How to Compress Vocals

Good vocal compression makes the vocal consistent, so it sits forward at a steady level instead of jumping between quiet and loud words. The most reliable way to do it is serial compression: one compressor to catch the fast peaks, and a second, gentler one to level the overall body. Two compressors each doing a little sound far more natural than one doing a lot. Once you understand what the four main controls do, the settings make sense.

What do the controls actually do?

Every compressor comes down to four controls, and vocals need all four set with intent.

• Threshold sets the level where compression begins. Lower threshold means more of the vocal gets compressed. Set it so the compressor works on the louder words, not the whole take.
• Ratio sets how hard it clamps once past the threshold. 2:1 is gentle leveling, 4:1 is firm control, 8:1 and above is heavy limiting-style compression.
• Attack sets how fast the compressor grabs the signal. Fast attack catches peaks and consonants; slow attack lets the initial transient through, keeping the vocal punchy.
• Release sets how fast it lets go. Too fast sounds pumpy, too slow stays clamped into the next word. Aim for a release that recovers by the next syllable.

Makeup gain then brings the compressed, now-quieter signal back up, so the more consistent vocal sits at a forward level.

What settings should I start with?

For a single, general-purpose vocal compressor, these are a solid starting point you can adjust by ear:

• Ratio around 3:1 to 4:1 for controlled but natural leveling.
• Attack around 5 to 15 ms so consonants punch through before the clamp.
• Release around 40 to 100 ms, or auto-release, timed to recover between words.
• Threshold set for 3 to 6 dB of gain reduction on the loudest words - watch the meter, not just your ears at first.

If the vocal sounds squashed or lifeless, your attack is probably too fast or your ratio too high. If it still jumps around, lower the threshold or raise the ratio a little.

Why use serial compression?

A raw vocal can swing 15 dB or more between quiet and belted words. Asking a single compressor to control that means heavy gain reduction, which sounds unnatural and pumps. Splitting the job fixes this.

• First compressor - catch the peaks. Fast attack, medium release, ratio around 4:1, pulling 3 to 6 dB on the loudest words. This tames the biggest jumps.
• Second compressor - level and glue. Slower, gentler, ratio around 2:1 to 3:1, pulling another 2 to 4 dB fairly constantly. This evens out the overall body and adds cohesion.

Because each stage only removes a few dB, neither has to work hard, and the result is a smooth, consistent vocal with no obvious pumping. This is how professional vocals stay glued and forward through a whole song. It also pairs with clip gain - smoothing the biggest jumps by hand before compression means the compressors work even less.

How do attack and release change the sound?

These two controls shape character, not just level.

• Faster attack catches more of the transient, which controls consonants and can soften punch. Slower attack keeps the vocal's front-edge energy and feels more lively.
• Faster release makes the compressor recover quickly, raising perceived density and energy but risking a pumping feel. Slower release is smoother but can dull dynamics if it stays clamped.
• For an aggressive, upfront vocal, lean into faster attack and release. For a natural, transparent vocal, use a moderate attack and a release timed to the syllables.

Set release by ear against the tempo - you want the meter returning to zero just before the next word, so the compression breathes with the performance.

Should I use parallel compression on vocals?

Parallel compression - blending a heavily compressed copy under the clean vocal - is a powerful way to add density and consistency without squashing the dynamics of the main signal. Instead of compressing the vocal hard and losing its life, you keep the natural track upfront and mix in the crushed version underneath for body.

• Send the vocal to a parallel bus and compress that copy hard - ratio 6:1 or higher, fast attack, pulling 10 dB or more.
• Blend the crushed copy in low, under the clean vocal, until the vocal feels denser and more even but still breathes.
• Use it to lift buried words - the parallel copy raises the quiet detail without pulling the loud words down, which keeps the vocal punchy and present.

Parallel compression is how you get an aggressive, upfront, dense vocal that still has dynamics, which is why it is so common on modern pop and rap leads.

How does compression fit with the rest of the chain?

Compress after EQ, so the compressor reacts to a tonally balanced signal rather than clamping down on mud or harshness. De-ess after compression, because compression raises sibilance. A clean order is clean the take, tune, EQ, compress in two stages, de-ess, then add space. For the full sequence see how to make vocals sound professional, and for the EQ that comes first, how to EQ vocals.

Compression is also the biggest tool for perceived loudness, because a consistent vocal always feels louder than a swinging one. If you want that consistency without dialing in two compressors by hand, Vocal Polish applies balanced compression as part of one pass, then you can master the full track at Sauce Mastering. More in the vocal guides.