This video is the second part of our exclusive series on compression controls. After learning about the compression basics such as threshold and ratio in the first episode, it's time to take it a step further and explore all of these lesser-known controls that can leave you puzzled.
But be puzzled no more. Fab Dupont takes it upon himself to demystify settings such as Hold, ARC, Look-Ahead, etc, and describe them to you as clearly as possible. As always, every compressor control is illustrated by down-to-earth audio examples, so you can truly understand what the setting does and its different applications.
This follow-up video is perfect for anyone comfortable with the basic compression settings, and willing to master the compression tool in its entirety.
Before watching this video, make sure you've seen the previous episode.
Once logged in, you will be able to click on those chapter titles and jump around in the video.
- 00:00 - Start
- 00:40 - Soft Knee
- 06:21 - Automatic Release Control
- 09:50 - Automatic Gain Control
- 12:08 - Wet/Dry Ratio
- 14:27 - Depth (a.k.a. GR Limit)
- 16:58 - Look Ahead
- 21:49 - Side-Chain Controls
- 24:25 - EQing the Side-Chain
- 28:55 - External Side-Chain Trick
Good morning children! Today, we're going to talk about Compression Controls, part 2.
What? Again? What do you mean? We just did that in Compression controls, part 1! One! Yeah, but we just scratched the surface. So here we go! Now that you are fully familiar with the basic controls thanks to the amazing Compression Controls Part 1 video, I'm talking about controls such as Threshold, Ratio, Attack, Release, and Gain, we're gonna be able to get into the road less travelled, such as Soft Knee. Contrary to popular belief, Soft Knee is not referring to engineers praying on their knees to try and get the right setting because they don't know what they're doing.
No, no, no, that's not it! What the Soft Knee control does is allow you to force the compressor to have a gradual transition between uncompressed and compressed signal.
Instead of having: Signal is not compressed / Threshold / Signal is compressed, you have a: Signal is not compressed / It gets louder... Oh! It gets compressed a little bit, and then more and more and more.
Look at this amazing graph! Alright! This first graph is regular compression.
The signal is below the Threshold, it's not compressed.
Oh! Now it's over the Threshold, it gets compressed. Amazing.
This is the second graph. This is with Soft Knee enabled.
Note the very gentle soft curve that looks like a knee, hence the name.
This signal is way below the Threshold, nothing happens.
The signal is getting near the Threshold. Ah! It gets compressed a little bit.
And this signal is over the Threshold, so it gets compressed as per the settings that you had even when Soft Knee was off.
Here's the amazing Will Knox.
Alright. So we have a verse that's at level x, and then we have a pre-chorus that's a little x+1 louder, which is fine, because we want the intensity to go up.
Then we have the chorus where he's literally screaming at us.
And that's a problem.
How would you set a compressor to handle this problem? In my personal case, I like that the pre-chorus is a little louder than the verse, and I like the dynamic that he chose. I'm not gonna change that.
However, the chorus hurts my ears, so I'm gonna have to take care of that.
The idea is to set the compressor up so it can handle just the chorus and not touch the verse and the pre-chorus.
So here's the Avid Pro Compressor.
And here's what it does.
You noticed that it does basically nothing.
Maybe 1 dB here and there, which is basically nothing.
I set the Threshold up for that.
I set it just above the level of the verse and pre-chorus vocals, so that the compressor is transparent on those 2 sections.
Now let's look at the chorus.
So now, the chorus gets slammed. Right? You could tell me: "Yeah, but what if you raised the Threshold and just handle the chorus?" Very good point! But that's not what I'm showing you right now.
In this particular case, I'm gonna show you how Soft Knee can be used to have an even, kind of transition between slightly compressed and more compressed, between the pre-chorus and the chorus. Check it out.
I'm gonna put the Soft Knee on stun.
Now we're using the Soft Knee, so the compressor is gonna start compressing below the Threshold.
Which means the verse, which was not compressed before, is gonna get more compressed. We have to compensate for that loss in level if we want to be able to compare apples to oranges.
So I'm gonna give it a little... kick! Now I have Soft Knee on stun, so a lot of Soft Knee.
You notice here in the interface that it shows you that it's a soft knee.
Everything else is the same, except I gave it 3 dBs of gain, so we can hear it at the same level.
So this is now the same exact segment, with the Soft Knee.
It's not a perfect setting, but notice how much smoother the transition is.
I'll play back the old one. I was at 5 dBs here, and no Knee.
So this is getting killed. And then now...
Obviously the Threshold is too low.
I'm gonna raise it a little bit. But now you can have a smoother transition.
It's so nice! How can you integrate this notion? Here's a way you could think about it.
If you want something that is smooth and hides the onset of compression, and more even between low and high compression settings, use a lot of Soft Knee.
Some compressors give you an amount of Soft Knee, some compressors just say On or Off.
This one gives you an amount of Soft Knee.
If you want to do special effects, and you want to do that snare thing that goes "Krr Rrr Rrr!", no Soft Knee.
Fair? Let's move on! Let's talk about Automatic Release Control, ARC, or Program Dependent Release, PDR, or Auto Recovery, which are all the same.
I really enjoy the Auto Recovery one, because it implies that the transition between compressed and uncompressed, which is the recovery, is a recovery from an illness for example, which is very fitting. So what does it do? Sometimes, it's hard to find one single Release speed for your program material.
Sometimes, you need long and short, and it's hard to have both at the same time.
But Auto Recovery gives you both at the same time. Let me show you.
Here's a snare drum. It is quiet, and then loud, and then quiet.
I play it for you.
Just like I said.
Now I'm gonna turn the compressor on with a fixed Release time.
Check out what it does.
Listen very carefully to the difference in the noise floor right after the hit, on the short snares, and the loud snares.
It makes a mess. Why does it make a mess? It makes a mess because the compressor clamps down on the snare.
When it has a very fast Release time, you get the clamping on the snare, then it comes back up right away. And what comes back up? All the garbage that's at the end of the snare.
On a quiet snare, it's ok.
On a loud snare, all that garbage comes up. It's a problem.
Let's listen to the same example, with Automatic Release Control.
You have to keep in mind that this is manufacturer dependent.
For example, here's a 33609 compressing this with a very fast Recovery.
It's a mess.
With A1, which is their first Automatic Recovery setting.
As a comparison, check it out with a fast recovery.
It's much better. They also have A2 which is also automatic, but slower, for other case scenarios.
I think A1 is better in this case.
There are other situations where Auto Recovery is very useful.
Say you're compressing a whole mix, the whole song.
Say you have a loud chorus.
You're probably using a slow Release time, so you're not pumping.
Now what happens at the moment when you hit the bridge, and the bridge is this big? It's gonna get clamped on, because the compressor doesn't know yet that it's very small, because you have a long Release time.
Automatic Recovery is great for that.
Conclusion number 1. You should spend a little bit of time listening to your different compressor's Automatic Recovery circuits because they're all different, because they depend on the taste of the guy who made it up.
Conclusion number 2.
If you're not quite sure, turn it on, it can't hurt.
While we are on the subject of automatic stuff, let's talk about Automatic Gain Control. What might that be? Good question.
When you're compressing, you are losing signal.
And it requires you compensating for that with the Gain Make-up.
You're compensating for all the signal that the compressor is taking away from you, fiercely. Now...
Wouldn't it be nice if the compressor knew how much level you are losing, so that it can compensate for that automatically? Hence the Automatic Gain Control.
In this particular case, I'm using the Fabfilter Pro C.
I did not choose this because of the name.
Let's listen to Wendy Parr sing very well without compression.
Let's compress the hell out of Wendy.
We're losing a lot of level. This is uncompressed.
Ah! But turn the Auto setting on, and... Uncompressed...
Compressed! So you saw the compressor is now guessing how much level I've lost, and automatically compensating for it the best it can.
You'll notice that it's not perfect, because if you think about it in the physical world, it's pretty impossible to imagine, because this is a dynamic process.
So it's pretty amazing that it already does this well.
The benefit of this, I think, is that when you're tweaking settings it can be a pain to have to think of so many settings: Threshold, Ratio, and then Gain to compensate.
On top of it, you have the Attack and Release, and all the other stuff.
It's kind of nice to have a ballpark level compensation while you're tweaking Threshold and Ratio for example.
Let me show you.
You noticed that I was playing between Threshold and Ratio, but I didn't really have to think about level because this circuit did that for me. More or less.
It's so nice! Speaking of Output settings, let's discuss the Wet/Dry Ratio, or Mix Ratio that you can find on some compressors these days.
What is that? Good question.
Don't tell anyone, but that is the magical parallel compression setting of some of those compressors. Let me show you.
Here's Wendy again. She's currently uncompressed, hence Dry.
Here she is Wet! Hence, compressed all the way.
Obviously, it's slightly over-the-top setting.
Why would I do that? Here's why! I can go from the Dry signal, uncompressed, to this extremely compressed signal, and blend the uncompressed and compressed signals together with one knob. Check it out.
Ok, now listen. This is the uncompressed signal.
And here's with the parallel compressed signal.
Notice how it's very steady and very dense, and stays in place more and has more steadiness to it, but you still hear all the transients.
That's because the transients are from the Dry signal, meaning the uncompressed, pure signal, you get all that detail, and then the density and the "Vrrr!" thing comes from the Wet signal, and the combination of the two is often very interesting to use, especially if you can do it with just one knob as opposed to send it to an Aux and do all that stuff that we showed you in other videos.
It is a very good idea to try that Mix/Dry Ratio stuff if your compressor has it, when you have a compression setting you like, but is a bit much, and you've already tried changing the Threshold and tried changing the Ratio, and it doesn't do it for you.
By going from 100% compressed and blending back in a little bit of the uncompressed signal, you can soften that effect, while keeping most of the characteristic of it.
Here's a compressor control that's actually relatively new.
It's not every day that you have a new control on your compressions, because all the other ones were not enough. Here's a new one.
It's called various things of course.
In some compressors, it's called Depth, other compressors it's called GR Limit, and I'm sure there's other names.
In this particular case, on my example in the Elysia, it's called GR Limit.
What does it do? It limits the compression... No! Not that way! It doesn't limit the compressed signal as in a limiter on the compressed signal, that would be confusing. Other compressors do that though.
It prevents the compressor from compressing more than what you tell it to. Think about it.
Say you have a vocal. Say it's Wendy Parr.
And say you set the compression pretty heavy, because you're looking for a special sound.
So you have a low Threshold, and a high Ratio.
And you're happy with that.
Except for this one moment where she leaned towards the mic and it makes the compressor go crazy. But everything else is cool.
What if you could tell the compressor: "Look. Do whatever you like, but never compress more than 12 dBs. Ever!" So the compressor could compress 8 dBs when it needs to, 10 dBs when it needs to, and when that enormous peak comes in, and it says: "Wow! That's a lot of level, let me go crazy!" it says: "Oh! But wait! I can't do more than 12." Let me show you.
Here's the dry vocal.
So you hear that "On its way," she leaned towards the mic, or maybe she turned her head to change the cue or something like that.
So I'm gonna compress this so I'm happy with the beginning.
You saw that, right? The "On its way" gets taken care of, but too much.
So I can take the Gain Reduction Limit setting... I turn it on, and say: "Ok, don't compress more than 7.6 dBs." How about... 9! See how natural that is? Without the Gain Reduction Limit, it sounds like this.
It's very subtle, but it's very cool.
Here's a control that is on a lot of compressors, and that a lot of people don't know what to do with. It is pretty simple.
So we're gonna go over it quickly. It's called Look Ahead.
It has absolutely nothing to do with what your mother insisted you do before you cross the street! No, no, no! Here is what it's for.
Sometimes, when you compress a lot, and you use extreme Attack and Release settings, you demand so much of the compressor that it doesn't really have the time to take care of it for you, and then freaks out, and then distorts.
Let me show you.
What's going on here? What's going on is that I am in a All-Button mood today, which is a mood you can't really shake, and then I want my compressor to have a fast Attack, to catch the fingers, and a really fast Release to make sure that it bounces nice.
And what does that do? That makes my compression distort.
As opposed to...
So I like how tight and bouncy it is with it in, but the distortion cramps my style. What to do? Well, you could slow the Release down.
It still distorts a little bit, because bass signal does that to compressors mostly.
Or, you could use what they call the Look Ahead.
What the Look Ahead does is it looks... ahead! Hence the name, Look Ahead! Fabulous! Because this is a computer, and it has control over time, you can tell it: "Look. We got time! Look ahead at what's going on, take your time, process, and give it to me without distortion, please." It sounds like this.
You'll see me raise the Look Ahead time.
Listen to the distortion, the edge on top of the sound go away.
It's not perfect, but this...
Is a lot better than this...
So with the Look Ahead, in extreme situations like this, you can actually get a lot closer to the ideal setting you had in your head, provided you had it in your head, than if you didn't have the Look Ahead. That's what it's for.
Another situation where you might find yourself pushing the limits of your compressor, and getting distortion, is when using extreme settings on bass instruments.
You could have your Attack and Release settings so fast that they are actually faster than the lowest audible frequency cycle.
And that sounds like this.
That's the clean bass.
That's compressed... to smithereens.
The first solution that comes to mind is to slow the Release down.
Or slow the Attack down. But we saw earlier with Look Ahead that it's not necessarily what you need. Sometimes you need that setting.
You just don't want the distortion.
What you could do is use the Hold function of your compressor, if it has it, to tell the compressor to hold a little bit between the Attack phase and the Release phase, and set the time so it's just long enough to be longer than the lowest audible frequency cycle.
It sounds like this.
I get to keep my Attack speed, I get to keep my Release speed, and I add a little bit of time in between to avoid distortion.
In my opinion, the most important thing about a compressor is not the kind of compressor it is, like Opto, or FET, or Vari-Mu, in my opinion, the most important thing is the Detector circuit, because the Detector circuit is what tells the VCA, the Vari-Mu, the FET or the Opto what to do.
So... What's more important? The person that does something, or the person that gives the order? In this case, it's who gives the order. And the Side-Chain gives the order.
Detector circuit and Side-Chain are the same thing, depending on what country you live.
Let me show you a few special Side-Chain controls that are really, really important to master.
The first one is kind of easy and makes a lot of sense.
I've seen it on UREI compressors from the 80's, and also on some new plug-ins that are starting to come out.
What's the difference between Peak and RMS? To simplify: Peak is how loud music is for your recorder, and RMS is how loud music feels coming out of the speakers.
I am definitely simplifying, but this is a good way to remember it.
Peak is how loud music is, RMS is how loud music feels.
You could decide to have your compressor's Detection circuit point at the actual real level, or at that average feel-like level.
Say you're compressing a whole CD's worth of material, 2-track material.
You might want to use Peak, because you have a certain edge you're trying to reach there. Right? Say you're compressing a vocal, and you want it to feel steady, and on average the same, use RMS Detection. I'll show you the difference between the two, and you must practice to really get the feel for it, because it's a feel thing. Check it out.
This is Wendy Parr. Uncompressed.
This is her compressed, with RMS compression.
Meaning, the Detector is looking at the RMS value of the signal, and telling the compressor: "Yo! This is what I think." And this is the exact same phrase with a Peak style Detector.
Obviously, this is much more compression, let me make up the gain.
Seen like this, it may seem a little subtle, and pointless, and kind of like technical for no reason.
But in some situations, only a Peak style Detection circuit will let you achieve what you need, and in some cases, only an RMS style Detection circuit will let you achieve what you need.
If you are limiting or crushing your track, it is very likely that your processor has a Peak style Detection circuit.
It does it behind the scenes, you don't know.
If you're leveling or averaging a track, very likely it's an RMS circuit.
Most processors don't give you the choice.
If you have the choice, I recommend you start with RMS, because it's kind of foolproof, and it's closer to what you hear, so your compressor hears what you hear.
If you have big peaks that bother you, I'll let you deduct the solution to that problem.
Within the context of our chapter 'Side-Chain is everything', let's discuss EQ'ing the Side-Chain.
As a reminder, the Side-Chain is the director that tells the compressor what to do.
What the Side-Chain hears is key in telling the compressor what to do.
Say if you have a very bright Side-Chain.
Then everything bright is even brighter and will make the compressor compress bright stuff more.
Say if you have a very fat Side-Chain.
Then everything that is fat is gonna make the compressor compress more.
It's kind of like driving a boat.
You've got to push right to go left, and push left to go right.
So say if you want to compress to highs more, you have to make the Side-Chain brighter, and if you want to compress the lows more, you've got to make the Side-Chain deeper.
Since every compressor already has its own kind of bias of the sound, based on the taste of the designer, it's nice for you to be able to impart your own taste of that.
And that is what EQ'ing the key, EQ'ing the Detector, or EQ'ing the Side-Chain is about.
In this compressor for example, there's a whole section down here with an EQ.
But what does it do? Well say I have this song.
And somebody decided it's still too dynamic.
So I turn the compressor on.
Listen to the bottom.
Every time the bass drum hits, the compressor freaks out because there's a lot of energy at the bottom of this track.
For a compressor, whether it's low or high it doesn't care.
It cares about the amount of energy.
So every hit at the bottom compresses the whole track a lot.
What if we told the compressor: "Don't worry about the bottom.
Don't pay attention to the bottom." How would you do that? By EQ'ing the Side-Chain, removing the bottom information of the Side-Chain, so that the compressor doesn't see it. It sounds like this.
I'm gonna turn the Side-Chain on, and I'm gonna listen to it.
Then I'm gonna pick a spot. This is my high-pass filter of the Side-Chain only.
The reason why this is blinking is to remind me: "Hey! You're not listening to the signal.
You're listening to the material that the Detector is listening to, from which decisions will be made about the level of your signal." Think about it. Let's listen to it.
So now, my Side-Chain is gonna hear this...
Instead of this...
You noticed that with it on, there's basically no bass drum left.
So what does it do to the actual final sound of the compression? We are now going to listen to this track compressed, with the Detector only listening to a little bit of the high end of the signal not the whole thing.
As a reminder, the exact same settings without the high pass on the Side-Chain sound like this.
Listen to the bass drum. With.
With the high pass on the Side-Chain, you get your bass drum back.
There are a lot of possible applications to this.
Imagine. Say you listen to your Side-Chain.
You find the 'S' that bothers you, and you boost that 'S'.
That means that the compressor is gonna react to the 'S' more.
That's called a De-esser! So you could turn your compressor into a De-esser.
Lots of possible applications! Ok, if you insist, we can do one more thing.
It is the most overused Side-Chain trick in the history of the universe: the external Side-Chain trick! Let me show you.
Here's a very pleasant signal.
How's your brain? How's it now? Alright, enough! I have a compressor, you can see it there.
I'm going to turn the Side-Chain on.
In this particular case, when you turn this on, it's actually external Side-Chain.
Every compressor is different, you've got to learn it.
And here in Pro Tools, that is the source of the external Side-Chain.
And here, I'm gonna set up the Send on this track, pre-fader, to send whatever is happening on this track, through this Aux, through this bus, into the Side-Chain of this compressor. If I press Play right now...
Nothing happens! Why does nothing happen? Because if I remove the external Side-Chain key right here...
The signal gets compressed.
If I turn the external Side-Chain key here...
The signal does not get compressed, because right now, the compressor is waiting for someone to tell it what to do.
But there's nothing going on on this track, so let's make something going on on this track.
Let's pick an instrument, like Boom.
And I'm gonna keep the compressor up so you can see what's going on in both. So now...
That sounds like a root canal without anesthesia. But! If I solo the bass drum here, and maybe go half time...
Right? And I press Play...
And I turn this on...
The easiest way to really hear what's going on is for me to turn the direct signal of the bass drum off, so we don't hear it, and to only hear what it does to the compressor's Side-Chain, hence to the output of that sine wave signal.
Shall I say that again? Ok! The easiest way for you to understand what's going on is to no longer listen to the bass drum as a sound but think of it just as the key, or the Side-Chain, to the compressor that is on the sine wave, and to realize what has been happening in pop music for the last 5 years.
Check it out! So yes, you can use it to animate a pad, so instead of the pad sitting there like a dead fish, you can animate it with whatever, your bass drum track, your snare drum track, your kazoo track, whatever you want.
You can also go and invent a new way.
So it's time for me to wrap this up so you can go play with Side-Chains, internal, external, and Peak and RMS, and all those things, to invent a new way of using a compressor, so you can beat the competition to the punch! Et voilà!
Once logged in, you will be able to read all the transcripts jump around in the video.
Fab Dupont is a Grammy winning NYC based record producer, mixing/mastering engineer and co-founder of pureMix.net.
Fab has been playing, writing, producing and mixing music both live and in studios all over the world. He's worked in cities like Paris, Boston, Brussels, Stockholm, London and New York just to name a few.
He has his own studio called FLUX Studios in the East Village of New York City.
Fab has been nominated for Grammys 6 times, including two Latin Grammys and has received many other accolades around the world, including Victoires de la Musique, South African Music awards, Pan African Music Awards and US independent music awards.
Toots And The Maytals
Parts of this site and some files are only accessible to pureMix Pro Members or available to purchase. Please see below our membership plans or add this video to your shopping cart.