attack and release times explained - please

[especht]

Hi, guys. I 've got a dumb question for you.

I read in a lot of postings (such as a recent Kevin Killen article where people describe the attack and release rates for compressors as fast, slow, or medium.

Could I get your opinions of what you consider fast, slow, or medium as expressed in milliseconds?

Bonus points will be awarded if you include tips on when you like to use a particular attack or release rate!

[Kiwiburger]

Don't get hung up on the numbers. There is no official standard - so they are meaningless, except as a knob pointer.

If I said "1 ms" - what does that mean? 1 ms to do what? To reduce the gain from what to where?

It's even more confusing with software compressors - i've found no logic to the numbers.

In a fairly rough way, we can expect fast attacks (few ms or below) will chew into the waveform itself and cause distortion. Attacks of around 50 ms will let large transients through uncompressed before clamping down on the tail. Attack of several seconds may appear to do nothing, except with long release times they will basically float the average volume of a track like an automatic gain control.

Short release times allow the compressor to recover fast, in time for the next transient. If you like to hear a lot of room sound getting compressed, you need fast release times.

Something like that - but the numbers just do your head in. You can adjust different compressers so they sound about the same, and the numbers can be completely different.

[especht]

I appreciate the info.

[Mastering101]

just look at it like your trying to tweak in your favorite radio station until it sounds great!!!!!!lol

[andychamp]

Some compressors (like the UA1176) are so fast that they're able to track low frequencies, like sustained bass notes, as if they were a signal's envelope. This relatively fast operation of the comp's gain control causes AM of the audio signal, which can lead to distortion.
That's too fast.
Some compressors (often used in mastering) are so slow that, when an element drops out of the mix or submix, it can take several bars for the mix to return to full volume.
That's too slow, unless that's exactly the effect the engineer wanted to achieve.
I personally like a compressor to attack slowly enough to let through a drum's attack phase before it controls its "boom". Then I want the release to be fast enough to let the room ambience through uncompressed. It's a cool effect if you can get it to do this in sync with the song's tempo.
For vocals, you may want the exact opposite: attack fast enough to catch uncontrolled dynamics (P's, T's, O's and AH!'s are bad), but release so smoothly as to be unnoticeable.
Get a compressor where you can manually set the attack and release times, feed it any signal you can think of, play around and you'l develop your own definition of "fast" and "slow".
AND: the gain reduction lights do not have to be on all the time.

enjoy!

[nosebleedaudio]

As a starting point release time increase with lower frequency, you dont want a short release time with something like a bass guitar or kick.
Compressors like the LA-3 have a fixed release time thats long, longer than .5 seconds.

[e-cue]

Originally Posted by especht

I appreciate the info.

Attack and release times aren't the whole ball of wax. There's another big consideration (among others)- knee settings. The knee is basically the point where the compressor starts it's gain reduction. A hard knee is instantaneous. A soft knee (or over easy) ramps up, and the compression ratio gradually increases from 1:1 to whatever your ratio is set at (if you have a ratio control), so that the transition from uncompressed to compressed is more gradual. To over simplify- think of hard knee attack like when you turn a light switch from off to on and soft knee like a light dimmer you fade up from black. The difference between Hard Knee and Soft Knee is usually more obvious at high compression ratios.

When I'm doing 2 mix compression at the end of a mix or if I'm mastering and I have a knee control, I usually lean toward the soft knee side, which is less obtrusive. I'll set my ratio higher, like 6:1 or 10:1, whatever, turn the threshold down so I'm doing like 6 db of damage, then I tweek the knee setting till I hear the benefits of using compression without loosing a lot of the punch I need to retain. Then I back everything off (ratio 2:1, threshold so it's shaving a db or 2 tops) and I consider this my ballpark area. Sometimes it sounds very different, and you just have to start over, but most of the time I have much less tweeking to do, or at least have a better understanding of how the program material is going to react. My attack times are usually slower, and my release times very often (but not always) fall around a 16th note.

Then again, many compressors don't have a knee, slope, or overeasy control. Also, plenty of people sometimes engage their 2 mix compression during the beginning of a mix.

No rules, just observations.

[e-cue]

Originally Posted by effectsnut

just look at it like your trying to tweak in your favorite radio station until it sounds great!!!!!!

Finalizer Wizard?

[u b k]

to me, 1ms is a fast attack because it squashes the transient to the point where there's no meaningful 'pop'. 10ms is a slow attack because there's a big, uncompressed 'pop' at the front of the sound.

under 50ms is a fast release because it's incredibly audible (it pumps) and the source generally sounds distorted (not a bad thing) to my ears. 300ms is slow because the compression gets way less noticeable, the envelope is natural.


gregoire
del
ubk
.

[Sigma]

front end "tit" i just set attack bet 5-10 ms and release to program source

ovd compression-- 150 ms attack and over a sec release

rounding --1ms and below to takh off the edge

google compressor limiters like the la-2 , la -4 , dbx 160 etc to see what their times were ..and emulate if using a DAW plug in

i like opto's for alot of stuff ..that slow down from -3 to 0 is nice