Compressor as an EQ

[hubb]

Been thinking about this thread.

Another idea that wont squash the life out of a mix entirely.

If you have an eq like gliss eq that has dynamic filtering (focusing on transients that are available in a certain freq. range rather than just boosting a curve), you could have a look at your whole mix and then accentuate the most important peaks, like the fundamental, the snares body, the fundamental doubled up and tripled and so on.

[SunkLo]

I think it's time I showed you guys my knob...

[+] Spoiler

The center LED lights up depending on the parameter the knob controls. So on the attack knob it'll light up during the attack phase, and on the threshold knob it'll light up when the signal exceeds the threshold, etc.

Too bad it'll look like turd at a smaller size
It's hard to cram a bunch of glitz into a 50px footprint. Gotta practice some pixel packing I guess.

[nowaysj]

how lonG ARE Sounds in attack phase?

[titchbit]

how lonG ARE Sounds in attack phase?

[SunkLo]

Depends on your attack time innit

[nowaysj]

f
t
range

[SunkLo]

f
t
range

[+] Spoiler

[Genevieve]

how lonG ARE Sounds in attack phase?

Watch this thing with the audio of this in the background and you're in heaven

[nowaysj]

Just cause a guy has only his left hand in the darkness to type...

What is the range of times for the attack phase? .5-250ms?

[titchbit]

250 ms sounds pretty short. i'd say you'd want at least 5-10 seconds.

[SunkLo]

Somewhere in the middle of your answers. Anywhere from instant to about 5 seconds.

Attack time is a misnomer though, really you're setting an attack decay speed. The way it works is by moving a fraction of the way to a target. So say 0dB gain reduction is occurring, signal goes above threshold, compressor requires 4 dB of gain reduction. The attack envelope takes the distance between the current value and the target value and multiplies it by the attack coefficient. So say the attack coefficient is 0.1, it'll move 10% of the way to the target value each sample. Because of this, it never actually reaches the exact value, just gets closer and closer. Think of moving to halfway between you and a fence, then moving to half that distance again, etc. The way most people calculate attack time is how long it takes for the exponential decay to get "close enough" to the target value. With "close enough" being defined as some arbitrary percentage.

Compressors like the 1176 where you're adjusting attack speed instead of time are actually more accurate to what's actually going on. Even though everyone complains "the knobs are wired backwards!"

[titchbit]

oh fuck i was talking about synths i forgot we were talking about compressors lol

also that makes a lot of sense sunklo. Like a half life basically.

[Sharmaji]

Compressors like the 1176 where you're adjusting attack speed instead of time are actually more accurate to what's actually going on. Even though everyone complains "the knobs are wired backwards!"

literally kept me up at nights. "MAYBE I JUST DON'T GET IT!?!?!?!" please don't do that on a plug looking fwd to seeing what you're developing tho!

[SunkLo]

If I did I'd be sure to label it Attack Speed not Time and have "Fast" and "Slow" written at the extremes of the range. Might still do that but have fast on the left and slow on the right just so it's still intuitive to people who are used to time based controls.

You use a VST host Sharm?

[Samuel_L_Damnson]

honestly man, i use reason but i'd buy this comp off you when its done to use in reaper. I no nothing about vsts but this looks interesting. plus i bet it will be more unique than the average.

[Sharmaji]

If I did I'd be sure to label it Attack Speed not Time and have "Fast" and "Slow" written at the extremes of the range. Might still do that but have fast on the left and slow on the right just so it's still intuitive to people who are used to time based controls.

You use a VST host Sharm?

logic and ableton, so i'm like 90% AU

[SunkLo]

Ah damn. I think I could still compile you a mac vst to check it out in ableton.

Today I'm gonna try to implement filtering in the sidechain and main signal path, as well as an option to use a transient detector instead of the standard envelope follower. And a wet/dry heh. I also... might... put in separate attack/release for the VCA gain depending on whether it's compressing or expanding... Timing stages are like the fucking hydra, this thing's gonna have a million of them by the time it's done. Might have that as an advanced option so it's not super overwhelming. Trying to keep the complexity fairly restrained without sacrificing flexibility.

[Sharmaji]

^ killa. would be super happy to check it out.

sounds like something crazy involved along the lines of the things elysia makes... innnnnnnteresting.............

[nowaysj]

IMO:

Simplicity > Flexibility

========

The process of music production is complex enough. Anything that can remove complexity and still achieve results is a god send. Don't be fooled by a reductive linear argument here, of course there is a sweet spot. Your mission as a designer is to build around that sweet spot.



And Sharmaji:



To losing sleep over the 1176.