Skip to content
  1. Blog
  2. Mixing

MixingMixingCompressorGetting started

What a Compressor Is: Narrowing the Gap Between Loud and Quiet

A compressor turns down only what goes over a set level, evening out the volume. What threshold, ratio, attack, release and make-up gain mean, a starting point for vocals, and the signs of overdoing it.

Written by the site’s operator3 min read
UREI 1176LN compressors in a rack
A studio staple since the 1960s. Its controls work much like the compressor plugins you use today.Photo: Ville Hyvönen / CC BY-SA 2.0 / Wikimedia Commons

Record a vocal and the chorus is loud while the whispered verse is hard to hear. Moving the fader up and down never quite fixes it.

That is what a compressor is for. In a phrase: it automatically turns down only the parts that are too loud, narrowing the gap in level.

What a compressor does

A compressor watches the level constantly and turns the sound down only while it exceeds a set level. Quieter parts pass unchanged, so the gap between loud and quiet shrinks.

Because the gap is smaller, you can raise the whole thing, so the quiet parts become easier to hear. That is its main effect.

Only the excess above the threshold is reducedInput levelOutput levelNo compressionThresholdRatio 4:1
Only the part above the threshold is reduced, by the ratio. Anything below passes unchanged.

The five controls

ControlMeaningIn short
ThresholdThe level above which it starts workingWhere from
RatioHow much the excess is reduced (4:1 turns 4 dB over into 1 dB)How much
AttackHow fast it acts once the level goes overWhen it starts
ReleaseHow fast it lets go once the level drops backWhen it stops
Make-up gainRaises the whole signal to make up for the reductionRestore at the end
The gain reduction (GR) meter shows how much it is reducing right now.

Attack and release change the character

  • Fast attack: grabs the start of each note, rounding off transients and pushing the sound back.
  • Slow attack: lets the start through first, so word onsets and drum hits stay forward.
  • Fast release: recovers quickly, denser sound, but too fast makes the level pump.
  • Slow release: recovers gently and naturally, but too slow keeps the next note squashed.

A starting point for vocals

These are starting points to adjust by ear; the right answer depends on the voice and the song.

ControlStart here
Ratio3:1 to 4:1
ThresholdWhere loud passages show about 3–6 dB of GR
AttackAround 10 ms (slower if word onsets get buried)
ReleaseAround 50–100 ms (follow the rhythm of the words)
Make-up gainUntil it is about as loud as with the compressor bypassed

Signs of overdoing it

  • Breaths and mouth noises get loud. Quiet sounds are being lifted too much.
  • Flat, lifeless delivery. Dynamics are part of the performance; not everything needs to be the same level.
  • The level breathes or pumps. Release too fast, or too much reduction.
  • Word onsets get squashed. Attack too fast.

For heavy reduction, two gentle compressors in series often sound more natural than one working hard.

How limiters and maximizers relate

A limiter is a compressor with a very high ratio, set so the level never goes past a ceiling. The last device in mastering, to prevent clipping, is usually a limiter. A maximizer is easiest to think of as that limiter combined with a way to raise the overall level.

Summary

  • A compressor reduces only what goes over, narrowing the level gap.
  • Where from (threshold), how much (ratio), when it starts (attack), when it stops (release).
  • For vocals, start around 3:1–4:1 with 3–6 dB of GR and adjust by ear.
  • Compare at matched level; overdoing it shows as loud breaths, flatness and pumping.

Combined with EQ for tone, that covers the mixing basics. The full workflow is in where to start.

Tools for this article

Read next

Mixing

What EQ Is: What to Cut, How Much, and Why

EQ is a volume control for each range of pitch. What you hear in the lows, mids and highs, why to start by cutting rather than boosting, what Q means, and what to try first on a vocal, with concrete frequencies.

Mixing

What reverb is: giving a voice a room to sit in

Reverb adds the reflections a room would give a sound. What pre-delay, early reflections and decay time mean, how room, hall and plate differ, why reverb goes on a send, and starting points for blending a cover vocal into a backing track.

Mixing

Mastering basics: what happens after the mix

Mastering is the last step: turning a finished mix into something ready for streaming or CD. How it differs from mixing, the order of EQ, compression and limiting, streaming loudness (−14 LUFS, −1 dBTP), export formats, and how to approach a cover.

Mixing

Where Does Harshness Live? Finding and Taming Piercing Frequencies

Most piercing sound comes from a very narrow peak somewhere around 2–5 kHz. Why hearing is so sensitive there, how to find the peak by sweeping a narrow EQ, how many spots and how much to cut, and how sibilance differs.

Mixing

What a limiter is: a level that is never crossed

A limiter catches anything louder than a set ceiling, instantly, so the track does not clip. How it differs from a compressor, why the ceiling goes at −1 dBTP, what true peak means, and how much limiting is too much.

Mixing

What a maximizer is: making a track louder, and why not too loud

A maximizer raises a track’s overall loudness while a limiter holds the peaks down. How it relates to a limiter, how streaming services even out volume (loudness normalisation), what −14 LUFS means, and the signs of pushing too far.

All articles

Sponsored links