Why Compress
Three reasons to compress, in order of importance:
-
Control dynamics. A vocal that whispers one line and screams the next is hard to sit in a mix. Compression narrows that dynamic range so the quiet parts are audible and the loud parts don’t blow past everything else.
-
Shape tone. A fast attack on a compressor clamps down on the transient of a drum hit, making it rounder and pushing it back in the mix. A slow attack lets the transient through and compresses the body, making the hit punchier and more forward. Compression changes the character of a sound, not just its level.
-
Glue. Bus compression on a drum submix or a mix bus makes the elements feel like they belong together. The shared compression creates a unified dynamic envelope — the whole bus breathes as one thing.
The Parameters
Every compressor has the same core controls, even if the labels vary:
Threshold: The level above which compression begins. Everything below the threshold passes through unaffected. Lower the threshold and more of the signal gets compressed.
Ratio: How much compression is applied above the threshold. 2:1 means for every 2 dB the signal goes above the threshold, only 1 dB comes through. 4:1 is moderate. 10:1 is heavy. Infinity:1 is a limiter — nothing goes above the threshold. Watch: Visualize Compressor Ratio
Attack: How quickly the compressor engages once the signal crosses the threshold. Fast attack (0.1–1ms) catches the transient. Slow attack (10–30ms) lets the transient through and compresses what follows. This is where the tone shaping happens.
Release: How quickly the compressor stops compressing after the signal drops below the threshold. Fast release sounds transparent but can cause pumping. Slow release is smoother but can squash the tail of a sound.
Makeup gain: Compression reduces level. Makeup gain brings it back up to match the uncompressed signal. Always A/B with and without the compressor at matched loudness — if the compressed signal is louder, you’re hearing volume, not improvement.
Knee: The knee controls how the compressor transitions into compression at the threshold. A hard knee means compression kicks in abruptly — below the threshold, nothing happens; above it, the full ratio applies instantly. A soft knee eases into it, applying gradually increasing compression as the signal approaches and crosses the threshold. If your compressor offers a knee control, it’s offering you a compromise between precision and transparency. Watch: Why Does a Compressor Have a Knee
Types of Compressors
Different compressor designs have different sonic character because of how their gain reduction circuits respond:
VCA (Voltage Controlled Amplifier): Fast, precise, transparent. The SSL bus compressor is the archetype. Good for bus compression and anything where you want control without obvious coloration.
Optical: Smooth, slow-responding, musical. The LA-2A is the classic. The optical circuit creates a naturally soft knee and program-dependent release. Excellent on vocals and bass.
FET (Field Effect Transistor): Fast, aggressive, colored. The 1176 is the standard. At extreme settings (all buttons in), it adds harmonic distortion and excitement. Great on drums, vocals that need edge, and parallel compression.
Variable-Mu (Tube): Warm, thick, with natural saturation. The Fairchild 670 is the legend. Slow attack character makes it a natural choice for bus compression and mastering.
You don’t need all four types. But understanding how they differ helps you choose the right tool. An 1176 on a delicate acoustic guitar is the wrong compressor. An LA-2A on aggressive drums is too slow.
Compression Strategies
Serial Compression
Vocabulary
Serial Compression
Using two or more compressors in sequence, each doing moderate work rather than one doing heavy lifting. Common approach: a fast compressor first (catching peaks, 3-4 dB), followed by a slow compressor (evening dynamics, 2-3 dB). The result sounds natural because neither compressor is being pushed hard.
Instead of one compressor doing all the work, use two (or three) in series, each doing a little. The first catches the big peaks with a faster attack. The second smooths the overall dynamics with a gentler ratio. Each compressor does less work, which sounds more natural than one compressor working hard.
A common vocal chain: an 1176-style compressor first (fast, catching peaks, 3–4 dB of reduction), followed by an LA-2A-style compressor (slower, evening out the overall performance, 2–3 dB of reduction). Neither compressor is working hard. Together, they create a vocal that sits in the mix without sounding compressed.
You can also use two of the same type in series — a fast VCA followed by a slower VCA, for example. The principle is the same: distribute the work.
Parallel Compression
Blend a heavily compressed copy of a signal with the uncompressed original. The compressed copy brings up the quiet details — room ambience, sustain tails, subtle performance nuances — while the original preserves the natural dynamics and transients.
Set up a parallel compression bus: send the signal to an aux, compress it hard (high ratio, fast attack, fast release, heavy gain reduction), then blend the compressed return underneath the original. Adjust the return fader to taste. A little goes a long way.
Parallel compression is especially effective on drums. It adds weight and sustain without killing the transient snap. Watch: Parallel Processing — Have Your Cake and Eat It Too
Bus Compression
A compressor on a submix bus (drum bus, vocal bus, mix bus) creates glue. The elements react to each other through the shared compressor — when the kick hits, the whole drum bus dips slightly, and the recovery creates a breathing feel.
Settings are typically gentle: low ratio (2:1–4:1), medium attack (10–30ms to let transients through), auto or medium release, 1–3 dB of gain reduction. The compressor should be barely working. If you can hear it compressing, it’s probably too much.
A word of caution about mixing into a bus compressor on the master: personally, I keep it bypassed most of the time and periodically check it to hear what’s going to happen when I engage it. If you leave it on while you’re building the mix, you can end up pushing level into it without hearing the consequences — it pushes back at you, and your mix decisions get filtered through its reaction. That’s fine if it’s intentional, but it can sneak up on you. Think of bus compression as more of a finishing-stage decision.
Limiting vs. Compression
A limiter is just a compressor with a very high ratio (typically infinity:1) and a very fast attack. Nothing goes above the threshold. Use limiters for:
- Catching stray peaks that would clip the output
- Mastering (loudness maximization)
- Bus limiting to keep submixes under control before they hit the master
Don’t use a limiter where you need musical dynamics control — it’s too heavy-handed. That’s what compression is for.
Hearing Compression
The hardest part of compression is hearing it. Beginners often compress without knowing what changed. Here’s how to train your ear:
- Bypass constantly. A/B the compressor on and off at matched gain. If you can’t hear the difference, either the compression is too subtle or you’re not listening for the right things.
- Exaggerate first. Set a high ratio (10:1), low threshold, fast attack. Hear what aggressive compression sounds like on this material. Then back off until it sounds right.
- Listen for the release. The release is where compression becomes audible. When the compressor lets go, the signal jumps back up. A pumping, breathing sound means the release is too fast or the compression is too heavy.
- Listen to what happens between the notes. Compression brings up the quiet stuff — room noise, string buzz, breath, reverb tails. That’s the tradeoff. Watch: Trouble Hearing Compression
Gain Staging Through Compressors
If your compressor is reducing 6 dB of gain, add 6 dB of makeup gain to match the output to the input level. Then A/B the compressor — bypass it and compare. If it sounds better bypassed, either the settings are wrong or the track doesn’t need compression.
Not everything needs compression. A well-recorded acoustic guitar with consistent dynamics might be perfect with just a fader ride. Compressing it “because that’s what you do” will make it smaller and less alive. Listen. Decide. Don’t default.
What to Practice
- Compress a vocal track using serial compression: an 1176-style first, an LA-2A-style second. Aim for 3 dB of reduction on each. A/B against the uncompressed vocal at matched loudness.
- Set up parallel compression on a drum bus. Crush the parallel copy (10:1, fast attack, fast release, -15 dB of reduction). Blend the return underneath until you hear the room ambience and sustain come up. Note where it transitions from “enhanced” to “obviously compressed.”
- Put a bus compressor on your mix bus. Set it to 2:1, 30ms attack, auto release, and aim for 1–2 dB of reduction. Listen to how the mix breathes differently. Does it feel more cohesive? Does anything get lost?
- Mix a full song without any compression. Just faders, pans, and EQ. Notice what problems remain that compression would solve — and which ones it wouldn’t.
