Compression Masterclass: Every Technique Producers Need to Know

The Four Parameters: Threshold, Ratio, Attack, Release

Threshold determines when compression begins — signals above this level get compressed, signals below pass through unchanged. Lower threshold = more compression applied. Start by setting the threshold until you see 3–6 dB of gain reduction on peaks.

Ratio determines how much the signal is reduced. 2:1 is gentle (for every 2 dB above threshold, output increases by 1 dB). 4:1 is moderate. 8:1 is heavy. 10:1+ is limiting. Use 2:1–3:1 for vocals and acoustic instruments. Use 4:1–6:1 for drums and bass. Use 8:1+ for aggressive dynamic control or effect compression.

Attack determines how quickly the compressor responds. Fast attack (0.1–1 ms) catches transients and reduces punch. Slow attack (10–30 ms) lets transients through and compresses the sustain — this preserves punch while controlling level. For drums, slow attack preserves the hit. For vocals, fast attack controls plosives and breaths.

Release determines how quickly the compressor stops compressing after the signal drops below threshold. Fast release (20–50 ms) causes pumping and breathing effects (used in EDM intentionally). Slow release (200–500 ms) provides smooth, transparent compression. The release should be timed to the tempo — too fast causes pumping, too slow causes the compressor to never reset between hits.

Compressor Types: VCA, FET, Optical, Tube

VCA (Voltage Controlled Amplifier): The most transparent and versatile type. Fast, precise, and clean. SSL G-Series bus compressor is the most famous VCA design. Use for: bus compression, drums, anything that needs precise dynamic control without coloring the sound. Plugins: Waves SSL G-Master, Plugin Alliance bx_townhouse.

FET (Field Effect Transistor): Fast, aggressive, and colored. The UREI 1176 is the iconic FET compressor. Known for its ability to add energy, presence, and 'forward' quality to sounds. The classic 'all buttons in' mode (20:1 ratio) on vocals and drums is legendary. Use for: vocals, drums, bass, anything that needs energy and aggression. Plugins: Waves CLA-76, UAD 1176.

Optical: Slow, smooth, and natural. The LA-2A is the iconic optical compressor. Uses a light-dependent resistor (LDR) for gain reduction, which creates a gentle, program-dependent response. Use for: vocals, bass, pads, anything that needs transparent level control without audible compression artifacts. Plugins: Waves CLA-2A, UAD LA-2A.

Tube / Vari-Mu: The slowest, warmest, most colored type. The Fairchild 670 and Manley Vari-Mu are the icons. Tube compressors add harmonic saturation and warmth while compressing. Use for: mastering, bus compression, anything that needs warmth and glue. Plugins: Waves Puig-Child, UAD Fairchild 670.

Parallel Compression (New York Compression)

Parallel compression is blending a heavily compressed version of a signal with the dry, uncompressed signal. The compressed version adds density, sustain, and perceived loudness. The dry version preserves dynamics and transients. Together, you get the best of both worlds: loud AND dynamic.

Setup: send the signal to an aux channel with a compressor. Crush it — threshold at -30 dB, ratio 10:1 or higher, fast attack, medium release. Blend the crushed signal underneath the dry signal at -10 to -6 dB. The result should sound like the original signal but with more weight, density, and presence. If you can hear the compression working, the blend is too loud.

Parallel compression works on everything: drums, vocals, bass, synths, even the master bus. It is the secret weapon behind 90% of commercially loud mixes. The reason it works: in traditional compression, increasing the ratio and decreasing the threshold always costs you dynamics and transient impact. Parallel compression avoids this trade-off entirely.

Sidechain Compression

Sidechain compression uses one signal to control the compression of another. The most common use: sidechaining the bass to the kick, so the bass ducks every time the kick hits. This creates a clean frequency pocket for the kick without reducing the bass level permanently.

Beyond kick-bass: sidechain pads and synths to the vocal so they duck slightly when the vocal is present, creating space for vocal clarity. Sidechain reverb tails to the dry signal so the reverb swells between phrases but does not compete with the direct sound. Sidechain everything to the kick for the 'pumping' EDM effect (French house, future house).

For the pumping effect: use a compressor with a very fast attack (0.1 ms), slow release (200–400 ms), and heavy ratio (8:1+). The signal will pump up and down with each kick hit. For transparent ducking: use a fast attack but a fast release (30–80 ms) with a moderate ratio (3:1–4:1). The signal ducks briefly on the kick transient and recovers quickly.

Multiband Compression

Multiband compression splits the signal into frequency bands and compresses each independently. This allows you to control the dynamics of the bass, mids, and highs separately — essential for mastering and bus processing where a full-band compressor would either over-compress the bass or under-compress the mids.

Typical mastering multiband setup: Band 1 (20–120 Hz) ratio 3:1, slow attack, fast release — controls sub bass peaks. Band 2 (120–1 kHz) ratio 2:1, medium attack and release — controls mid-range dynamics. Band 3 (1–8 kHz) ratio 2:1, fast attack, medium release — controls harshness peaks. Band 4 (8–20 kHz) ratio 2:1, fast attack, fast release — controls sibilance and cymbal peaks.

Warning: multiband compression can cause phase issues between bands and create an unnatural, 'squashed' sound if overused. Use gain reduction of no more than 3–4 dB per band. If you need more control, fix problems at the source (individual track EQ and compression) rather than trying to fix everything with multiband compression on the master.

Bus Compression and Glue

Bus compression is applying gentle compression to a group bus (drum bus, synth bus, vocal bus, or master bus) to 'glue' the elements together. The compression responds to the combined signal, making the elements interact dynamically — when one gets louder, everything else ducks slightly, creating a cohesive, unified sound.

Classic settings for the SSL-style bus compressor: ratio 2:1 or 4:1, attack 10–30 ms (let transients through), auto-release or 100–300 ms, threshold set for 1–3 dB of gain reduction on peaks. This is the sound of virtually every professional mix since the 1980s. The compression should be felt, not heard — if you can hear it pumping, the threshold is too low.

Apply bus compression early in the mixing process, not just at the end. Mix into the bus compressor — set it up first, then build your mix. This way, your balance decisions are made in the context of how the compressor responds to the combined signal. Applying bus compression after the mix is finished often requires rebalancing everything.