1. Don’t Always Play By The “Rules”
People just learning how to use compression are often given guidelines as to the “best” settings for various applications—these are frequently recommendations as to what Ratios (and sometimes Thresholds) to use for certain tracks or instruments. For example, the standard textbook suggestion might be to stick to around a 4:1 ratio for vocals, 10:1 or so for bass, etc.; another common recommendation is to set Threshold so only the loudest notes result in gain reduction. The idea is to guide neophytes into not over- or under-applying the processing, and those are, in fact, good suggestions—but they should be taken only as starting points. Every player and every recording will be a little different when it comes to level variation and dynamic swings, so one track may work fine with a standard recommendation (like the ones above), while another may end up sounding too choked, or not controlled enough. Don’t be afraid to drastically alter Ratio and Threshold settings for an individual situation—the meters can help, but ultimately your ears will have the final say.
For example, while a medium Ratio (4:1-6:1) and a Threshold that hits just the loudest notes of a vocal track—a common approach—may work just fine, a much lower Threshold, that affects everything but the softest notes, coupled with a gentler Ratio (1.5:1–2:1) might provide a more subtle application of dynamic control. Aspects of the recording, like lip smacks and breath noises, will be affected differently with these two methods, and depending on the specific part, one may provide far better-sounding (smoother, more natural) results.
2. Don’t Limit Yourself To Level Control
From a purely technical standpoint, the goal of compression is to reduce dynamic range—this usually means taking a track that has wide level swings or occasional notes that jump out of the track, and providing more consistent dynamics. But compressors shouldn’t be thought of only as problem-solvers—they can be employed as effects too. Careful application of compression can affect the thickness and presence of a recording, even one that doesn’t really require dynamic control.
The most classic example of this is the way compression is used to shape the envelopes of drums. Specific settings of the Attack and Release controls can be utilized to change the balance between the attack portion of drum hits (or any percussive sounds, for that matter) and the decay portion of notes. For example, with a suitable Threshold and Ratio (6:1–8:1+), drums can be tightened up by applying a slower Attack (~15–20 ms or so) and a longer Release (~500+ ms). This will allow the initial impact transients to pass through un-compressed, while clamping down on the “boom” that follows, making for a tighter drum sound—sort of like adding damping to the drums after the fact. The opposite settings—a fast Attack (≤ 1–2 ms) and fast Release (< 50 ms)—will clamp down on the transient while allowing the rest of the note (the “boom”) to come back up, ringing out more strongly. This can take a tighter drum sound and give it a wide open, ringy character (like removing damping after the fact)—this is a common technique for making drums sound “bigger”, and some compressors are famous for it, which brings us to the next suggestion.
Audio example 1—Compression applied as an effect to drums: A) Original; B) Tighter drum sound (slower Attack/Release); C) “Bigger” drum sound (faster Attack/Release):
3. Don’t Just Stick To The Default
Most DAWs come with a standard Compressor, and novice engineers/mixers can be tempted to just use multiple instances of this included plug-in, especially once they get comfortable with it. But there are many different types of compressors out there (available as both software and hardware), often modeled on the very different circuits used in classic hardware compressors. Many of these have very different response characteristics—even identical settings will yield distinctly different results, and some are especially well suited to certain applications.
The most well-known of these different compressor circuit types are:
- Opto compressor: Opto is short for optical—an internal light and sensor are employed for level-detection and gain reduction. Opto compressors, like the Teletronix LA-2A, are smooth—signals can be slammed without really sounding compressed. They’re many engineers’ secret weapon for vocals (though they’re good on lots of other signals as well).
- FET compressor: An FET (Field Effect Transistor) provides lightning-fast response, making this design great for drums and percussion (among others), with the FET circuitry adding a bit of welcome edge when pushed hard. The classic design is the Urei 1176, which is famous for making drums sound “big” (as described above).
- VCA compressor: The Voltage-Controlled Amplifier is the standard modern compression circuit—it’s fast and tight, and compressors that use this design (like the dbx 160 and SSL Buss Compressor) are good at tightening up dynamics that are all over the place (think slap bass, drums, percussion—just about anything really).
These compression circuit types (along with others) are widely available as third-party plug-ins (and even built-ins). Instead of sticking with the familiar stock compressor, it’s well worth it to give them a try, especially on certain tracks—not only will many of these alternate compressor designs respond more “musically”, but thanks to simplified controls, some of them make it even easier to get the best results.
4. Don’t Be Afraid To Double-Up
While using a single compressor can often achieve the necessary results, it’s not uncommon for there to be multiple compressors used in series on some sources. The most obvious example of this is a multi-tracked drum kit—typically individual compressors (even different types) may be applied to the individual kick and snare, and sometimes overheads (which also contain those drums). All of these individual drum tracks are usually subgrouped through an Aux, and it’s possible for there to be another compressor added there, affecting the entire kit. The same thing might be done with background vocals, or other layered instruments—individual compressors to control each track, plus overall compression on a subgroup of the parts, to tighten up the blend.
If multiple compressors are employed, you’d probably want to go a little easier within each individual processor as far as gain reduction, so the overall effect isn’t too squashed. Sometimes, if you’re having trouble taming the dynamics in a difficult recording, feeding one compressor into another may help—one can be set with lower Ratio & Threshold, for a gentle squeeze, while another is used more as a limiter (higher Threshold/Ratio), for a tighter clamp-down on occasional dynamic jumps (loud notes) without overly squashing the track as a whole.
5. Don’t Get Stuck In-Line
The most common way to apply compression is as in-line processing—the compressor is inserted in the signal path, and the output is the 100%-compressed signal. But there’s an alternative to this—Parallel Compression (a.k.a. “New York Compression”). With this technique, a compressor is inserted either on a duplicate track, or in an Aux (with the signal from the original track routed via a Send to that Aux). This copy is then compressed—often more aggressively than usual—and then the compressed version is blended in with the original un-compressed track.
This approach will yield a slightly different effect than a standard in-line application—the compressed version can be really slammed, for a strong effect, but the original un-compressed version will provide any missing transient snap and dynamic accents that would otherwise have been sacrificed by heavy compression. Parallel (NY) Compression is especially popular on drums, and sometimes massed vocals, and is definitely worth experimenting with—some compressors make it even easier to apply by adding a Mix control (%), which blends the dry (un-compressed) and wet (compressed) versions of the track without the need for extra routing/auxes.
Audio example 2—Some drums with parallel compression (gradually) applied:
6. Don’t Squash Synths Too Much
This last one is more specific to the use of compression with synthesizer (and sometimes sampler) sounds. Acoustic instruments (and the best sampled emulations) can be compressed fairly heavily, controlling level (volume) variations without sacrificing musical accents and dynamics. That’s because when loud and soft notes are played on an acoustic instrument, it’s not just volume that varies to provide those critical dynamics and musical accents that give the performance its character. The tone and attack of loud and soft notes also varies, so even if strong compression strips away the volume variations of a musical line, the differences in attack and tone between loud & soft notes will still be there to preserve the pattern of accents/dynamic variations, even if the volume variations themselves have been smoothed out of existence.
But many synth sounds don’t have these variations—a bubbly synth sequence may depend entirely on level variations for its rhythmic/accent pattern, and heavily-applied compression could rob that line of its internal rhythm and much of its musicality. When applying compression to electronic sounds, you should be aware of this possibility—if it seems that a synth figure is reacting badly to being compressed, you could either back off the squash a little, or (better) add some velocity-based ADSR Attack and LPF Cutoff variation to the patch itself (via the synth’s programming parameters), to insure that any internal dynamics will survive even the heaviest application of compression.
Compression can be one of the more difficult types of signal processing to truly master, but while you’re getting a handle on it, it might help to keep two thoughts in mind. First, “do no harm”—make sure any compression applied doesn’t have a negative effect on the musicality of the track(s); And second, think outside the box—different compressor settings, types, and applications can often be the best way to get the most out of this very important aspect of production.