EQ & Compression

[DATASOUND]

In a sound system application: Just wondering in what order others use EQ & compression.

Do you like the EQ first or the compressor first? And Why?
Or, do you feel there is no difference?

[RamRod 1]

I don't know. I think that I would EQ 1st. If I didn't have an analizer I would 1st get a mix level, and then very very carefully bring up the master on the mains until I could hear a particular frequency starting to feedback and then try to find it with the EQ and pull it down a bit. Then continue doing that until I got everything fairly flat.

Then add some compression.

NEXT!

[DATASOUND]

Thanks RamRod,
I'm just trying to get different perspectives on how others prefer to do things. And why?

But to respond to your reply, I'm not wondering in what order one may choose to dial them in or how they would dial them in. I'm curious to know, in what order others are hooking them up in the signal chain. Wether its inserting a compressor on a channel(usually before the channel EQ) or inserting a compressor on a group(after the channel EQ's).

Since it sounds like you are talking about the main PA system, that strikes another question.

If you(any reader) had a compressor and graphic EQ on the mains, in what order would you hook them up? And why?

[LHSL]

In a sound system application: Just wondering in what order others use EQ & compression.

Do you like the EQ first or the compressor first?

Datasound,

It really depends on what you are trying to achieve. I'll first start off by saying that using a compressor on the main mixes or the monitor sends is a very very bad idea. Here's why:

When you compress a signal, what you are effectively doing is reducing its dynamic range. This will reduce the volume of the loudest peaks. Many people will then raise the overall gain of the system then (raising the master faders, or the output gain of the compressor to "make up" the lost volume. This ends up raising the average power of the signal over time. If you follow the 2x rms or sometimes 4x rms rule as some people do, on sizing your amplifiers, AND you use compression, beware. You could cook your voice coils very quickly.

This is not to be confused with limiting. Limiters are very fast, high ratio compressors (ideally infinity:1) that catch extreme peaks from allowing the cones to jump the gap.

Back to the compression before or after EQ deal. On the channel strip when you insert a compressor into the mixers insert jack, it is almost always pre-EQ. This is desired. You want to control dynamics before you shape the sound.

Here's why:

If you boost a frequency on the channel strip EQ, and the compressor was post-eq, when ever that frequency was loudest it would cause the compressor to kick in if it was above the threshold. (This can actually be used as an effect, I'll explain later).

You'll notice that most people open up the gates and compressors when they are "voicing" a channel or trimming gain or EQ. This allows the signal path to be unaffected by the dynamics processing during that time.

Using post-EQ dynamics as an effect:

This works much like a "De-esser." If you have a particular frequency range that is giving you trouble when the signal gets to a certain level you can use this to your advantage.

Buy boosting a frequency before the compressor, the compressor will react first to that frequency. This is largely not used because it affects the normal signal as well. However, you are probably familiar with a "side-chain" input. By inserting an EQ on the sidechain of a compressor, you can boost the freqency you want to dynamically remove. That is, boost 6k on the EQ in the side chain, then set threshold so that whenever the "esssy" singer sings, it will kick in and squash the "esssssssss."

To sum up. normal operation is dynamics first, EQ second. For certain effects you can do EQ first dynamics second (to control proximity effect or something). But by in large, that will be done using a sidechain input.

[lonewolf]

In general, you want the compressor before everything else in a signal chain, whether its for a channel or out of a mixer's main (limiter) or even a guitar.

On a PA, the compressor establishes the signal's dynamic operating range. That way you can keep the signal from clipping and/or overloading the rest of the equipment. Boosting the EQ into a compressor can also have some undesirable effects like causing the compressor to kick in when you really didn't want it to.

[DATASOUND]

Your reply did answer my question in a round about way by telling me how you prefer to do it.
I do have a console that has (pre EQ) and (post fader) send and return jacks on every channel. Yes, 4 jacks not including the line & XLR jacks. So I do have the option to insert a compressor at either point of the signal chain. But that isn't the norm, I agree. This doesn't even address putting a compressor on a subgroup.

Both are great reply's. Enjoyed

As LHSL said, "It really depends on what you are trying to achieve". This I know, and was exactly what I was fishing for.
The problem I have with the idea behind JUST using a compressor Pre EQ is this. Say your mains are flat(insert target curve here) for the room. To taste of course because we all know this can't be true everywhere in a room. Anyway, you compress first bringing the dynamic range to a desired range. Now if you do any EQ to a vocal or an instrument after the compressor, this will affect the frequency response of that signal. Now when the person singing or playing goes to a note you have cut or boosted, the actual level of that signal will change accordingly. So at some notes it will be hot, even though your system is flat. At other note the signal may drop below the noise floor. Even though your system is flat

So basically if you compress first, then EQ, your boost or cut will increase the dynamic range again. Think of it. Say you compressed a guitar (inf/1). Not that you would, but just to make it easier to understand the point I'm trying to make. So no mater what goes in, only one leval will come out. Then you boost 440hz example, then when they play that note, it will be louder then other notes. Hence you have just increased the dynamic range. Grant it, this increase in dynamic range would be far less then that of a singers voice getting louder and softer at the mic.
Putting the compressor after the EQ would still control dynamic range.

My experience has been if you insert a compressor on a channels strip, it will not only be Pre EQ but it is usually Pre Aux's too. If your running monitors from FOH, you will be compressing the monitor signal (for that channel) too. That's not the best thing to be doing. That's another topic in itself.

One down fall to inserting a compressor on a subgroup would be how one very dynamic instrument/vocal can upset the whole apple cart.

As for using the side chain, I have done it. But only as a de'esser.

I do use some compression on the mains when I'm running a small system that can't handle the dynamic range. (big room, loud band, small PA). Just enough to keep it above the noise floor and out of clipping like you said lonewolf. 3dB of reduction usually get that job done.

Anyway, think about my view on it, and let me know what you think?

[LHSL]

The problem I have with the idea behind JUST using a compressor Pre EQ is this. Say your mains are flat(insert target curve here) for the room. To taste of course because we all know this can't be true everywhere in a room. Anyway, you compress first bringing the dynamic range to a desired range. Now if you do any EQ to a vocal or an instrument after the compressor, this will affect the frequency response of that signal. Now when the person singing or playing goes to a note you have cut or boosted, the actual level of that signal will change accordingly. So at some notes it will be hot, even though your system is flat. At other note the signal may drop below the noise floor. Even though your system is flat

So basically if you compress first, then EQ, your boost or cut will increase the dynamic range again. Think of it. Say you compressed a guitar (inf/1). Not that you would, but just to make it easier to understand the point I'm trying to make. So no mater what goes in, only one leval will come out. Then you boost 440hz example, then when they play that note, it will be louder then other notes. Hence you have just increased the dynamic range. Grant it, this increase in dynamic range would be far less then that of a singers voice getting louder and softer at the mic.
Putting the compressor after the EQ would still control dynamic range.

Datasound,

I think this subject is just kind of wrapping around itself. Step back and think... if you have a frequency boosted, yes, it will be "hotter" than the others, but that's what you wanted? right? You boosted it.

I wouldn't think to hard on it. Most compressors like a 166, 1066, etc. aren't going to react so fast as to only ever squash or make a certain frequency very pronounced. They are more of an averaging type compressor. In the end, the whole signal get's squashed that much on average. That's probably not the right word for it.

If you are looking to tame certain frequencies, you might want to look at a multiband compressor. But I just don't think it's necessary.

If you have equipment of a reasonable level of quality (mostly dependent on speakers and microphones) you should not have much to worry about.

Just remember a few things...

1) the channel EQ is there to get an input ship-shape. boomy this, essssy that, harsh this, dead that. You cut out what is bad, and maybe sweeten what is good a little

2) compression on the channel strip is only meant to HELP control the dynamics. Lonewolf mentioned that it "establishes" your dynamic range. That isn't really true. The SOURCE establishes the dynamic range... compression helps manage it and fit it in to your "dynamic range budget."

3) compression on the outputs is generally not a good idea. If you use it, avoid high ratios, and your pronouced frequency problem will not be an issue.

My experience has been if you insert a compressor on a channels strip, it will not only be Pre EQ but it is usually Pre Aux's too. If your running monitors from FOH, you will be compressing the monitor signal (for that channel) too. That's not the best thing to be doing. That's another topic in itself.

Let's address it. Short and sweet, you are correct. This is the #1 reason I tell musicans to stay away from compressors. Leave that to sound guys like you and me. Compression on monitors sends only aggrevates feedback (or an unstable system ready to feedback). If they use a compressor on the main outs, they cook voice coils. Compression and Musicians don't mix IME.

If you must have compression on the channel strip AND are doing mons from FOH, the best way to solve that is just split off the vocal channels in to two strips. I pretty much ALWAYS run two strips for each vocal when doing mons from FOH. I run one strip for mains, and one strip for aux sends. No compression on the aux sends. AND you get a whole extra EQ section to customize the monitor signal for a different set of speakers and help tame feedback in rare instances.

One down fall to inserting a compressor on a subgroup would be how one very dynamic instrument/vocal can upset the whole apple cart.

Agreed, I generally never do compression on a subgroup. I do often use my backing vocals subgroup to feed an extra effects processor.

I do use some compression on the mains when I'm running a small system that can't handle the dynamic range. (big room, loud band, small PA). Just enough to keep it above the noise floor and out of clipping like you said lonewolf. 3dB of reduction usually get that job done.

If you use a compressor on the lead vocals, and use a limiter before the amps, you will get 100% of the same results, and never have to worry about cooking a voice coil. Limiters are stop lights though... once you dial them in, don't run them into red. You will pay dearly. I still stay compression on the main outs is a dangerous game of russian roulette. Especially if things get out of hand.

[lonewolf]

Now if you do any EQ to a vocal or an instrument after the compressor, this will affect the frequency response of that signal. Now when the person singing or playing goes to a note you have cut or boosted, the actual level of that signal will change accordingly. So at some notes it will be hot, even though your system is flat. At other note the signal may drop below the noise floor. Even though your system is flat

You can't EQ individual vocal notes without a precise parametric EQ with the Q set at minimum. I can't imagine any situation why you would ever want to set up a notch gain filter at any particular note, especially on vocals. With vocals, the goal should be to increase clarity and decrease sibilance while maintaining the desired tone. This 1st 2 involve frequencies that are outside the vocal range and aren't generally affected by different vocal notes. With good microphones, the 3rd takes care of itself with just a little tweeking.

There is such a thing as over-processing.

[lonewolf]

2) compression on the channel strip is only meant to HELP control the dynamics. Lonewolf mentioned that it "establishes" your dynamic range. That isn't really true. The SOURCE establishes the dynamic range... compression helps manage it and fit it in to your "dynamic range budget.

I didn't say dynamic range, I said dynamic operating range. When you put the source signal thru a compressor, you have established a new, lesser dynamic range to operate or "manage" as you said.

The dynamic range of a system is measured at the output, not the input and is established by every component that the source signals encounter.

[DATASOUND]

Now if you do any EQ to a vocal or an instrument after the compressor, this will affect the frequency response of that signal. Now when the person singing or playing goes to a note you have cut or boosted, the actual level of that signal will change accordingly. So at some notes it will be hot, even though your system is flat. At other note the signal may drop below the noise floor. Even though your system is flat

You can't EQ individual vocal notes without a precise parametric EQ with the Q set at minimum. I can't imagine any situation why you would ever want to set up a notch gain filter at any particular note, especially on vocals. With vocals, the goal should be to increase clarity and decrease sibilance while maintaining the desired tone. This 1st 2 involve frequencies that are outside the vocal range and aren't generally affected by different vocal notes. With good microphones, the 3rd takes care of itself with just a little tweeking.

There is such a thing as over-processing.

Just so you don't think I'm stupid, that example was a dramatization to emphasize a point. Maybe "note" wasn't the right word. "Frequency" maybe?
My scenarios aren't very good but I hope they helped.

Thanks for replys. The only thing I enjoy more then talking about sound, it doing it.

[DATASOUND]

LHSL, I'm probably digging to deep into all this. This really isn't anything I'm worried about or having problems with. Just looking for other perspectives on the subject. Sometimes I use a compressor before, sometimes after. As you mentioned before, "it depends on what your trying to accomplish". I have noticed most engineers prefer to use it before the EQ. And some that ONLY use a compressor before the EQ. That's what sparked this topic for me. Visiting BE's that refuse to run a compressor after the EQ. I had to cut about 4 feet of the jacket off of one of my patch snakes because he insisted on having the comps inserted into a channel. I can understand if he was subgrouping multiple vocals but it was just one vocalist. And he needed comps inserted on the drums(other end of console) and he refused to move channels. So, I had to strip the jacket of my patch snake to give me a longer fan reach. No, I didn't have any long 1/4" cables on hand. He tried to convince me that was the only way to run a compressor. And after hearing the job he did, his debate carried no weight.
In this profession, there are many things that have a "right/wrong" way. But I'm still not convinced that this application has a right or wrong way.

As I said to lonewolf, Thanks for replys. The only thing I enjoy more then talking about sound, it doing it.

[LHSL]

2) compression on the channel strip is only meant to HELP control the dynamics. Lonewolf mentioned that it "establishes" your dynamic range. That isn't really true. The SOURCE establishes the dynamic range... compression helps manage it and fit it in to your "dynamic range budget.

I didn't say dynamic range, I said dynamic operating range. When you put the source signal thru a compressor, you have established a new, lesser dynamic range to operate or "manage" as you said.

The dynamic range of a system is measured at the output, not the input and is established by every component that the source signals encounter.

On a PA, the compressor establishes the signal's dynamic operating range.

That is what you said... "establishes" It establishes nothing about the signal (signal, your word). The signal is the signal, it has a dynamic range. Call it dynamic operating range, I don't care. That's the same thing.

A compressor mangages (manages, my word) the dynamic range of a signal to fit it into the dynamic range of the system. Which is just another way of saying what I originally said. Maybe you meant something else, but your words were poorly chosen then. I appologize if I mis-interpereted what you meant.

[LHSL]

I have noticed most engineers prefer to use it before the EQ. And some that ONLY use a compressor before the EQ.

I've never seen an engineer do it any other way in practice.

That's what sparked this topic for me. Visiting BE's that refuse to run a compressor after the EQ. I had to cut about 4 feet of the jacket off of one of my patch snakes because he insisted on having the comps inserted into a channel. I can understand if he was subgrouping multiple vocals but it was just one vocalist. And he needed comps inserted on the drums(other end of console) and he refused to move channels.

I would have wanted the same thing changed on the system... but I would have just probably bypassed the house graphic and called it a day. (if the system was of decent quality)

If there was only one comp available, I would have insisted it be used for input strip on vocals as well. I would have worked with you on the channel order however. But, I usually have some insert cables with me.

[lonewolf]

No, I meant what I said. By definition, a compressor is a device that establishes limits to reduce dynamic range. Mathematically, that's all it does. If you want to call that management, that's fine with me.

I did say signal...I just didn't say source. The signal is no longer the source once it goes thru the compressor.

[LHSL]

No, I meant what I said. By definition, a compressor is a device that establishes limits on dynamic range. Mathematically, that's all it does. If you want to call that management, that's fine with me.

I did say signal...I just didn't say source. The signal is no longer the source once it goes thru the compressor.

I'm not about to get in a pissing contest with you on this stuff. However, a compressor does NOT establish limits on the dynamic range.

If I have a signal with a 80 dB dynamic range, a noise floor of -50dBu and I compress it 2:1 starting at 5dBu, the output would reach 17dBu with compression and 30dBu without. If that signal changes to 100 dB of dynamic range with the same noise floor the output would reach 27.5dBu with compression and 50 without. Most pro equipment can accept up to about 20dBu in most instances... sometimes a little less, some times a little more.

The point is, a compressor limits nothing. The signal still has a dynamic range.

Something that limits the dynamic range of a signal... that would be a limiter. Which is essentially a compressor with a very fast attack and a infinity to 1 ratio. Some compressors double as limiters. But in essence, a true "compressor" does nothing to limit the dynamic range of a signal.

The only way to ensure your amps never clip is to use a limiter. And when a limiter says red... stop. Running a compressor all night on the main outs of a system is a bad bad bad idea. If the system is already too small for the gig, then all a compressor will do is exacerbate the problem. If you are into the compressor at all (i.e. gain reduction) you will load the speakers with a higher average power over time. This will cause the voice coils to heat up and melt over years of use, or maybe possibly the first time you try it, if you hit the system too hard.

If you have a compressor on the outs and are never in gain reduction (no red) then just pull it out of the signal chain and use it where it counts, or if it has the ability to be set as a limiter, use it as a limiter.

This stuff is pro sound 101.

[lonewolf]

http://en.wikipedia.org/wiki/Audio_compressor

I can't help you, but I can recommend Carnegie-Mellon and PSU as two excellent local engineering schools.

[LHSL]

http://en.wikipedia.org/wiki/Audio_compressor

I can't help you, but I can recommend Carnegie-Mellon and PSU as two excellent local engineering schools.

Oh, so now we do the typical rockpage thing and make it personal. Real cute, real professional.

from your cited link:

Audio level compression, also called dynamic range compression, volume compression, compression, limiting, or DRC (often seen in DVD player settings) is a process that manipulates the dynamic range of an audio signal.

A compressor reduces the dynamic range of an audio signal if its amplitude exceeds a threshold.

Do you see the word "establish" or "limit" in the context of compression?

They go on to say:

A limiter is a compressor with a higher ratio, and generally a faster attack time.

Which is exactly what I said, twice, in two seperate posts before this one.

You are not exactly helping your argument here.

[lonewolf]

Like I said, I can't help you, but if you go to a good engineering school, you might understand my use of the term "limits". If you are a pro-audio enthusiast without a formal engineering background, you probably think I am referring to an audio limiting per se, but I am not.

"Limits" is a mathematical/engineering term used by engineers to describe constraints in any system whether its robotics, pneumatics or audio, etc. If there ever was an element in a system that imposes limits or constraints, it is a compressor.

My use of "establish" simply means to "set up"or "bring about". I'm sorry if my vocabulary is much more resplendent than the typical pro-audio enthusiast.

That's all I have to say about the subject. If you want to know more, you will simply have to go to college.

[LHSL]

Like I said, I can't help you, but if you go to a good engineering school, you might understand my use of the term "limits". If you are a pro-audio enthusiast without a formal engineering background, you probably think I am referring to an audio limiting per se, but I am not.

"Limits" is a mathematical/engineering term used by engineers to describe constraints in any system whether its robotics, pneumatics or audio, etc. If there ever was an element in a system that imposes limits or constraints, it is a compressor.

My use of "establish" simply means to "set up"or "bring about". I'm sorry if my vocabulary is much more resplendent than the typical pro-audio enthusiast.

That's all I have to say about the subject. If you want to know more, you will simply have to go to college.

Thank you for demonstrating to everyone that you are both childish and incapable of admitting when you failed in your word choice.

All it would have taken was a simple explanation of the word and the argument would have ceased. Instead, you chose to let it drag on and let doubt linger, trying to prove you were "the bigger man." Now you choose to be pedantic about word choice. I'm a reasonable man, and when someone says "I'm sorry, by saying X, I actually meant the "other" meaning X" or, "sorry, I mis-spoke, I meant Y," I pretty much cop it up to the human condition. We all make mistakes.

I'm familiar with mathematical limits. They do teach us that sort of thing, even though I majored in Comp Sci. Your condescending tone is unwelcome, and really uncalled for. And even though I am a college grad, how the fuck am I supposed to know what you meant by "limits" unless you say so. We are talking audio, not mathematics. Though now it seems convenient for you to claim this and remove all responsibility of yourself.

To cite an example, here's your original post:

In general, you want the compressor before everything else in a signal chain, whether its for a channel or out of a mixer's main (limiter) or even a guitar.

On a PA, the compressor establishes the signal's dynamic operating range. That way you can keep the signal from clipping and/or overloading the rest of the equipment. Boosting the EQ into a compressor can also have some undesirable effects like causing the compressor to kick in when you really didn't want it to.

Where exactly did you use the word "limits" in the context of mathematics? As in, as x approaches 1 the limit of f(x) approaches infinity?

The only word I see there is "limiter."

[DATASOUND]

Chill guys.
I wasn't trying to start a feud here. A debate is fine, but lets not get carried away with word usage.
"establishes"... "manages"... either or, I know what you both meant.

"Can't we all just get along"

Oh, and in case you didn't understand my OP question?

In a sound system application: Just wondering in what order others use EQ & compression.

Do you like the EQ first or the compressor first? And Why?
Or, do you feel there is no difference?

I wasn't asking for a lesson on what compressors do, or what they are used for, or how they work. Simply tell me how you prefer to put them in the signal chain. And why?

[lonewolf]

Chill guys.
I wasn't trying to start a feud here. A debate is fine, but lets not get carried away with word usage.
"establishes"... "manages"... either or, I know what you both meant.

"Can't we all just get along"

Oh, and in case you didn't understand my OP question?

In a sound system application: Just wondering in what order others use EQ & compression.

Do you like the EQ first or the compressor first? And Why?
Or, do you feel there is no difference?

I wasn't asking for a lesson on what compressors do, or what they are used for, or how they work. Simply tell me how you prefer to put them in the signal chain. And why?

Hey Dean, I apologize to you if I helped to take your thread out of whack. It was not my intention to get into an unnecessary battle of nomenclature. As long as you knew what I was talking about, that should have been enough.

[LHSL]

In a sound system application: Just wondering in what order others use EQ & compression.

Do you like the EQ first or the compressor first? And Why?
Or, do you feel there is no difference?

I wasn't asking for a lesson on what compressors do, or what they are used for, or how they work. Simply tell me how you prefer to put them in the signal chain. And why?

I'm sorry for getting into so much detail. Based on the question, I thought a little explanation was in order. It always bothers me when people told me "do it this way" with no explanation. I like to provide some background.

To put it bluntly, you will nary run into a sole that uses compression directly after EQ. If they are trying to de-esss something, they will either use an EQ in the side chain of compressor, or just use a purpose built de-esser.

You will also very rarely run into anyone that truly understands the concepts of compression, and what it is really doing, that will put one on the main outs of a system. Let alone run it into any kind of gain reduction. So before or after EQ on the main outs? not even applicable in my opinion.

When used on the group outs for say backing vocals, the best solution is a multiband compressor. Mostly due the complexity of the signal at that point. Yeah, that's sort of post EQ... but now it's a new combined signal. I would personally rather have enough 166s or 1066s to go around and use on the individual channels pre EQ.

So, EQ before compression? Pretty much never for me.

[HBSPro]

For a Front of house Rig Live you would Never Compress before you EQ,
compressor After the Eq Signal. that way your not squashing the signal your trying to clear up.

another way is to insert the compressor if you do not want to run it into the chain,

I use a Soundcrafty Mixer into 2 Ashly GQX 3101 then into a DBX 1066 with the Compressor not in Stero mode. with about 2.1 Ratio attack and release on Auto. into 2 BSS FDS 340 Crossovers Runing the Rig in Stereo.

[LHSL]

For a Front of house Rig Live you would Never Compress before you EQ,
compressor After the Eq Signal. that way your not squashing the signal your trying to clear up.

another way is to insert the compressor if you do not want to run it into the chain,

I use a Soundcrafty Mixer into 2 Ashly GQX 3101 then into a DBX 1066 with the Compressor not in Stero mode. with about 2.1 Ratio attack and release on Auto. into 2 BSS FDS 340 Crossovers Runing the Rig in Stereo.

HSBPro... No knowledgable engineer will run a rig this way. FOH EQ is for system tuning. The goal of a system EQ, crossover (or a loudspeaker management system) is to provide a linear, flat system. A white canvas if you will.

Mix compression, channel EQ, channel compression, gates, fx, all make a "mix" That "mix" should sound good on any medium. Tape, house, system, headphones, etc.

If you are fixing a "mix" with FOH EQ, you are shooting yourself in the foot.

That said, arguing over where to put a compressor on the main outputs is a bit rhetorical. It's unecessary and is likely to lead to cooked drivers on a small system, and just totally F's with the dynamics on a large, capable system.

[HBSPro]

Excusse me LHSL im sorry for not posting my wording exactly like you can understand it. but to tell me i can not run my rig this way is bull Compressors are used for Speaker management also. there is nothing wrong with using a compressor after your house Eq . if there was why do you see it on all the riders. I agree that you can over compress something . but from mixer to EQ to Compressor out to the mains or your crossover is a common thing. and you do not need to go to school for that.

be cause a guy ask someone how they do something or why is not the reason to bash the hell out of a guy. you can over Compensate for alot through out a system. and maybe the term cleaning up was not the proper, Wording. I end up with a Flat EQ using my set up the way it is. and use my FOH EQ for Removeing things not to boost them.