Snare compression specifics.

You can even count how long the release can be without listening to it at all. The formula is "one second in milliseconds / hits per second / (beats per minute / minute in seconds)" and same with numbers (4th notes at 120bpm):

1000 / 4 / (120/60) = 1000 / 4 / 2 = 125ms

or 16th notes blastbeat at 251bpm:

1000 / 16 / (251/60) = 1000 / 16 / 4.1888... = 14,94ms = 15ms

So that is the time maximum that you have before the next hit comes, but I would still advice you to decrease the amount with atleast the amount of attack you inputted earlier + an 64th note. Just in case, unless it sounds like shit.

Then just put the output gain to what is the average of the gain reduction of the solid hits.

edit: remembered the formula wrong, that gives the earlier result (on 120bpm 1000ms is 1/2, 500ms is 1/4, 250 is 1/8, 125ms is 1/16 notes)... I'll have to recheck it.

edit2: According to this, it seems to be (1000 * (60 / bpm)) / hits...

(1000 * (60 / 120)) / 4 = (1000 * (0.5)) / 4 = (500) / 4 = 125... no, it still gives me the same result, with a different formula.

edit3: ah, it WAS basically the same formula, just a different method :)

edit: or if it is

1000 / (4 / (120/60)) = 1000 / (4 / 2) = 1000 / 2 = 500ms

Then it is correct... Same formula, just the subdivisions were done first

edit: fuck it, just make it sound good :)

Another formula:
(4th notes at 120bpm) 60000/120bpm/4th notes=125ms

(or 16th notes blastbeat at 251bpm) 60000/251bpm/16th notes=14.94ms
 
with a 30ms attack I find it pretty hard to do much make up gain since your are letting the attack go through untouched.

I like the idea of using two comps, one for the attack and one for the decay. especially on original snare tracks.
 
i'm with most everyone else, in that i typically have a 25-30ms attack, quick release, ratio usually at 4:1, and that's for the snare along with the kick and toms

then they all get sent to a parallel comp with has a really fast attack, longer release, and stupid high ratio