What is Multiband Compression – How to Use It in Your Mix

Multiband compression is a tool in mixing which is used to provide more control over your audio than you get in a conventional compressor. Let’s identify what a multiband compression is and how to use it in your mix.

What is Multiband Compression

First let’s define multiband compression.

As the name suggests, this is a compression tool with one notable difference over conventional compressors: it separates your audio into different frequency bands of your choosing.

Each frequency range can be compressed to different degrees with unique settings with some being untouched altogether if you choose.

Why Use Multiband Compression

Multiband compression is useful when you have specific frequencies you need attenuated rather than the entire track.

A prime example is in de-essing, the process of attenuating the harshness of a vocal due to sibilance.

In actuality a de-esser is nothing but a multiband compressor which specifically targets the 6-10k (give or take) frequency range where sibilance “S”, “T” and other sounds act up.

We can create our own makeshift de-esser with a multiband compressor. In this example, I’m using FabFilter’s Pro-MB multiband compressor:

As you can see, it’s a rather simple use of the multiband compressor. We’re simply creating one band at 7k which ranges up to 12k and down to 5k.

Find the worst offender of sibilance in your vocal track and use that to set your threshold accordingly, generally attenuating this peak by roughly 5dB with the rest of the settings.

Combining an instant attack and release along with an aggressive ratio of 8:1 and a hard knee, we’re immediately and exclusively targeting those sibilant peaks on the high end of our vocal.

With the threshold we set, the compression only engages when those sibilant peaks are playing, so otherwise the high end of the vocal is untouched. And because it’s multiband compression, everything below 5k is untouched.

We’d likely put this de-esser early on in our vocal chain, then further down we can add a conventional compressor to get a more even and energized vocal (see my tips on vocal compression).

It’s a transparent way to clean up the harshness caused by sibilance without affecting the rest of the track. Check out my tips on how to further deal with a sibilant vocal, by the way.

There are plenty of other applications for multiband compression. As I talked about in my comparison of multiband compression versus dynamic EQ, multiband compression is oftentimes better suited for macro level adjustments in your mix.

I like to use a multiband compressor on my drum bus, mix bus, and it’s arguably most commonly used in the audio/song mastering stage.

Rather than making a static cut at 500Hz to add clarity to a master, a mastering engineer might create a band via multiband compression in this range to pull it down if and when the mix mud stacks up and needs to be controlled.

Dynamic changes like this keep your mix evolving to the listener, even when they don’t know exactly what about your song is changing as they’re listening. It’s all about keeping the listener engaged, and multiband compression is a useful tool for doing just that (or even just controlling the sibilance in a vocal track).