Hey guys!
I'm trying to find graphs/curves from the most known studio hardware compressors so I can study and maybe replicate them in MCompressor.
I googled this already, but it's mostly general info, not specific models.
If anyone could point out some resources I would greatly appreciate it!

 

Last edited: Nov 10, 2018

You won't find graphics of the curves online but you will find downloadable manuals in .pdf format.

So, for example, the Manley Vari-Mu has a full set of curves in its manuals.

 

Thanks for the idea! It never crossed my mind that these things had manuals
I'll update the op with whatever i find ;)

 

Try to measure/analyse original LA-2A, 1176, Fairchild with DDMF PluginDoctor with hardware input and post it here :D

 

Hardware, schemes and general curves:


Opto compressor curve:




Vari mu compressor with 6V6 tubes:



VCA compressor gain reduction limit:




Self labeled:




Not from the original hardware, but an emulation:




And just because you seem to interested in the grafic response curves:



And finally a general overview of some different compressor functions

 

Last edited: Nov 11, 2018

Believe me, if i could lay my hands on any of these i wouldn't bother finding their response curves
I would post them, though, after request!

Thanks for the reply and the effort. Are any of those from the real hardware units? Because the real high-end, analog, hardware compressors is what i am really interested in...

 

Last edited: Nov 11, 2018

I edited the post. If you look at the curves (the little waves) and the labeling (600Ohm load), they all from hardware, although I'm not sure if the UA LA2A is a scheme or a real measurement (the curves look too perfect to me). But, apart from the latter, there were no hint that they were made with genuine hardware. The 1176 is labeled as a hardware clone.

 

It's genuine data from UA.

https://www.uaudio.com/blog/digital-emulation-of-distinctive-compressors/

You'd expect an optical compressor to be flat like that as the side chain is dependent on the light element soaking up any short term volatility in the signal. It should be remembered that the characteristics of LA2A is as much about variation in the attack and release characteristics over signal strength as it is about the simple curve profile.

 

Exactly. Which is why I doubt MCompressor is tweakable enought to adjust the right parameters in the overall behavior of both attack and release curve and the compression process itself. A type of compression (opto, fet, vca etc) isn't just relying on the curves af at and rl, its behavior relies on the types of conponents used, technics, which are already what can explain the way they behave in contact with the signal you feed them.
A compressor that both allows you to tweak the - many - right parameters beside attack and release curves, and give you templates of the most famous compressors with their parameters already ready, a fine starting point and a good way to learn both compression and the differences between the different type of compressors, is Compassion from DMG Audio (use the "mods", in the preset panel, to get access to the modeled comp : "classic fet", "classic opto", "british type 1", "type 2", "Very Mooh" (lol), etc, you'll recognize which are reproduced). Use the advance settings panel to understand what they change in the settings. Probably read the manual, it's a lot of things to take in and diggest.

 

Last edited: Nov 11, 2018

I was wondering about something that seems relevant to this thread. SPICE coding apparently allows to simulate a hardware device by inputing its components, does that mean that you can recreate a compressor digitally and make it into a plugin by just inputing the schematics into SPICE? I know this might sound like a dumb question but I have little knowledge of coding or electricity.

 

Well, i don't intent to replicate those units I would be a fool to think i could. But, combining the curve with attack/release times, the hi-pass filter of an LA-2A for example, i could come close to their
behavior, i think. Also, trying to replicate these is a good way of learning how and when to use each type of compressor.
I tried Compassion, but as i recall it was very cpu intense to use in every channel.

 

Attack and release curves are just that...Attack and release . You'll still have a digitaly responding compressor (in the way it reacts to the signal beside attack and release which are only a little part of it) no where close to what you're trying to replicate. The circuits, the material of the original unit is what explains the behavior of every aspects of a compressor. This is why there are emulations, and even with those - unless we are talking sampled IRs - there are always major differences. It's not that easy to replicate hw.
Compassion has a marginal use on the cpu, even if you use oversampling, then it goes up a little.

 

I totally agree with you! But, maybe, i didn't make my self understood. Let me give you an example:
I/We know that LA-2A is great on vocals, with its slow attack and soft knee. I could find its attack times, as a number it is clear and it is absolute. Now, the term "soft" for the knee is, well, vague. This is what i'm trying to determine. Not to replicate the hw, just to be able to test and use "soft knee" on my vocals, just apply the general principles.
Now, i get that the materials and circuitry are responsible for the character of the hw, but i'm not after the character.

 

As you can see in the pic above the knee changes with the frequency and the amount of input more or less drastically. So, there won't be any number for you to determine.

 

I understand what you mean, I probably wasn't clear enought, and that's my bad.
The response of the Opto cell to the program is well....program dependant. It's not a fixed number wether we're talking attack, release, or the way it behaves in its compression. Which is why it's so great on vocals. It's not something that you will get only by messing with the filter in the sidechain or by matching an attack curve (only). And I'm not talking about its "caracter" here (I'm guess you're talking colours, saturation). I'm talking about how it does its job because of how it was designed and with which material, components and circuits it was made.

 

Exactly.

 

Exactly exactly
Test and experiment and learn. Now i got an idea of how to approach this kind of compression in the digital domain.

 

Yes indeed. You got an idea of what cannot be done in the digital domain without the right tools.

 

The Teletronix LA-2A was a levelling amp. That's why it uses optics. It's purpose was to glue a sound together rather than etch a shape out of it. The key features are invisibility, transparency ans smoothness. That's why it's so famous on vocals. Most amateur vocalists are all over the place, but the really good pros can sing to within a fader touch of dead level. These great vocalists don't need vocal riding, they need something like an LA 2A to bring it all together. Tom Jones, it's like he's got one built into his throat, all done with mic technique and experience.

 

Alright. You obviously don't understand what i'm trying to do here.
Because the quest is what can be done.
Yeah, i know what you're gonna say: "Nothing"
But i can get the transient of a snare stand out with a digital compressor. And that's something. And it requires certain amount of attack time an a certain type of a knee. I'm questioning for the latter. So...
Anyone with curves of real, highly regarded compressors? I want to study them!