Login / Register
 
Opto vs electro compression
New Reply
Subscribe
gritzildino
Thread Starter
#1
4th February 2011
Old 4th February 2011
  #1
Gear maniac
 
gritzildino's Avatar
 
Joined: Sep 2010
Posts: 170

Thread Starter
gritzildino is offline
Opto vs electro compression

Ive been reading some waves manuals on C4 and Linearphase Multiband. They talk about two different compression settings Opto and electro... I dont really understand the uses for the Opto settings. The Electro settings make sense to me.
Well. First it says in the manual that an opto compression has fast release times at high gain reduction and slow release times at zero gain reduction...
"release times"!? Well, I am pretty sure that instead of "release time" they must mean like high ratio reduction and low ratio reduction instead of release times... I mean just from looking at different compressors like my logic compressor, i can see that "opto" curve in my head. And they surley must mean different rations as the release tmes are set on another parameter...
Anyway ... the uses for electro, I think I understand. As the sound gets louder it becomes more squased. So u can use this to get a loud lower in dynamics master for example....
But the opto.. it compresses more when it is at lower volumes.... I cannot really think of a use for this... Maybe, Sidechaining against a kick or somthing.... So that the volume level will jump down really quickly or somthing.. IDK.
Any one getting good use out of opto compression?
#2
4th February 2011
Old 4th February 2011
  #2
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
The manual is stating the case correctly: opto compressors have a variable recovery (release). The harder you dig in, the quicker it releases from the deep part of the compression. This can help them to be more 'transparent' and less prone to pumping when doing more intense gain reduction on highly dynamic sources like vocals.

Think of it this way: if you do, say, 8db of reduction on a peak, a typical opto will release the first 4db very quickly, then release the remaining 4db's much more slowly. So if the comp needs to react to another fast peak, it can do so quite handily, and it's not doing a full-on squeeze-and-release in the process which can be distracting or aggressive and may not be the desired effect or the right vibe.

I love opto's!


Gregory Scott - ubk
__________________


Kush Audio: Where High End Keeps Getting Higher

.........

Kush Audio: Where High End Keeps Getting Higher
____________________
#3
5th February 2011
Old 5th February 2011
  #3
Gear Guru
 
AllAboutTone's Avatar
 
Joined: Nov 2005
Location: South
Posts: 12,767

AllAboutTone is offline
I would like to hear even more on this topic.
#4
5th February 2011
Old 5th February 2011
  #4
Gear addict
 
Surfkat's Avatar
 
Joined: Dec 2005
Location: Va / NJ
Posts: 406

Surfkat is offline
At my place it's Opto on the vocal ALL the time, hardware or software.
__________________
"After silence that which comes nearest to expressing the inexpressible is music." - Aldous Huxley
#5
5th February 2011
Old 5th February 2011
  #5
Gear Guru
 
Joined: May 2008
Location: Mountain View, CA
Posts: 12,682

Dean Roddey is offline
I wasn't aware of the initial partial release being faster as the GR goes up. I thought it was fairly constant with a little variation, and the secondary release is the GR dependent part? The LA-2A, being kind of the canonical example, has an initial release to about 50% that's from like 40ms to 60ms depending on the state of the optical cell, and then it can be anywhere up to a few seconds to get back to zero if you hit it pretty hard.

That's one reason why so many folks use them in conjunction with an 1176 or other fast compressor because it's so nice to hit it kind of hard and therefore it stays in GR mode the whole time and just rides the signal, so you don't have to worry about the 10ms attack time, which can be pretty slow. But, if you have a break between sections of the song, it has time to reset and then the first syllable of the next line comes out way too loud

Maybe some others also have GR dependent initial release as well, I dunno.
__________________
Dean Roddey
Chairman/CTO Charmed Quark Systems, Ltd
www.charmedquark.com

Be a control freak!
#6
5th February 2011
Old 5th February 2011
  #6
Lives for gear
 
Ernest Buckley's Avatar
 
Joined: Feb 2006
Location: Brooklyn, NYC
Posts: 2,220

Ernest Buckley is offline
Quote:
Originally Posted by gritzildino View Post
Ive been reading some waves manuals on C4 and Linearphase Multiband. They talk about two different compression settings Opto and electro... I dont really understand the uses for the Opto settings. The Electro settings make sense to me.
Well. First it says in the manual that an opto compression has fast release times at high gain reduction and slow release times at zero gain reduction...
"release times"!? Well, I am pretty sure that instead of "release time" they must mean like high ratio reduction and low ratio reduction instead of release times... I mean just from looking at different compressors like my logic compressor, i can see that "opto" curve in my head. And they surley must mean different rations as the release tmes are set on another parameter...
Anyway ... the uses for electro, I think I understand. As the sound gets louder it becomes more squased. So u can use this to get a loud lower in dynamics master for example....
But the opto.. it compresses more when it is at lower volumes.... I cannot really think of a use for this... Maybe, Sidechaining against a kick or somthing.... So that the volume level will jump down really quickly or somthing.. IDK.
Any one getting good use out of opto compression?
I use opto most of the time. It works great on vocals, bass and the stereo buss. Electro is more for drums (fast attack with pumping sound). The Waves C4 manual is a beast. Once you figure out how to use that plug, you`ll absolutely love it. Then theres the C6...
__________________
"Shes So Ready"


You don`t need any more gear, you need to re-write the chorus.
#7
5th February 2011
Old 5th February 2011
  #7
Lives for gear
 
CJ Mastering's Avatar
 
Joined: May 2010
Location: South Florida
Posts: 2,646

CJ Mastering is offline
Gritz, just do some A/B comparisons and you should be able to hear the difference between opto and electro. make sure to have allot of gain reduction so you can hear both electo and opto at work.
__________________


Mixing And Mastering Studio -Professional Mixing & Mastering Studio Providing Online And Attended Sessions.
Mixed And Mastered Best Album Of The Year By En-Sound
#8
5th February 2011
Old 5th February 2011
  #8
has all the gear he needs
 
Unclenny's Avatar
 
Joined: Jan 2007
Location: Terra Firma
Posts: 7,938

Unclenny is offline
Quote:
Originally Posted by u b k View Post
......opto compressors have a variable recovery (release). The harder you dig in, the quicker it releases from the deep part of the compression.
"Honey, you spend hours reading those posts! Don't you pretty much know everything by now?"

Nope.

I know about opto compressors now, though.
__________________
"The main thing is to have a gutsy approach....but use your head." Julia Child

Leonard Scaper
Quote
2
#9
5th February 2011
Old 5th February 2011
  #9
Lives for gear
 
Sigma's Avatar
 
Joined: May 2006
Location: phallicdelphia
Posts: 4,624

Sigma is offline
opto's are and were REAL world compressors that had a certain release rate at greater than 3 dB compression and then a slowed down rate from that speed from 3 dB of compression to no compression..these characteristics were emulated by SOFTWARE plug ins

an ELECTRO was a name and concept created by WAVES software to do the inverse of the classic opto --ie release quicker as you approach no compression


it never existed in the non software world
__________________
"The notes I handle no better than many pianists. But the pauses between the notes, ah, that is where the art resides." Artur Schnabel


http://miketarsia.com

https://www.facebook.com/mike.tarsia
#10
5th February 2011
Old 5th February 2011
  #10
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
Quote:
Originally Posted by Dean Roddey View Post
I wasn't aware of the initial partial release being faster as the GR goes up. I thought it was fairly constant with a little variation, and the secondary release is the GR dependent part?
While it's true that the first stage of release in an opto like the LA2a is 'fixed', that doesn't mean it's always releasing at the same speed in that first stage.

Using 50ms for 50% recovery as the example: if the LA2a hits a peak of 16db reduction, it has to release 8db of its gain in 50ms. If it hits a peak of 8db of reduction, it has to release 4db of its gain in that same 50ms window.

Releasing 8 db in 50ms is a faster release than releasing 4db in 50ms. Same release time constant, faster release.


Quote:
it's so nice to hit it kind of hard and therefore it stays in GR mode the whole time and just rides the signal, so you don't have to worry about the 10ms attack time, which can be pretty slow.

This is a misunderstanding of what attack time is. Attack is not the amount of time it takes for a comp to 'grab' once threshold has been crossed, attack is the amount of time it takes the comp to reduce the gain by Xdb. X is fluid and, while the official spec is 10, I've spoken to designers who prefer 6 and one who prefers 8.

So attack time is always relevant and always affects the envelope even when a comp is swimming in reduction and the signal never goes below threshold. You can verify this by taking a vca compressor, settting fastest attack and release on a snare, burying the gain reduction meter, then opening up the attack time. Even though the meter stays in heavy reduction, you will hear the leading edge change shape as the attack gets slower. Same with a strumming acoustic that never drops below threshold, you'll hear greater transient definition as you open the attack.

This stuff is all pretty confusing, I still get confused by it and I'd be surprised if I'm not still misunderstanding some of the trickier aspects of the interactions. I know I get foggy when I start to factor in ratio and the way it interacts with attack and release, and how higher ratios multiply speeds even more. When you start talking about vari-mu's and the way the ratio increases as you go farther into reduction (deeper into the knee), I confess I can't fathom the math yet.


Gregory Scott - ubk
#11
5th February 2011
Old 5th February 2011
  #11
Gear Guru
 
Joined: May 2008
Location: Mountain View, CA
Posts: 12,682

Dean Roddey is offline
Quote:
Using 50ms for 50% recovery as the example: if the LA2a hits a peak of 16db reduction, it has to release 8db of its gain in 50ms. If it hits a peak of 8db of reduction, it has to release 4db of its gain in that same 50ms window.

Releasing 8 db in 50ms is a faster release than releasing 4db in 50ms. Same release time constant, faster release.
I think that's possibly just a semantic argument? In both cases it's 50%, and the perceived level change is the same in both cases, because it's all relative. So it's changed the same relative amount in the same amount of time. So I don't think that that's necessarily the right way to look at it?


On the other thing, I think you are maybe misunderstanding what I was saying. With a comp like the LA-2A, because of the saturation of the opto cell, once it's up doing like 6dB of gain, it's basically continually clamping down. The reaction to changes in the signal are obviously not instantaneous just because it's already compressing, but because of the slow release, it's still got plenty of compression going on already. So it's reaction to a change upwards in signal level while it's already in to substantial GR isn't anything like what it will be if you let it completely reset and hit it with the same level of signal. In the latter case it can overshoot badly for a pretty quick level change because it'll be doing almost no compression for a good bit of the 10ms attack time, whereas in the former case it's already compressing and the cell is well saturated so you don't get that kind of reaction.

The same applies to any compressor really, AFAIK. If you set up a compressor with a 200ms attack, and you feed it a signal when it's not in GR mode, then you do the same when it's already in GR mode, it's not the same thing. If that weren't the case, you wouldn't have to worry about seting the release too long and whacking the transient of the next beat. The reason you need to be careful of the release is that it can still be in GR mode by the time the next transient comes along.

Though, in the 2:1 ratio, the LA-2A's continual riding of the signal and somewhat slow reaction isn't the sort of thing that's going to remove all of the dynamics from a signal, and I guess that's why it's more like riding the signal in a nice way.

But anyway, I get what you are saying. The reaction time is the same. The difference is just that in one case GR is already going on, and you are just adding to it and keeping it in GR mode, and the other case it's doing no GR until it kicks in.
#12
5th February 2011
Old 5th February 2011
  #12
Lives for gear
 
Sigma's Avatar
 
Joined: May 2006
Location: phallicdelphia
Posts: 4,624

Sigma is offline
USE your ears..it can be confusing for anyone..especially dealing with complex dynamics levels and waveforms and real energry verses apparent voulme..

foor a monent when i go tprotools i looked at thetime factoras of attack and release of percussive stff and thenthe space between thenans sussed out some simple thing for certain pirposes.. yes it cool for some aplication to "retun to ) compression before th enext kick or snare hit ..but in otheres it's not..

hey when i compress overheads a lot of tomes i use a 150ms attack and a 1-1.5 second release ...at 1.5 to 1 and about 3 -5 dB of Squish.. the 150 lets the snare pass to a degree than the clamp sustains the cymbals longer and thicker ..inside that 1.5 release when the snare hits again the compressor is usually moving back up and doesnt get smacked down by the snare as t would with a fater attack rate

you have to listen close..and then realize,,if it's so insignificant to not be important F


now moving a release on a snare from 100 -200 or 500 .. there is a BIG difference just as going from super fast atack to 1 3 or 10 ms on the attack..

it's dependent on what you want to hear

damn i wanted a crush with a
solw attack and long sustain on a guitar part..everything was great exepth the intro of the power chords as it wasn't clamped..

so all i did was dupicate the audio a bar early...let the compressor clamp on that and then lifted that regions's audio outta the mix [kinda like "preloading the compressor"]

all the knobs interact...attack and release the same on the same program material but with varied compression either light or heavy will cause you to make fine adjustment on the attack and release setting
gritzildino
Thread Starter
#13
5th February 2011
Old 5th February 2011
  #13
Gear maniac
 
gritzildino's Avatar
 
Joined: Sep 2010
Posts: 170

Thread Starter
gritzildino is offline
"This is a misunderstanding of what attack time is. Attack is not the amount of time it takes for a comp to 'grab' once threshold has been crossed, attack is the amount of time it takes the comp to reduce the gain by Xdb. X is fluid and, while the official spec is 10, I've spoken to designers who prefer 6 and one who prefers 8."

Wow. I definitely thought that attack was the amount of time before the compressor 'Grabed' the signal above the threshold. Ha. So many instances in my tracks I set the attack to as low as possible because I didnt want the beginning of the sound unaffected but the tail affected by the compressor.... This is going to dramatically change the way i mix. Thanks. thumbsup
gritzildino
Thread Starter
#14
5th February 2011
Old 5th February 2011
  #14
Gear maniac
 
gritzildino's Avatar
 
Joined: Sep 2010
Posts: 170

Thread Starter
gritzildino is offline
Would you say, as an example, when you are compressing a master that you use faster attack/release on the higher frequencies and slower attack/release on the lower frequencies?
#15
6th February 2011
Old 6th February 2011
  #15
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
Quote:
Originally Posted by Dean Roddey View Post
I think that's possibly just a semantic argument? In both cases it's 50%, and the perceived level change is the same in both cases, because it's all relative. So it's changed the same relative amount in the same amount of time. So I don't think that that's necessarily the right way to look at it?

I do see the semantic issues created by my initial statement, you're right. I said the release is quicker the harder you dig in, which is ambiguous.

Think of it this way: if I travel 2 miles in an hour and arrive at point X, and you travel 1 mile in an hour and arrive at point X, our journeys were the same length of time, but I traveled faster than you. In the same way, the first stage release on the comp is always the same length of time, but the harder you've hit it the faster it releases.

So the underlying issue here is not semantic, it's critical for designers, and it's especially critical to understand this if you're trying to model analog behavior in dsp. But I do acknowledge that it isn't remotely critical to know any of this in order to write a song or craft a mix that touches people's hearts, so I make no pretense about what 'critical' really means in the larger context.



Quote:
So it's reaction to a change upwards in signal level while it's already in to substantial GR isn't anything like what it will be if you let it completely reset and hit it with the same level of signal.

That's true, but not because of the attack time. If you hit a comp with a signal that's 6db over threshold and it grabs, then while it's still 3db in reduction you hit it with another 6db peak of energy of equal bandwidth, it *will* respond identically in terms of the attack time and how long it takes to achieve the reduction it needs to do (usually the aforementioned 10db).

But it won't sound the same when grabbing that second volley of energy because it is working in a different part of the knee, so the curve of that reduction in gain has an entirely different shape (and, therefore, sound).

That transient pop you're used to hearing when a signal first crosses threshold is due to the interaction of the attack, the ratio, and the knee combined, and in my experience the knee is the strongest of the three in determining why one comp grabs differently from the next.


Gregory Scott - ubk
#16
6th February 2011
Old 6th February 2011
  #16
Gear interested
 
Joined: Sep 2009
Location: Lugano
Posts: 10

sirio.bm is offline
@u_b_k: This is the best definition of attack time I've ever seen. You are a god.
#17
6th February 2011
Old 6th February 2011
  #17
Lives for gear
 
Joined: Dec 2004
Posts: 1,223

Flymax is offline
To me, optos behave like human ears in how they react to volume.
#18
6th February 2011
Old 6th February 2011
  #18
Lives for gear
 
vernier's Avatar
 
Joined: Mar 2006
Posts: 6,367

vernier is offline
Variable-Mu for me. Or none.
.
.
#19
6th February 2011
Old 6th February 2011
  #19
Lives for gear
 
Piedpiper's Avatar
 
Joined: Jan 2009
Location: Between the Notes, Iowa
Posts: 3,085

Piedpiper is offline
Speaking of variable mu, how is a variable mu different from very soft knee with a high ratio?
__________________
Tim Britton
producer, engineer, musician, audio sales
http://www.piedpiperprod.com
http://uilleanpipes.com

row, row, row your boat...
#20
6th February 2011
Old 6th February 2011
  #20
Lives for gear
 
cinealta's Avatar
 
Joined: Sep 2009
Posts: 2,989

cinealta is offline
Quote:
Originally Posted by u b k View Post
So if the comp needs to react to another fast peak, it can do so quite handily, and it's not doing a full-on squeeze-and-release in the process which can be distracting or aggressive and may not be the desired effect or the right vibe.
So VCA does full squeeze/release cycle before re-squeezing a new peak?
#21
6th February 2011
Old 6th February 2011
  #21
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
Quote:
Originally Posted by cinealta View Post
So VCA does full squeeze/release cycle before re-squeezing a new peak?

Only if the program drops below threshold and the GR circuit has time to fully restore unity gain.

What I was getting at is that vca comps have a linear release (unless the designer purposefully tweaked it to be otherwise, a la Ruben's anti-log on the mpressor).

With a linear release, if you want the comp to recover quickly from heavy transients, you set the release to be short, in which case it will swing back towards full recovery much more quickly and often, which means the compressor artifacts will be more noticeable.

With opto's, they naturally tend to release the louder stuff faster and let go of the continuous stuff more slowly, so they're less likely to fully release dynamic signals like a voice or acoustic guitar. So they sound less grabby on these peaky sources, more natural.


Gregory Scott - ubk
#22
4th February 2012
Old 4th February 2012
  #22
Gear interested
 
Joined: Apr 2007
Location: Denver, CO
Posts: 11

yeahimsteve is offline
Where's the "like" button for this thread? Haha.......

This is good stuff. Thanks for the information, everybody!
#23
5th February 2012
Old 5th February 2012
  #23
Banned
 
Joined: Aug 2011
Location: Where Angels Live
Posts: 71

Mr. Shore is offline
Quote:
Originally Posted by u b k View Post

This is a misunderstanding of what attack time is. Attack is not the amount of time it takes for a comp to 'grab' once threshold has been crossed, attack is the amount of time it takes the comp to reduce the gain by Xdb.
Gregory Scott - ubk
I'm so tempted to say "I couldn't disagree more" with the above statement. Interestingly enough both versions of what an attack time is are equally offered as definitions. I still believe the word "attack" was used to precisely describe the time the compressor needs to "attack" the signal over given threshold. It would've been another word to describe the "time" the comp needs to compress to a certain rate. If this is what you're saying an "attack" is then "release" should be the time the compressor holds the signal from the level (as GR) reached by attack and specified by ratio until that signal reaches a below threshold value (0dB) and compression stops. I couldn't agree with this either...
#24
5th February 2012
Old 5th February 2012
  #24
Lives for gear
 
Old Goat's Avatar
 
Joined: Apr 2007
Location: Eastern Ozarks
Posts: 4,712

Old Goat is offline
Quote:
Originally Posted by REX View Post
Well the guy modifies Distressors and resells them under his own brand name so I imagine he knows what he's talking about.
Beat me to it.
#25
5th February 2012
Old 5th February 2012
  #25
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
Quote:
Originally Posted by Mr. Shore View Post
I still believe the word "attack" was used to precisely describe the time the compressor needs to "attack" the signal over given threshold.

You can believe it with all your might, but you'd still be wrong.

Attack is not a one-shot affair where once you're over threshold the attack has happened and doesn't happen again until you dip below threshold and come back over it once again; attack is happening *any time* a signal is increasing in amplitude while over threshold. IOW, if the compressor has to push back against a signal, it is attacking it. The question is how fast is it pushing back (attack) and how hard (ratio).

The easiest way to demonstrate the reality of attack is to drive a drum bus hard into compression at a high ratio, try 10:1, with the fastest attack and a fast release, say 50ms. Bury the signal in gain reduction, so that the signal never goes below threshold.

Now that you've got a signal swimming in compression, adjust your attack and notice how the envelope and the sound change radically. According to your definition (or at least the way I'm reading, which I admit might be wrong), attack would be irrelevant as the signal is always over threshold.

But that's not what happens, what happens is that the compressor is taking longer to achieve XdB of gain reduction, *regardless* of the fact that the signal is not crossing back and forth over threshold. Depending on how you've got it set up, a compressor can be constantly attacking and releasing a signal that goes over threshold and remains there.

Don't let the fact that I actually design these things convince you that I have a clue, because lots of designers don't really understand compressors any more than the average joe. Just do yourself a favor and run the above test, I suspect it'll clarify the situation.


Gregory Scott - ubk
#26
5th February 2012
Old 5th February 2012
  #26
Pragmatic Snob
 
u b k's Avatar
 
Joined: Oct 2004
Location: The Land of Sunshine
Posts: 12,246

u b k is online now
Quote:
Originally Posted by Piedpiper View Post
Speaking of variable mu, how is a variable mu different from very soft knee with a high ratio?

The ratio of vari-mu's gets higher the harder you drive them. For example, when you're just over threshold you're at a low ratio, say 1.5:1, but when you're 7dB into reduction the ratio might be more like 5:1 and by 15dB of reduction you might be at 20:1.

It can also be the case that the knee is continuously variable as well, I know the Manley's knee gets softer as the ratio gets higher. But knee is to ratio as barometric pressure is to humidity in that most people understand the latter but not the former, so it's easier to just talk about ratio.


Gregory Scott - ubk
#27
5th February 2012
Old 5th February 2012
  #27
Lives for gear
 
Joined: Dec 2006
Posts: 544

flatfinger is offline
"attack" could be called "gain reduction speed" When a real person rides the fader, he can pull it down gently or quick like once he's made up his mind to do so . The problem is that if he's quick draw magraw and is alway moving the fader in a darting fashion , it most times results in an audio artefact known as " Pumping ".


So once the detector in the side chain determines that the signal is over threshold , it says to the robot guy who has his robot hand on a ity bitty fader in yer comp ....turn it down beotch ! , then the "Gain Reductio speed " is applied ! ...( you know ; to the ity bity fader in there !)
attack setting is usually how long it's going to take him to get 2/3's of the way turning it down as much as the present gain law requires (the RATIO or slope requires...) so if you are at a 2:1 ratio and signal in is 10db only 5db is supposed to get out ,

( GUESS WHAT ??? if yer at a low ratio , you get to how far down it is to be turned down ( the full turn down !) more often than if you are at a high ratio !!!!! But of course threshold AND material come into play as well !!! BUT ; if you have a sky high ratio and less that fast " Gain reduction speed" you may never turn it down as much as the gain law says your supposed too !!!!!)

The thing is ( and this is the squishy part where different comps sound different even if the dials are sorta set the same ...) Have you ever had to replace those volume pots on yer guitar ?? ever made the mistake of putting linerar pots ( which made making fine adjustments at low volume a real bitch !) the "rate of reduction" of the "robot fader" in the comp can be varied and is part of the units signature . THe sony Oxford soft comp ( and some others as well) will even let you choose between logarythmic and linear laws.


UBK was explaining to us that the gain reduction device ( the thing acting as the fader ) has a real hitch in it's step if it's an opto !! Instead of a nice linear relationship ( like a pot that changes it resistance to the movement of it's shaft ( EVERY MM OF TRAVEL CAUSES CHANGE)....those dam things have a dead spot in their travel where nuttin happens !!!!! OK just an attempted analogy )

This was discovered by a real "happy accident " . Word got out mighty fast however .( you can't keep a secret in this town !!)


SO when you try to put a rigid THIS attack and THIS release it doesn't always work for music that has some stagger in it's step ( who gave that singer all those free drinks ???) That's why the properties that make the release "vary" with some factor are mighty helpfull ( Although the intredip amoungst you Could automate those compressor dials in your DAW Too !!





I first heard of all this goobly gook in a KJAERHUS Golden comp manual ( softy , now defunct) They called it PDR or "Program dependant release"

Quote:
Program dependent envelope times are used to reduce the pumping effect and maximize loudness.
Basically it will reduce the attack time if there is a short time of lower signal level, and reduce the
release time if there is a short peak with a higher signal level (such as a kick drum). This behaviour
is particulary desirable with mixed program material where it will apply fast and effective
compression to peaks while giving a more natural sounding release phase with lower harmonic
distortion after a harder or longer lasting compression. This effect was first discovered as a natural
memory effect in the LDR / El-panel set-up discussed earlier, and has since then been implemented
in alot of popular compressors.

auto release algorithms work in one of two ways (or both). One is related to the amount of gain reduction taking place - faster release time is allocated to smaller amounts of gain reduction and longer releases to large gain reductions to reduce pumping. Another method is for longer release times to be allocated for bass frequencies and shorter for treble to reduce wave gain riding distortion.







Don't forget to turn the knobs till it sounds good then .....
__________________
ENJOY THIS HYPERBOLE ; IT'S FREE!
#28
5th February 2012
Old 5th February 2012
  #28
Lives for gear
 
Piedpiper's Avatar
 
Joined: Jan 2009
Location: Between the Notes, Iowa
Posts: 3,085

Piedpiper is offline
Quote:
Originally Posted by u b k View Post
The ratio of vari-mu's gets higher the harder you drive them. For example, when you're just over threshold you're at a low ratio, say 1.5:1, but when you're 7dB into reduction the ratio might be more like 5:1 and by 15dB of reduction you might be at 20:1.

It can also be the case that the knee is continuously variable as well, I know the Manley's knee gets softer as the ratio gets higher. But knee is to ratio as barometric pressure is to humidity in that most people understand the latter but not the former, so it's easier to just talk about ratio.


Gregory Scott - ubk
Thanks. I understand what a vari mu is which is why I asked. How would your example in the first paragraph differ from a compressor set at 20:1 with a 15dB knee?
#29
5th February 2012
Old 5th February 2012
  #29
Gear nut
 
Joined: Feb 2011
Location: Sydney, Australia
Posts: 138

voxhumana is offline
If anyone is into DIY, this opto-compressor kit from JLM is killer. I've got two of the 500 series and am planning to get several more. There's also a stereo kit with power supply and rack kit (more expensive than two of the 500 series for obvious reasons).

JLM Audio Shop - LA500 opto comp kit

Price is $349 in Oz dollars (basically the same as US dollars at the moment)
#30
5th February 2012
Old 5th February 2012
  #30
Lives for gear
 
DrFrankencopter's Avatar
 
Joined: Sep 2003
Location: Ottawa
Posts: 978

DrFrankencopter is offline
Quote:
Originally Posted by Mr. Shore View Post
I'm so tempted to say "I couldn't disagree more" with the above statement. Interestingly enough both versions of what an attack time is are equally offered as definitions. I still believe the word "attack" was used to precisely describe the time the compressor needs to "attack" the signal over given threshold. It would've been another word to describe the "time" the comp needs to compress to a certain rate. If this is what you're saying an "attack" is then "release" should be the time the compressor holds the signal from the level (as GR) reached by attack and specified by ratio until that signal reaches a below threshold value (0dB) and compression stops. I couldn't agree with this either...

You could implement 'attack' as you described above on a digital compressor with relative ease, but it would be a challenge to implement in the analog domain (not impossible though).

Most analog compressors have their attack and release times established by a simple resistor plus capacitor network (RC circuit). See http://en.wikipedia.org/wiki/RC_circuit ...but don't get put off by all the equations, just skip down to the graphs showing the capacitor charging up. Note that the time at which the cap is 63% charged is indicated on the graph. This is the 'time constant' of the circuit.

The way a typical compressor side chain works is by:
1. calculating input level by turning the AC music signal into a DC signal (rectifier circuit)
2. Compare the DC level from step one against a threshold (this can be as simple as putting a reversed diode after the parts that convert the signal to DC...in this case any DC level below 0.7 V would not pass through to the next stage of the compressor)
3. Vary the gain of the thresholded signal from step 2 to implement a ratio control
4. Send the signal from step 3 to the attack/release RC circuits
5. Send the DC control signal from step 5 off to the audio path gain control circuit

Note at as soon as the calculated level exceeds the threshold there will be gain reduction, but the speed of it will be determined by the time constant of the RC circuit in the attack network.

Some compressors use paralleled RC networks to create more complex attack/release curves. Now you effectively have two time constants. Usually if you see an 'auto' release on a compressor that's what it is.

Now onto opos. Optos have their own set of time constants associated with them since their photo resistors do not respond immediately to changes in light level. These time constants work in conjunction with the time constants set by the RC networks. Plus as a further complication the time constant of the opto element changes as a function of the intensity of the light source.

See http://www.datasheetcatalog.org/data...lmer/VT500.pdf on page 9 for an example of this. The harder you drive an opto, the shorter its time constant is.

I hope this is sufficiently accurate, and I hope it helps.

Cheers

Kris
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+ 
 
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
pootkao / Geekslutz forum
31
e-cue / Geekslutz forum
14
pennyco / Mastering forum
25
jlsgear / So much gear, so little time!
5

Forum Jump

SEO by vBSEO ©2011, Crawlability, Inc.