Slate Digital VMR compression modules

compression tutorial: beginner

hey, hey, hey, look who's back! so good to see ya.

Compression is arguably the most important tool in the mixing engineer’s kit.

It is THE go-to tool that engineers reach for when they want characteristics like “punchiness,” “up-front-ness,” “smoothness,” “evenness” and “loudness.” Unfortunately, slightly overdo compression and you start to traverse the land of “squashed,” “unpleasing,” “overbearing,” even downright unlistenable. Sounds like compression is worth a discussion, no?

To make matters worse, not only can you ruin a good production with too much compression, it is BY FAR the most difficult effect to hear when you aren't used to listening for it. I still have problems listening for it after almost eight years of mixing audio! The combination of "hard-to-hear" and "can-ruin-a-mix-easily" characteristics of compression can truly create a frustrating situation for mixing and producing newcomers.

Although it is a tricky tool, I don't want the difficulty to scare you. Compression can be really powerful even if you don't quite get it yet. After my explanation here, I will again share a video link, and we can work on compressing a track together. I will also give some very useful beginner's tips at the end of this blog.

So, what is compression? Well, I like to think of it as a volume fader.

Now wait.

I also said that equalization (EQ) was like a volume fader. What’s going on here? I can explain. While eq acts as a precise fader, compression (in its simplest form) simply functions as an automated fader. Instead of focusing on equalizing frequencies, a compressor, well.. “Compresses” the difference between the loud bits and the soft bits of your audio. I like to think of a friendly little robot turning the fader down when the volume jumps too loud.

Okay, compression. But first, some house cleaning.

As a newcomer to audio, you may not be aware that sound maxes out at “0 dB.” That means that the loudest that any audio can get in any system is.. yeah, “0 dB.” Anything above "0 dB" is called clipping. clipping basically sounds awful. Trust me. On the other end, the quietest audio can get is "negative infinity dB". That's obviously too quiet. So when we record, produce, and mix, we generally aim for around the -10 dB to -5 dB mark. That way our sound is loud enough to listen to, and loud enough for our computers' algorithms to process correctly, without overloading the system (fig. 1)

Fig. 1: good volume, not enough volume, and clipping

fig. 2: transients and decays from a kick drum sample. Nathan can't spell "beginning."

A “transient” is a large spike in audio with a short decay. The perceived loudness of a transient is much less than that of a held-out note, and a compressor will react differently to a transient than a long note.

A transient, specifically, is the large spike at the beginning of any audio region, decay describes the amount of time it takes to reach silence after the initial transient (fig. 2).

Also notice, I don't know how to spell "beginning." Pro audio engineer here. 👍🏻

Okay, house cleaning done.

Let’s go over the controls that a compressor may have. Some compressors have more controls, some have less. That does not make any compressor inferior to another. I'm going to cover the six most frequent controls your compressor may have.

My DAW of choice is Logic Pro X. I’m going to use the included compressor. If you own Logic, perfect. If you don't, I guarentee your compressor will function similarly. Compressors will fall under the “dynamics” page of your plugins picker.

I want to break compressor parameters down into three main groups:

**(As a Logic Pro user, you can change the type of compressor you’re using by clicking through the different buttons near the top of the interface. Notice the amount and type of controls change depending on the compressor that you select. We will stick with the Platinum Digital model for our discussion since it is the default compressor when you load the software, and because it contains all six parameters.)**

Carousel imageCarousel imageCarousel imageCarousel imageCarousel imageCarousel imageCarousel image

different compressors that you can use in Logic Pro X

parameter group 1: makes the compressor work

These controls work in tandem, and will ultimately get your compressor to start doing things. I suggest starting by playing around with these tools.

fig. 3: use for reference with each parameter group

threshold (orange)

When you set your threshold, you’re telling the compressor the volume level that you want it to “guard.” Anything underneath that volume level is totally fine--it won't get tampered with. Anything above starts to get compressed. If I set my compressor’s threshold to -5 dB, but my audio is peaking at -7 dB, the compressor will do absolutely nothing. If my threshold is at -80 dB, then the compressor will lconstantly engage. However, before it starts compressing, you must change the..

ratio (red)

The ratio is the amount of volume that the compressor automatically reduces after audio raises above the threshold. Simply put, if the compressor has a 2:1 ratio, then for every 2 dB that the audio climbs over the threshold, only 1 dB will be let through. So a 2:1 ratio cuts the audio above the threshold by half. A 4:1 ratio quarters the audio over threshold, and a 100:1 ratio just flat out smooshes it. Many compressors start out on a 1:1 ratio, which is not helpful.. because (as you may be able to figure) compressors do nothing in this mode.

Carousel imageCarousel image

fig. 4: the squiggly white line on the graph reprisents the amount that the compressor is turning the volume down (this is called "gain reduction")

parameter group 2: makes the compressor work faster and slower

These controls adjust the speed that the compressor reacts to volume over the threshold:

attack (green)

the faster the attack, the faster the little robot buddy (remember him?) reduces the volume of audio once it goes past the threshold. You might think that we always want the fastest attack, but this is not the case. We often want a slower attack to let a little bit of the original volume through. This increases punchiness, believe it or not.

release (yellow)

When the volume goes underneath the threshold again, the compressor must relax to its default position. The release dictates how quickly our little robot buddy resets the volume back to its normal place. A quicker release can make audio sound really up-close, while a slower release can make your track sound distant and quiet. Both are desireable effects depending on the circumstances.

Carousel imageCarousel image

fig. 5: the squiggly white line on the graph reprisents the amount that the compressor is turning the volume down (this is called "gain reduction")

parameter group 3: miscellaneous controls

These tools don't really belong in the above groups in my opinion. But they are still important.

knee (blue)

Think of a human knee. When your knee is hard, like... in real life... I hope.. there is a specific point that the bend takes place. If you’re an Adventure Time character (if you don't know what I'm talking about.. where do you live?), you have a soft knee. The knee sits in the area of the threshold, and provides a logarithmic average of the threshold. The softer the knee, the gentler the slope of compression. The hardest knee begins to compress right at the threshold, and compresses its maximum amount.

Makeup gain (purple, fig. 3)

Since the purpose and main goal of compression is to turn the volume of audio peaks down, ideally you would then adjust the volume so that the percieved loudest peak of uncompressed audio is roughly equivelant to the average level of the compressed audio. That's why we call it makeup gain: when you turn the volume down with compression, you want to make up for that lost volume with extra output gain.

Carousel imageCarousel image

fig. 6: soft and hard knees

So now you know all about compression. Now the trick is applying your newfound knowlege in a mix! Let’s go back to our EQ tutorial mix and see how I would use compression to make this mix pop.

Here's a free multitrack of a little beat I produced: Click here to get your free stems. Download these stems, import them into your DAW, and follow along with me on YouTube while I dial the compression on each track and explain my thought process. (YouTube video not yet available)

Hopefully this video helped you. Please type in the comments below if you have any questions, comments, or concerns! If you are struggling with an area of compression, I would love to help.

Have an awesome day, friend.

  • Nathan

Oh! I almost forgot. Here's a list of super helpful tips to get you started:

  • Don't let your compressor show more than ~5dB of gain reduction.

  • If you're hitting 5dB of gain reduction on one compressor, instead of compressing more with that same compressor, load up a completely new instance of compression and chain the two together. Like a compression conga line! Seriously. This works.

  • Especially use multiple layers of compression with vocals. I use at least three of four stages of compression on my vocals. Seriously.

  • To control a melodic instrument with lots of transients (like an acoustic guitar that someone is strumming away at) use a fast attack and a medium-slow release.

  • To control a melodic instrument with no transients (like a bowed-cello or violin,) try a medium attack and a medium release with a low threshold and a low ratio.

  • To get a snare drum to CRACK! , copy your snare drum to a seperate track, use a very slow attack and a very slow release. Blend in with the original snare drum track

  • Don't overcompress kick drums. Let those things breathe.

And most importantly...

  • YOU DO NOT NEED TO COMPRESS EVERY CHANNEL! Many channels of audio DO NOT NEED compression. Use your ears. Breathe. Everything is okay if you don't compress every channel.