but it is the time of year for giving. How does something like this stack up to the Disting MK4 algorithm?

I suppose the CV out is for controlling a hardware VCA?

1 Like

I have no idea how this stacks up to the Disting MK4. Besides Reaktor and Max/MSP, I have no experience with any modular equipment.

You are along for the journey with me in my attempt to learn what a compressor is, and what makes one compressor different from another. Actually, that sounds like a good separate thread.

I’m not sure CV is the best name for what that is but I’ll explain it. A compressor basically passes the audio input to the output untouched except for a component that turns the signal down. Sometimes it’s a VCA, sometimes it’s a resistor, or other things. In my design it’s just an expression that is equivalent to a knob, basically. The CV out gives you access to this value to do weird things.

One weird thing I’ve done with the more full version of my compressor is attach the the inverse of the CV out to the ratio knob, so the harder the compressor works, the higher the ratio. This is similar to a Variety-Mu compressor. You can also feed this back to the attack and release, so the harder the compressor works, the faster the attack and release times. Lots of fun stuff! I suppose you could forgo the effect of the compressor entirely and use the CV as a smoothed meter to control different parameters. I’ve always wanted a reverb send that increases when my signal gets louder before I compress the signal. Now I have one!

3 Likes

That’s a really interesting idea!

Ha, got hit with the auto correct there. It should be “Vari-Mu”

2 Likes

I’m not sure this is the case. Since this just got ‘refaced’ and is going in the “library” maybe this could be considered a little more closely. Also, I don’t know about everyone else, but documenting the module is helpful.

At least that’s how I made this one. The fun stuff that happens in the sidechain path determines how that component turns down the input signal. Whether that is fed from the input signal or an external source is up to the designer.

This is the main “character” of a compressor.

The part I suppose that I forgot is the make-up gain. If a compressor sounds “warm” or something, besides the compression itself, this may be due to the gain amplifier. In my compressor it’s digital. Hardware compressors may use tubes or other amplification methods that impart some saturation into the signal.

I could do a write up maybe. I’m hoping to start a thread soon on compressors, and people can follow my research into why compressors are the way they are.

1 Like

I would agree with @SteveX. A compressor is essentially a variable gain control where the gain is modulated by the level of an audio input signal. Ideally a compressor shouldn’t alter any characteristic of the signal other than it’s intensity. Of course modulating the signal gain over time distorts the output with respect to the input. Real circuitry, particularly circuits with vacuum tubes and vactrols as active elements introduce a considerable amount of additional distortion to the signal as it passes through the compressor. While this may produce a desirable end result, it is not a fundamental characteristic of the compressor.

1 Like

I have to go, but you introduced time again. There is event compression happening. Analog imparts characteristics because of the chemical makeup of the circuit parts. These parts absorb energy in various ways.

  • to be fair, I built my compressor so that when I play bass, the quiet notes will be louder and the reverse, which makes for more usable recordings when mixing. Also, I like how it gives my guitar more perceivable punch. Maybe I misled myself, because when I see a compressor with a sidechain input I tend to think of glue and colour as attractive properties in that type of compressor application. But if you think of the word ‘compress’ it does have something to do with a kind of density.

A compressor pedal for your bass? That’s really cool. Do you know what type of compressor it is?

I have a feeling that we’ll never get to that level of hardware modeling in Audulus. My approach is to do compression and saturation separately. I have a patch around here somewhere called Afterburner. It allows you to distort symmetrically, asymmetrically, and EQ the distortion to sound however you want. That could help achieve the color that you’re after.

Also, maybe I’ll upload the newer version of my Gravity compressor. Moon Walk is just a fun experiment, but it’s not a traditional compressor at all.

3 Likes

Yeah, sidechaining is really using a compressor in an off-label way that musicians really like. A conventional use would be ducking the soundtrack of a documentary whenever the narrator is talking. This helps the words from being drowned out, but also maintains an overall volume level in the typical meaning of compression. A lot of edm producers who use sidechaining simply treat it as an artsy amplitude modulation technique. They will sidechain a synth line and then remove whatever is sidechaining it to create a rythme for their dark house or dubstep. I guess it’s analogous to resonant filters, which you typically want to avoid in other electronic applications, but sound interesting so musicians love em.

This is my experience with compressors. I needed a lot of help from my Dad. Together we took a schematic and designed a board, and I asked the valve wizard some questions. I did all of the soldering, graphics, layout, etc. He drilled the box, helped order parts, got out his ancient oscilloscope, etc… but we troubleshooted together. At one point my Dad said, “if I can’t figure it out, there is no way you can.” It was not his finest moment. We got through it though.

I sometimes get ahead of myself here on the forums. I don’t really think that my views are solid, but I have to find my own ways to integrate information. So far I have done zero programming in Audulus. I have a messed up view of logic, which gets in the way of things. However, I am not so sure that formal logic is ‘complete’ in its contemporary form, even though it is the basis for our technological explosion.

This is an excellent series on compression and sidechaining:

1 Like

Nice that they included J-Dilla. When I was in Detroit for a baseball game, my Dad and I stopped by Dilla’s Delights, a grassroots donut shop run by his uncle Herm and Dilla’s daughters. If you are ever in Detroit…

1 Like

Works great! Thanks again.

1 Like

I checked the old forum and didn’t see “afterburner” anywhere.

Here’s an asymmetric distortion unit based on a spline node. By changing the spline you can make the distortion symmetric or even something weird. I also built a JFET model that generates asymmetric distortion. Asymmetric distortion usually introduces a significant DC component into the signal which you may want to filter before outputting the signal. The uJFET has an audio output that includes a DC Blocker as well as a modulation output that does not, but the spline based unit does not block DC.
Asymmetric Distorsion Saturator V1.0 demo.audulus (30.7 KB)
uJFET 2 V1.1.audulus (7.0 KB)
@biminiroad has also included several distortion models in the library reface, both as stand-alone effects and incorporated into various VCA models.

2 Likes

Afterburn.audulus (159.4 KB)
I found this (in the current Forum).
I think thats the patch you’re looking for.

2 Likes

I have been meaning to ask, would you be interested in helping me model a version of this that is more closely modeled after the TG-4 JFET VCA for eurorack?

Be glad to. I used a JFET transfer function I found online. Let me see what I can find on the TG-4.I saw your scope traces. That would be a pretty good place to start. Drive looks like a pretty hard clip. What frequency was the trace?

1 Like