5 Tips for Compression in Mastering


Perhaps you've read the previous article on equalization in the mastering process and find yourself a glutton for punishment. If not, welcome anyway! Please, feel free to read these articles in any order—you ostensibly have free will.

We're here to talk about compression, and what it can do for you in the mastering process. This article is going to assume a basic knowledge of dynamic compression on your part. If you're not familiar with what compression is, or how it works, then I recommend brushing up on these videos on various aspects of compression—parallel, sidechain, and series.

Still, we can sum up compression thus: Compression is the practice of decreasing an audio signal's dynamic range. The reasons for employing compression are varied, and include the desire to smoothen an inconsistent performance; the desire to bolster the "groove" or the rhythmic "feel;" the desire to raise the entire level of an audio signal (more about how this works later); or the desire for "vibe" and "character" specific compressors provide.

In mastering, each facet of compression comes into play. Here's how.

Compression for Inconsistency in Song Sections

If there are big disparities between the level of a verse and a chorus, your first line of defense is not a compressor plug-in, but gain automation that flows into further downstream processing. You can take that quiet verse and raise its level using a gain plug-in (or the input of a compressor plug-in) before it even hits an effect. This is arguably the most invisible kind of compression, yielding no sonic artifacts and accomplishing the goal of evening out sections of a song.

This technique doesn't just apply to whole sections: If you have a single, spiky transient that dominates a second of the mix—say a wayward snare or a loud vocal syllable—you can get away with pulling down that second of material with automation, so long as you don't pull it down too far (a dB will do). Also, it won't work over a prolonged period (we're talking a second here, rarely more, often less).

Now, on to compression plug-ins: In this use case, parallel compression might be the trick. Parallel compression involves duplicating the mix on a second track, slapping a compressor on that duplicate track, making sure your DAW's latency compensation allows for phase-coherent playback, and blending the outputs of these tracks to taste. Usually, the parallel-compressed track is affected rather heavily (with larger ratios, lower thresholds, and attack/release to taste), but blended low in the mix. Thus, when the mix gets super loud, the compressed track falls to lower levels and can be comparably inaudible; however, when the mix subsides to lower levels, the compressed track rises in gain, and yields an overall louder mix, as the blend of the two signals is proportionally louder. You can effectively squash the dynamic range of a mix this way without audible compression effects.

However, if you can't get the attack and release to sit right on a tune using parallel compression—if you always hear the unmusical pump of the compressed track returning to full volume—here's a more drastic approach.

On the stereo track, lower the threshold of a compressor to ensure the mix is ­always being compressed. It'll compress more on loud sections and lighten up for softer ones. This technique, though heavy-handed, can be of service, provided you're using the right compressor for the job (usually something relatively transparent), and provided the mix is relatively sparse. Don't use this technique on a very busy song, one with a lot of moving parts. It will clutter, and possibly distort, the goings-on. You wouldn't use this technique 45 out of 50 times, but for those five times, it's invaluable.

FabFilter Pro-C 2 - Compressor Plug-In

Compression to Help the Groove

Here, the technique you'd use is similar to bus compression in the mixing phase. Basically, you use the attack and release of the compressor to shape the feel of the music finely. You can do this by tuning the compressor to respond chiefly to the kick and snare. Here's the step-by-step guide.

  1. Set the ratio relatively low, say 2:1 or 1.5:1
  2. Set the threshold low for now, so that you really hear the compression
  3. Set the attack and the release to clamp down on the kick and snare hits
  4. Listen to the high-hats (or other busy parts—synth arps, guitar runs, what have you) while tailoring the attack, looking for the right "bounce" as the compressor returns to its resting place between snare/kick hits (hint: if it's got you dancing, you're on the right track)
  5. Bring the threshold higher until you're hearing the least amount of compression that will suffice
  6. Bypass/Engage the compressor to judge the difference
  7. Nod with satisfaction

Compression to Raise Overall level

Because compression restricts the dynamic range, compression can help you raise the overall level.

How? Here you must think of the digital ceiling—the absolute maximum level a signal can achieve before yielding distortion and nasty artifacts. A wild, errant snare hit might hit that digital ceiling way before the rest of the mix. However, if you restrict that peaky snare with a compressor, you have more headroom to push the entire mix louder as a whole.

To help you visualize this concept, imagine your closet. Say you have a bunch of winter jackets you'd like to put away for the season, which means piling them on a shelf by the closet's ceiling. You can't hang the jackets there; because they're too long, they won't fit—they'll spill down into the rest of your closet space. What do you do? You fold the jackets so they take less room, and now you can store them in a higher place.

Compressing the dynamic range allows you to fold the mix and place it higher, toward the digital ceiling.

Now, even though you can compress signal to make it louder, I'd say you should still avoid drastic measures in compression. The first rule of mastering is to do no harm to the mix, which means the moves are subtle. Not only that, but many digital limiters these days are quite adept at helping you achieve loudness while preserving transient heft. Thus, I'd reckon you're better off automating for unevenness, employing parallel compression, compressing for groove, or combining some measure of all three before hitting the limiter.

FabFilter Pro-L 2 Limiter Software Plug-In

Compression for Character

This is the last method we'll cover, and it involves using hardware or hardware-emulating plug-ins with tons of "vibe". All the previous facets apply—the dynamic-range squashing, the groove shaping, the loudness fostering—but the stand-out here is that the compressor has an obvious character that changes the quality of the mix.

Many compressors are engineered to be transparent; in this technique, you don't use those compressors. You use units like the API 2500, the Manley Variable MU, and others (also, don't discount the software equivalents—they're pretty good these days).

Manley Labs Stereo Variable MU Limiter Compressor Mastering Version

But doesn't this violate the first rule of mastering—first do no harm? Well, no: If the song is going for a Black Keys vibe, but lacks that grabby energy, it's well within your right to slap on a tube-powered compressor to secure that creaminess.

The decision is personal, but should be informed, so learn about the different kinds of compression, their characteristics, and which one is right for a given circumstance.

Compression in the Mastering Ecosystem

In the realm of mastering, all the processes are interactive; how you compress will influence EQ decisions and limiter behavior, so keep this in mind. For instance, if an equalizer is placed before the compressor, the frequencies emphasized by this equalizer will change how the compressor's triggered—and you can use this to your advantage or your detriment. Likewise, if the compressor precedes the EQ, the process can boast tonal-shaping properties, as it'll flatten the sound relative to which frequencies are the loudest ones (the ones triggering the compressor's overall swing downwards).

Therefore, whichever way you choose to go, I recommend that you master with all processes inline. We'll cover how best to set up your processes in later mastering-centered articles, but instantiate your equalizer, compressor, and limiter on from the get go; this will save you from having to re-tweak a whole bunch of settings once you switched an equalizer or limiter into the chain halfway through.


These are but a few of the considerations of compression in the mastering process. There's simply too much to cover for the space we have. However, these concepts are a great place to focus your initial attention when acclimating yourself to the discipline. If you read this article and find yourself hungry for more tips on compression, don't hesitate to reach out in the Comments section, and let us know!