Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[bob ebeling]

I've been engineering for 20+ years and 90% of the time I use a compressor I go to the fastest release and the slowest attack.  

In recent years I've explored and managed to speed some attack times up, as well as slow some release times down.  However, the fast attack thing is still somewhat of a mystery to me.  It always seems weird to hear mastering guys talk about needing faster times.

Part of this could be just defining fast I guess.  For instance, my Alan Smart C1, when at it's slowest attack is at 30 milliseconds, which is pretty fast.  But then there is the Tube Tech CL1B, where I love to have that attack knob fully clockwise usually (especially for vocals).  I like to set that knob near the middle of the range (12 oclock) for drum overheads.  Grabs quite violently then.

So my question to all is; When do you use fast attack times?  On what sounds?  AND with what compressors?  How much gain reduction?  What release times in conjunction?





 

[rankus]

I like faster attack times on vocals, but that's about it.

[bob ebeling]

What comp(s) are you using on vocals and what attack times are fast to you?  Are you just chopping off the real loud stuff?  

[j.hall]

1176, attack knob on "3".  i'm compressing the crap out of it.......

[bob ebeling]

What would you guess the milliseconds of that attack is J.?

Even with the old Rev D I still lean on the slowest attack/fastest release, I just don't know what the numbers are.

[rankus]

j.hall wrote on Mon, 12 January 2009 11:11

1176, attack knob on "3".  i'm compressing the crap out of it.......

Me too... 1176 but with Attack on 5.5 to 6 and release about the same... (JJ calls this the Bones Howe setting)  really good starting point IMO.

Similar starter settings on Bass gtr too btw.

[Fig]

bob ebeling wrote on Fri, 09 January 2009 21:39

I've been engineering for 20+ years and 90% of the time I use a compressor I go to the fastest release and the slowest attack.

Hi Bob - this is called "being in a rut".  It can also be considered a "safe" place to be set so as not to cause any harm.

But realize, time constants on dynamics processors are dependant on the envelope of the original signal and, IME, the tempo or pulse of the material.

Slow attack times on fast transients (snare drum for example) do not really "catch" the leading edge of the signal.  However, that may be what you are seeking too.
 

Quote:

In recent years I've explored and managed to speed some attack times up, as well as slow some release times down.  However, the fast attack thing is still somewhat of a mystery to me.  It always seems weird to hear mastering guys talk about needing faster times.

Some sounds require super fast attack times to grab transients - again, if and only if that is what is desired.  Claves, triangles, glockenspiel, etc.

Quote:

Part of this could be just defining fast I guess.  For instance, my Alan Smart C1, when at it's slowest attack is at 30 milliseconds, which is pretty fast.  But then there is the Tube Tech CL1B, where I love to have that attack knob fully clockwise usually (especially for vocals).  I like to set that knob near the middle of the range (12 oclock) for drum overheads.  Grabs quite violently then.

The twisting of knobs to get the right sound is the goal.  Try closing your eyes and ignoring the markings on the knobs.  The ears have it, understand.

Quote:

So my question to all is; When do you use fast attack times?

When the requirement is immediate grabbing of the leading edge of the sound(s) in question.  Or better, when it sounds good.

Quote:

On what sounds?

Which-so-ever signal(s) require such treatment(s).

Quote:

AND with what compressors?

Which-so-ever can do it (attack fast, that is - opticals need not apply).

Quote:

How much gain reduction?

As much (or little) as necessary.

Quote:

What release times in conjunction?

Which-so-ever makes it sound "right" or "good" or "cool" or whatever.

You mention mastering, and the use of dynamics processing in mastering can be quite different than on individual elements within a mix.  You should glean from my comments that IT DEPENDS (like most things audio).  Depends on what the sound is, what the desired post-process sound is, etc.

If you want answers - turn the knobs and take note of what the result is.

$0.02,

Fig

[bob ebeling]

Thanks Fig.
 
I don't think you really get me though.  Far from being in a rut or a 'safe' engineer, I'm more looking for interesting applications, of which you threw out glockenspeil and triangle.  
Those are intersting, I guess percussion in general like tambourine's would be a very interesting place to try faster times.  That is a great suggestion, even as I think about it I am imagining 'pushing' the tambourine back into the mix.  
Very cool.

'Use your ears'...uh, yeah.  'Just turn knobs and do what's cool'...uh, right.  

I also disagree that optical comps 'need not apply' to do fast jobs.  The mentioned Tube-tech cl1b does some amazing things to a mono overhead drum mic when dialed faster.  

Your answers about 'whichever is fast, whatever can do it, etc..' uh, yeah.  I thought this would be a good subject for people to share some specific settings, not discuss basic philosophy.

But definitely thanks for the percussion idea.

[Fig]

Quote:

I don't think you really get me though. Far from being in a rut or a 'safe' engineer,

Not trying to accuse, Bob.  Anytime folks say “90% of the time” – I just figure its their own fault.  No offense.

Quote:

I'm more looking for interesting applications, of which you threw out glockenspeil and triangle. Those are intersting, I guess percussion in general like tambourine's would be a very interesting place to try faster times. That is a great suggestion, even as I think about it I am imagining 'pushing' the tambourine back into the mix.

Right on.  Consider the time constant of the dynamics processor as an envelope shaper.  You want suggestions – try “Transient Designer” from SPL, yikes – but not just for fast, either.

Regarding tambourine – kinda depends on the style.  I might stretch out the attack time if its one of those “shaking”-type parts as opposed, say, to the regular hip or hand tap that has a very individual start and stop rather than full of RMS.  Otherwise, stretch the release time so even the below threshold stuff remains gain reduced to avoid pumping (unless that’s what you seek, again).  Once in the situation, that’s where the turning of knobs becomes the requirement, you see.

Quote:

'Use your ears'...uh, yeah. 'Just turn knobs and do what's cool'...uh, right.

Perhaps intended for other readers?  Adjustment and experimentation is almost always a good recommendation – gets us outta the “90% of the time" stuff, dig.  Don’t take it personal friend, it’s a community.

Quote:

I also disagree that optical comps 'need not apply' to do fast jobs. The mentioned Tube-tech cl1b does some amazing things to a mono overhead drum mic when dialed faster.

Indeed.  I guess I shouda said, any device without Attack and Release available (most opticals in my mind) need not apply?  I have great respect for TubeTech, though I do not own one.  Perhaps you can tell us what comps you have available, otherwise we’re making wish lists, right?

And don’t get me wrong, I love me some opticals – funny how a lamp’s “ramping” can be so natural to our other sense… hearing.  Warms my heart, actually.

Quote:

Your answers about 'whichever is fast, whatever can do it, etc..' uh, yeah. I thought this would be a good subject for people to share some specific settings, not discuss basic philosophy.

How’s this for a philosophy:  run the tape backwards and set the Release really fast.  I can’t help it if you do not like my “answers”??  We’re just talkin’ here.

A “specific” setting is exactly that, and I feel fundamentally derivative to whatever else happens to be going on at the same time.  Silver bullets aren’t out there, IME, you gotta notice when a current setting ain’t working and adjust accordingly – sometimes that means patching in something else (or nothing at all – sometimes a fader move is better than ANY dynamics processor, mute or unmute are pretty fast, too).  

Again, don’t get me wrong, its not like medium settings on time constants ain’t the way to go – I mean, some of those beasts are on “auto” (program dependant, even – whatever).

I’m also not in a big hurry to have a comp grab a leading edge – in all the live sound I do, I love to have the signal’s attack remain, well, intact.  But if its gonna stay loud for long – like maybe a belting singer – roll some o’ that GR in, eh?

Hope you don’t mind if I share.  Its why I come here at all.

My belief that "it depends" is exactly that -- a belief.

Quote:

But definitely thanks for the percussion idea.

Not to belabor it, but instantaneous “anything” takes a lotta horsepower from “somewhere”.  A triangle is the best example I can think of (short of, say, a click track pulse) with an envelope slope approaching the order of 1 instantaneously.  Things like that can really benefit from a “rounding”, if you will – assuming that’s the goal.  Otherwise, print it low… right?

I keep going back to your original post, and it mentions mastering.  Adding dynamic processing to full program material is significantly different, though, which I am certain you know, Bob – I mean unless it’s a tambourine soloist or something and then we’re back to the same discussion.

Keep it light.

[bob ebeling]

Amazing post Fig.  Thanks so much.  You really have me thinking about attack in relation to the envelope now.  It's a nice new connection for me.  
I learned everything from the ears forward, no training, trial by fire so that all goes without saying.  You have a nice knowledgable grasp of these concepts and great ability to spell it out.  I especially dig the visualizing I get when you're talking about the belting singer.  You've put some well needed pictures in a visual part of my brain now.  Always doing things by ear and what is cool is probably what has left these mysteries for me.  
I've used the Transient Designers and had a 2 channel in the rack for a while.  When you need to tweak up a soft kick or snare they're wonderful.  
The Tube Tech cl1b is a sweet ride for vocals if the singer is not exploring dynamic range every other line.  It definitely has formed the major part of alot of drum sounds with its impactful grab (attack at noonish).  
Other comps are all pretty standard, 1176, distressor, DBX 160vu, La2a, some real old Alison Gain Brains, Alan Smart C1.  I need something in the Neve family, and that is definitely an attack-time sound.  
Using the ears with eyes closed or not really looking is pretty much how I got to here.  When I grab the attack and go from slow to fast, I usually hear the sound getting farther away, sinking into the speakers.  Being that I'm usually mixing rock music, the battle is usually getting everything to come forward, hence the attack time gets lengthened.
For the last two years I moved from NYC and started working on some different music styles pretty much full time.  Out of the old comfort zone but into new places thankfully.  That has really made me start to explore the numbers behind the operation.  Needing to understand more vs. shooting from the hip.
Also, wanting to conceptualize mixes more from the start and work on spatialization within them.  Looking at attack times, there are great possibilities to explore with creating more of a front to back layering.  
I had part of this talk with a mixer buddy from NYC last month and he said 'your mixes are great, what are you worried about', and I couldn't accurately explain to him that I'm just at aplace where I need to understand more regardless.  

Thanks again Fig, very very helpful.  

[Fig]

bob ebeling wrote on Tue, 13 January 2009 19:53

Amazing post Fig.  Thanks so much.

My pleasure.  It’s a topic close to my heart is all.  It only seems complex because it is – and then, suddenly – crystallization.  Putting that into words can be more taxing than merely dialing it up, understand.

Quote:

I'm just at aplace where I need to understand more regardless.

 

Aren’t we all?

You may like this interesting tidbit, then:

Attack and release times on most manufacturers’ devices are measured in time since a signal has risen above threshold (for attack on a comp) and doesn’t “let go” until after the signal has dipped below threshold (for release).   These measurements of time based on a level ranging from merely milliseconds to beyond a second or more for some devices.

dbx is the only manufacturer I’m aware of (and correct me if I am wrong, y’all) that uses dB/mS as their time constant increments.  More of a vector quantity?  Perhaps I am misusing that term – I’m sure Andy Peters will correct me.  Anyway, to my ears, a clever way to squeeze the ratio setting in and out of effect.  I never woulda been able to describe it had I not opened my eyes to document the setting – ha!

Quote:

Thanks again Fig, very very helpful.

Hope it helps.  I’ve read every one of your forty-something posts so far, Bob.  I look forward to more conversation.

Here’s yer soapbox back, Mr. Hall.

NAMM-bound,

Fig

[grantis]

Bob,
I've used extremely fast attack times (<1ms sometimes) and fast releasees (<30ms) and about 15-20dB of GR on certain percussion instruments for a very deep ambient effect.  

I duplicate the track, compress the original like I would normally (slow attack, fast release), compress the duplicate as described above, then blend the duplicate in with the original.  On a mono source, turning the duplicate into a stereo track (with a slap delay or something) can be really cool too....or try panning the original hard left and the duplicate hard right.

I sometimes find fast attacks useful for vocals as well.  10ms approx with about 1-3 dB of GR.  Then put on your normal compressor after that (slow attack fast release) and smash, smash, smash, it.

Grant

[bob ebeling]

Killer idea Grant, I've stumbled upon that before, kind of a panning effect getting something to jerk back and forth..drums, that's what it was.  But revisiting and perfecting that is full of possibilities.  I'm really into this depth thing right now and that could definitely be a trick up the sleeve to get vocals to simultaneously sit up front and/or fade back.  
That reminds me of a related subject.  I've been really into records lately that have the vocal in this nice relaxed position inside the mix.  I mean it's always in the middle, but there is a big difference in right up in your face or back there in the performance.  That first quick attack comp could help this to happen.  What comps do you typically use on your vocals?  
For 'placement' compression I've really been digging the McDSP Compressor Bank.  
Fig, on the DBX tip, yeah, there is something unique with the vector concept.  I mean, I've hauled around a 160vu forever and it's unique.  When tracking loud vocals or punchy hip-hop vocals where they are constantly changing their emphasis, just tapping the meter at 2:1 on a 160vu is like a brick-wall limiter.  Once again, I don't even know what the fixed attack and release times are, but guessing I would think the attack is 50-80?  The slower side of fast.  Using it for vocals leaves me with one issue usually, that it's harder to bring them forward more than where they sit when tracked. What dbx stuff do you use?  I've heavily considered grabbing a 162 to run hell-on-earth 'fruity loops' tracks through.  

[grantis]

Quote:

What comps do you typically use on your vocals?

First compressor varies.  Usually I start with the Digidesign Dyn3 plugin.  If that doesn't do it, I'll try the Massey comp.  90% of the time, one of those will work well.

Second compressor, is almost always an 1176.  

[Fig]

bob ebeling wrote on Wed, 14 January 2009 09:13

Fig, on the DBX tip, yeah, there is something unique with the vector concept.  I mean, I've hauled around a 160vu forever and it's unique.  When tracking loud vocals or punchy hip-hop vocals where they are constantly changing their emphasis, just tapping the meter at 2:1 on a 160vu is like a brick-wall limiter.  Once again, I don't even know what the fixed attack and release times are, but guessing I would think the attack is 50-80?  The slower side of fast.  Using it for vocals leaves me with one issue usually, that it's harder to bring them forward more than where they sit when tracked.

That's "auto-mode" though, you know.  And hardknee.  We also have a 165 that is the first intro to the "soft-knee" that they coined "Overeasy".  A "ratio" issue, but of course all the parameters of a dynamics box interact.

Quote:

What dbx stuff do you use?

I was referring to my live stuff that breaks out time constants - 1066 (not the Neve, yo) is my favorite in this regard.  Others.

In the studio we have the reissue 160S (which then became the current 160SL) which has time constants (still in dB/mS, too).

Like I said we have a 165 for overeasy and the 160A for, you know, bass DI, neither of these have time constants.

Are you referring to the new reissue of the 162?  Or the old one?

Quote:

I've heavily considered grabbing a 162 to run hell-on-earth 'fruity loops' tracks through.

Fig

[bob ebeling]

The old one.  I've owned one before, I just can't remember that much about it, but I know it's closely related to the 160 vu.

How do you like the 160s?  What do you like it for?

[Fig]

bob ebeling wrote on Wed, 21 January 2009 07:44

The old one.  I've owned one before, I just can't remember that much about it, but I know it's closely related to the 160 vu.

Yeah.  We had an old 162 a while back.  Its effectively a stereo 160VU - unable to use it as two 160VUs (kinda like how the 1178 is a stereo 1176 - not (2) 1176es in on box).  Never tried the new one (162SL), but it sure "looks" good  

They're ok for that (strapped stereo comp/lim) - no time constants, though...

Quote:

How do you like the 160s?  What do you like it for?

Its good, too.  There's nothing in particular I would slate it for.  It DOES have time constants, and sounds nothing like our 160VUs.  When we run out of 1176, LA4, etc - we use the 160S.

It works just fine - its just not a "go to" until the "go tos" are gone.

Fig

[bob ebeling]

Yeah, that rings a bell.  I've heard that complaint before.  Seems alot of people are hoping to at least have a 160vu mode on all those new ones and can't find it.  It's kinda cool how usable the one preset is on those.  I would like to know what the attack and release time actually is on the 160vu and 162vu.
That guy that 'upgrades' 'em might know.

[mattrussell]

fast attack times are critical for many things.  a common thing i do on acoustic guitar is first running a waves renn. compressor to adjust the amount of pick (often using a very fast attack time with a fast release time) followed by my favorite compressor for Agtr, an LA-3A.  the combo has worked well for me for many years.

[bob ebeling]

In a mix session today several instruments (a bell track, a synth sweep, and some claps) we're done twice.  I took one of them and opened a compressorbank basic comp, dragged the attack time to it's fastest and then used the release time to create a 'fade/pan' away from the middle.  Great way to get some movement happening in a mix.
But the champion fast attack time  of the month award goes to the 808 kick.  There was a pair of slightly different sounding 'tuned' melody line 808's (club banger) both doing the same thing.  Cranking a super fast attack on one caused a nice HIT at the top of the notes, while the other uncompressed one eventually took over the sustain.  Both the client and I we're loving what was going on there and it became quite a feature of the song.  
Fast attack times can create very punch, tight, controlled tones.
Part of the key to quickening the attack time is usually that the release time also has to be pretty quick, or you're just squashing it down and drowning it.

[Jonah A. Kort]

I prefer to use the fastest attack time

then dial in the release time to the tempo of the song or the vibe of the source

hmmmm....

-jonah

[jonathan jetter]

grantis wrote on Wed, 14 January 2009 18:07

Quote: What comps do you typically use on your vocals? First compressor varies.  Usually I start with the Digidesign Dyn3 plugin.  If that doesn't do it, I'll try the Massey comp.  90% of the time, one of those will work well. Second compressor, is almost always an 1176.

i think it bears mentioning that even the slowest attack time on an 1176 is still pretty damn fast.

[Nicky D]

About the 162.  I use one all the time..that's a pretty special comp...nothing else sounds like it to me...great on drum buss, toms, OHDS, room....bass...always wondered what the attack and release settings were...they do seem to change depending on the input and ratio...but I might just be outta my mind!

[Etch-A-Sketch]

bob ebeling wrote on Fri, 09 January 2009 19:39

I've been engineering for 20+ years and 90% of the time I use a compressor I go to the fastest release and the slowest attack.   In recent years I've explored and managed to speed some attack times up, as well as slow some release times down.  However, the fast attack thing is still somewhat of a mystery to me.  It always seems weird to hear mastering guys talk about needing faster times. Part of this could be just defining fast I guess.  For instance, my Alan Smart C1, when at it's slowest attack is at 30 milliseconds, which is pretty fast.  But then there is the Tube Tech CL1B, where I love to have that attack knob fully clockwise usually (especially for vocals).  I like to set that knob near the middle of the range (12 oclock) for drum overheads.  Grabs quite violently then. So my question to all is; When do you use fast attack times?  On what sounds?  AND with what compressors?  How much gain reduction?  What release times in conjunction?

Fast attack times work great for nylon string guitar or Pizz Violin/cello and upright bass.  Anything that is really "plucky".  It tames the attack and brings out the tone.

[Nicky D]

I also think that fast attack times are best served in a back buss capacity, but maybe I'm missing something too.  I generally always lean towards slow  as I don't like the way a compressor tames(squishes) the transient...now transient designers are a different beast...I generally like using them as room/tail killers...