Learning how to compress vocals is about taming dynamics so every word sits at a consistent, audible level in the mix. You lower the loudest moments so you can raise the whole vocal, making it feel upfront and controlled. The core controls — threshold, ratio, attack, and release — are simpler than they look once you know what each one does.
This guide explains the controls, gives practical starting settings, and shows when to split the work across two compressors for a natural result.
Why vocals need compression
A sung vocal swings widely in volume: quiet verses, loud choruses, words that jump out, others that vanish. Compression evens that out so the vocal stays present without burying quiet phrases or letting loud ones spike. It is one half of the classic vocal duo — see EQ and compression fundamentals for how the two work together.
The controls, explained
- Threshold: the level above which compression kicks in. Lower it to compress more of the vocal.
- Ratio: how hard it clamps down above the threshold. 2:1 to 4:1 suits most vocals; higher ratios are more aggressive.
- Attack: how fast it reacts. Faster attack catches transients and sounds smoother; slower attack lets consonants punch through.
- Release: how fast it lets go. Set it to breathe with the song’s tempo so the vocal does not pump or sound choppy.
- Make-up gain: raises the now-controlled vocal back up to a useful level.
If these terms are new to you, our beginner’s guide on how to use a compressor walks through each one from scratch.
A reliable starting point
Settings always depend on the performance, but this gets you in the zone:
- Ratio around 3:1.
- Threshold set so you see roughly 3–6 dB of gain reduction on the loudest phrases.
- A medium attack so consonants still cut through.
- A medium-to-fast release that follows the vocal without pumping.
Adjust make-up gain so the compressed and bypassed versions are the same loudness, then judge whether it sounds better — not just louder.
How to dial it in by ear
Numbers get you started, but the final settings come from listening. Work in a sensible order so each control does an obvious job rather than fighting the others.
- Set the ratio first. Start gentle at 2:1 or 3:1. You can always push harder, but a low ratio keeps you honest about how much you actually need.
- Lower the threshold next. Bring it down until the gain-reduction meter moves on the louder phrases. Aim for the meter to rest at zero between phrases and only pull down when the vocal pushes — if it never lets go, the threshold is too low.
- Tune the attack to taste. Too fast and the vocal loses its consonant snap and starts to sound dull; too slow and the loud transients slip past uncontrolled. Sweep it while the chorus plays and stop where the words stay crisp but even.
- Set the release to the song. A release that recovers roughly in time with the tempo feels musical. If you hear the level “breathing” up between words in a distracting way, slow it down a touch.
Bypass often. Matching loudness when you A/B is the only way to tell whether the compressor is genuinely helping the vocal sit in the track or simply making it louder.
Serial compression: split the work
Pushing one compressor hard enough to fully control a dynamic vocal often sounds unnatural. A cleaner approach is two compressors in series, each doing a little:
- First compressor: gentle, catching the broad dynamic swings between verse and chorus.
- Second compressor: faster, evening out word-to-word level.
Two compressors each removing a few decibels sound smoother than one removing ten. This is a standard trick on professional vocal chains. A common pairing uses a slower, lower-ratio compressor first to manage the overall arc of the performance, then a faster optical-style or FET-style compressor to glue the detail. You do not need expensive plugins to do this — two instances of the same stock compressor, each set gently, already gets you most of the benefit.
Compression and automation work together
Compression is not the only tool for controlling level, and it should not do all the work alone. Volume automation — drawing the level up or down phrase by phrase — handles the big, deliberate moves a compressor cannot make musically, such as lifting a soft verse line or pulling back a shouted word. If you have not set this up before, see how to use automation in a DAW. A useful workflow is to ride the obvious problem spots with automation first, then let the compressor handle the fine, fast level changes that would be tedious to draw by hand. This keeps the compressor working in a comfortable range instead of slamming on a single out-of-place word.
Where compression sits in the chain
Most engineers compress after corrective EQ, so the compressor reacts to a cleaned-up signal rather than chasing low-end rumble or harsh resonances. Our guide on how to EQ vocals covers that step. Heavy sibilance is best handled with a de-esser, not compression. For the full vocal treatment, see how to mix vocals and the mixing and mastering hub.
Common mistakes to avoid
- Over-compressing: if the vocal sounds squashed or lifeless, ease off the ratio or raise the threshold.
- Judging by loudness: always match levels when comparing, or compression will always seem like an improvement.
- Fixing performance with a compressor: a wildly uneven take is better re-sung or comped together from the best takes; ride the volume with automation for the rest.
- Ignoring the release: a release that is too short can make the vocal sound grainy or pumping, while one that is too long keeps the compressor clamped and squashes the life out of phrases.
- Compressing a noisy take: raising the make-up gain also raises any background hum or room noise sitting under the vocal, so clean the recording before you squeeze it.
Frequently asked questions
How much should I compress vocals?
A good starting point is around 3–6 dB of gain reduction on the loudest phrases, with a ratio near 3:1. If the vocal sounds squashed or lifeless, you are compressing too much — ease off or split the work across two gentler compressors in series.
Should I EQ or compress vocals first?
Usually EQ first, so the compressor reacts to a cleaned-up signal rather than chasing low-end rumble or harsh frequencies. Many engineers then add a second, gentle EQ after compression for final tonal shaping.
What attack and release settings work best for vocals?
A medium attack lets consonants punch through while still controlling the body of notes, and a medium-to-fast release that follows the song’s pace keeps the vocal even without pumping. There are no universal numbers — set them by ear against the performance.
Does a busy mix need more vocal compression?
Often, yes. A dense, loud arrangement leaves less room for a vocal to peek through, so tighter dynamic control helps it stay consistently audible. Rather than pushing one compressor harder, add a little more gentle compression in series and lean on volume automation for the spots that still get lost.
