Learn Music and Audio Production | iZotope Tips and Tutorials
11 Creative Sidechain Compression Techniques
My friend and fellow contributor Phillip Nichols has already covered some common applications for sidechain compression. In this article, we’re going to explore some esoteric, creative, and all-around fun uses for the practice.
Learn how to use sidechain compression in your mixes with the new oscilloscope view in Neutron.
In a few of these tips, I might ask you to copy a channel—say, the snare—edit it a little, and mute its output, so that you can’t hear it, but are still able trigger the sidechain input of a compressor. This would be a dummy track.
Another piece of lingo you’ll see thrown around here is the verb “to key;” it’s pretty much interchangeable with the word “sidechain.” You could say “sidechain the hi-hat to the snare,” or conversely, “key the hi-hat to the snare.” It would mean roughly the same thing.
Lastly, some of these might seem a little utilitarian on the surface. However, they will constitute creative solutions to utilitarian problems, and thus, I’m hopeful you’ll find them quite handy all the same.
Okay, enough disclaimers; let’s get going!
1. Revisit the classic kick/bass synth trick
This has quickly become the defining sound of mainstream EDM, and you probably know how it works, but it’s worth repeating, in case you don’t:
You take a synth—usually a bass—and put a compressor on it; then you trigger its sidechain input with the signal of the kick. The attack and release parameters are your most important controls here, as they dictate how the bass will suck down when the kick hits, and how it will rise back up again. Get granular, for this is a big part of creating the groove.
Of course, the trick isn’t limited to bass elements; any synth, vocal, or indeed the whole mix itself can be keyed to the kick. Just make sure you stay within the boundaries of good taste as defined by the track.
A fun design we made. Kids, get along!
2. Sidechain to a delayed signal
This was mentioned briefly in our EDM mixing article, but it is worth expounding upon: you can add swing, swagger, groove, pocket, or other rhythmic variations to the previous trick by setting up a dummy track and delaying it ever so slightly.
For instance, you could send a kick to an aux track, put a fully wet, 70-millisecond delay on it, mute its output, and route that signal into the sidechain of a synth. Essentially, this would add an “attack delay” parameter to your sidechain compressor—or, depending on the delay, could cause the compressor to clamp down at weird intervals. Imagine using a multi-tap delay to trigger the sidechain; that would be bizarre!
You could also use this effect in sidechaining hi-hats to snares, synth-pads to claps, or literally anything to anything else.
3. Sidechain to a ghost
This topic was also covered in the EDM article, and it’s also worth expounding upon, as you can get a lot of mileage out of keying to elements that remain inaudible in the final mix.
Say you have a measure-long loop with three simple parts: a kick hitting on the first beat; a snare hitting on the third; and a synth bass playing a long, sustained note throughout the measure. This, in and of itself, could be quite boring.
Sure, there’s a lot you could do to spice it up, even with conventional sidechain compression. Key the bass to the kick, and the bass will swell up towards the snare; key the bass to the snare, and it’ll creatively duck out in the middle of the measure.
These are fun tricks, but they are in the realm of the expected. The listener will intuitively understand how the elements are linked, which might work out great.
Or you could mess with the listener’s sense of balance, and sidechain the bass to a ghost! It doesn’t have to be complicated: set up another track, put your least CPU-intensive soft synth thereon, and hammer out a staccato triplet pattern. Now mute the output of that track, rendering it a dummy.
Finally, set up a compressor on the bass, and assign the sidechain input to the dummy track you’ve just created. The bass will bounce in response to the groove of the triplet, creating a rhythm previously unheard in the piece.
You can do this with all sorts of rhythms—the key is to experiment.
Click the screenshot below to enlarge it.
4. Sidechain elements to a fill
If you have a track where the drummer plays an amazing fill—a wonderfully timed, perfectly-executed lift into the pre-chorus/chorus—and yet, the rest of the band doesn’t gel, try this out:
Bounce the fill down to its own stereo channel; how you do this varies from DAW to DAW, but many support this feature. Mute the track’s output, so it becomes a dummy track. Now, put a gate on the guitar, keyboard, or otherwise lackluster offender, and assign the sidechain of the gate to the dummy track. This bit of production trickery can create a special, cohesive moment.
5. Sidechain a sine wave to your kick
You might get an acoustic kick drum that just isn’t beefy enough or an electronic kick that lacks a little oomph. If that’s the case, try this:
Set up a sine wave on a new track, and set it to hang somewhere between 20–60 Hz (if the song is relatively diatonic, try setting the sine wave to the key of the tune). Next, place a gate on the sine wave. Look now to the gate’s sidechain input—here you will assign your original, lackluster kick.
Once that’s done, tweak the attack, release, and threshold to taste. You should wind up with a low, almost subsonic thud occurring simultaneously with the kick. Blending it into the mix, this sine wave beef up the kick considerably.
If you find that the attack time on the gate gives you too much click when the gate opens—or that a slower attack results in sloppy timing between the kick and the sine wave—try this bonus tip (and here’s where we get creative):
Duplicate your original kick onto a dummy track, nudging it earlier in time ever-so-slightly. Use the resulting, early kick as the sidechain input. Now you can open the gate slower on the sine wave, avoiding clicking artifacts. When doing this, make sure you keep the ultimate goal in mind: a perfectly timed kick/sine-wave combination. It’s very easy to mess up the timing.
Click the screenshot below to enlarge it.
6. Vocal music ducking
This is an old radio trick and could fall more under the utilitarian category, but it does have its creative applications. Basically, you take a track or bus, put a compressor on it, and assign your lead vocal to the compressor’s sidechain input. That way, when the vocals come in, they bring the level of the other element down.
In high-doses this is often called “ducking.” In mix applications, you’d probably employ the trick subtly, using it to supplement (and in some cases obviate) riding the vocal fader.
Slap the sidechained compressor on the background vocals and guitars, for instance, and the lead will cut through that much more.
You can get creative with this application by assigning throwaway vocals to the sidechain input—something like a specific delay throw, or a vocal motif that gradually fades in from silence to an overwhelming state. Those are a couple of examples.
I like to get creative and utilitarian at the same time: I use this technique with a multiband compressor. Rather than duck the whole drum bus down in a noticeable way, I’ll analyze the vocal and find its “meat”—the frequency bands conveying the most information (usually there’s some meat in the low-mids and some more in the upper mids).
I’ll slap a multiband compressor on, say, a drum bus, and dial these corresponding bands in accordingly. I’ll choose subtle settings, and key the sidechain input of this compressor from the vocal. If I’m doing it right, the vocals will cut through, but the effect of compression will be far less noticeable on the drums.
So, if my vocal is communicating strongly between 300–600 Hz, as well as between 1–5 kHz, I’ll put a multiband compressor (or dynamic equalizer) on the drum bus, set up the two bands as dictated above, and key the compressor to the lead vocal. The drums will only duck down in those particular frequency-bands whenever the vocals are strong enough to trigger them.
7. Use a processed signal as the sidechain input
You can also use an overly EQ’d input for creative purposes, especially in multiband compression. Here’s an example that is both creative and utilitarian:
Say your vocal sports sibilance in a frequency range lower than you’d expect (4 kHz or so), such that using a de-esser on this frequency-band has negative effects (i.e., it ducks down pleasant information in addition to the sibilance). Here you could duplicate the vocal onto a dummy track, place a de-esser on it, set the de-esser to 10 kHz, and place that de-esser in “audition” or “listen” mode; ostensibly, the only signals passing through the dummy track would be the harsh esses.
Look to your original vocal and set up a multiband compressor/dynamic EQ in place of a de-esser. Make sure this compressor/dynamic EQ has a sidechain input, and assign the previously-tweaked, de-essed dummy track to the sidechain.
If you set up a band in the problematic sibilance range, but key it off your dummy track, you’ll get great results: the only time you’ll hear ducking is when the compressor is triggered by the overly sibilated sound.
Click the screenshot below to enlarge it.
8. Sidechain the input of a synthesizer
Sidechaining isn’t only for compression and gating; Many synths, such as VocalSynth from iZotope, sport sidechain inputs. In the case of VocalSynth, this can be used to set up a vocoder, talkbox, and more. Other synths can grab their envelope controls from the side chain input—or, if they’re modular in design, you can assign the sidechain input to literally anything.
Experiment with putting a rhythm element of your mix into the input of a synth’s sidechain, and you just might find yourself wandering down some interesting rabbit holes indeed.
Click the screenshot below to enlarge it.
9. Sidechain two elements to each other
I’ve done this exactly once, with a dueling, overlapping guitar part. But it worked! Basically, if you have two distorted guitars (or similarly fuzzy synths) that engage in a sort of call and response pattern—and yet, react/overlap in a manner that renders them indistinct in the mix—try this out:
On one guitar/synth, set the sidechain to the other’s input, and vice versa. Of course, you’ll need to toy with the attack and release constants, but if pulled off successfully, you’ll ensure that the entrance of each phrase is distinguished every time, as it’ll immediately push the other element down into the mix.
10. Sidechain the room mics to the snare
If the drummer isn’t grooving, there are tricks you can employ besides basic editing. Often sidechaining elements to the snare helps; the hi-hat is a worthy candidate for this implementation.
In situations where I want to get that super-saturated, Black Keys rock’n’roll vibe, I may go a bit further, even if the drum part is already grooving—I’ll take the room mics, slap a compressor on them, and key them to the snare.
As with all these tricks, attack and release adjustments are of the utmost importance, but successfully accomplished, this trick works wonders. You’re basically taking the atmosphere of the drum kit and making it pump, suck, and breathe in response to the snare.
To do this right, you might need to duplicate the snare onto it’s on track and gate it to get rid of the ghost notes—you only want the strong backbeat hits to trigger this pumping. You may also want to edit out any fills from this track, and of course, silence the track’s output, creating a dummy track.
11. Mix triggered compression for printing stems
Printing stems for mastering engineers, production companies, labels, and music supervisors can be par for the course in this business.
This might seem impossible if you have a complicated stereo-bus chain—and sometimes, it is. But if your processing consists of a compressor and not much else, you can still print every stem so that each one behaves roughly the same as it did in the whole mix, and the stems sum together nicely upon re-aggregation. How, you may ask? Sidechaining!
Simply mix as usual, into your bus compressor, until you like the results. Once you’re done, bypass your stereo-bus compressor and print the mix without compression. Set up a stereo dummy track, move the printed mix there, and now, take a look at your original bus compressor:
If it has sidechain inputs, you’re in luck! Just route the dummy track to the sidechain input and print your stems as you normally would. The sidechain will react to the timing of the original mix, rather than the stems, so that each group-track will be processed similarly to how they would in full-mix context.
There are some caveats to keep in mind though: any processing you put on your master bus after the compressor will either not work as it did before, or will be nullified entirely; basically, this trick works best if your bus compressor is the last processor on your mix.
Also, pay attention to how the sidechain detector works. If you’re using the emulation of a vintage unit, there’s a good chance its sidechain detector operates in mono only; if that’s the case, your compressor will react differently from how it would’ve originally. The results, in other words, will not add up. Read the product manual or do research on the original unit, and you’ll know for sure if you can use this track.
Like many of these articles, the tricks offered herein represent a small fraction of the tricks you might come up with alone on a Sunday evening, up against a deadline. It is my sincere hope that they inspire you to get started down the wonderful rabbit hole of sidechain compression.
As always, experiment to your heart’s content—but always keep your eye on serving the final mix, which ultimately serves the song’s concept and execution. That, my friend, is the key.