How to compress vocals with a FET compressor

For a lead vocal, start on a clean revision at ratio 4 with a medium attack (3) and a medium-fast release (5). Drive INPUT until the loudest words show 3 to 6 dB of gain reduction. Filter the lows out of the sidechain so plosives stop pulling the whole phrase down.

A FET compressor is the classic tool for a vocal that has to sit on top of a busy mix. It reacts fast enough to catch every syllable, and it adds a little edge that helps the words cut through. The trick is to use that speed without making the voice sound squashed.

Step 1: pick the right revision

Not every revision suits every voice. In PINCH-76 the choice takes one click, and AUTO GAIN keeps the level steady while you compare.

Voice	Revision	Why
Pop, R&B, ballad	Rev D (LN)	Clean and even; the low-noise version stays out of the way.
Has to cut through a dense mix	Rev F	Broad presence lift around 3.4 kHz, mostly odd-order colour.
Rock, rap, gritty	Rev A	About three times the harmonic content of the others.
Gentle levelling, backing vocals	AE	Softer ratios (2, 4, 8, 20) and a slower top attack.

Step 2: set the amount with INPUT

There is no threshold knob. The ratio keys set how hard the compressor works; INPUT decides how much signal reaches it. Raise INPUT until the needle moves 3 to 6 dB on the loud words and comes back between phrases. If you want a louder, more pinned vocal, latch 4 and 8 together instead of 4 alone.

Step 3: attack and release

Remember that the timing knobs run backwards: clockwise is faster (more on that in attack and release explained by ear). ATTACK 3 lets the start of each word through, so consonants stay clear. RELEASE 5 recovers between words, so the vocal does not sound like it is breathing. If you hear the level pumping between words, slow RELEASE to 4.

Step 4: stop plosives from pumping

A loud "p" or "b" is mostly low-frequency energy. Without a filter it hits the detector, the compressor clamps down and the next few words drop in level. The sidechain HPF in the ADVANCED drawer at 100 Hz takes those lows out of the detector, while the vocal itself is untouched. Press LISTEN to hear what the detector hears.

The two-stage vocal chain

Most modern lead vocals are compressed twice: a fast compressor catches the peaks, then a slower one levels the phrase. Each one does a small job, so neither has to work hard, and the vocal sounds controlled without sounding squashed.

In PINCH-76 you can build this chain in one plug-in. Switch DUAL UNIT to SER and turn UNIT LINK off. Set unit 1 to Rev A, ratio 20, ATTACK 6, RELEASE 6, with only 2 to 4 dB of reduction. Set unit 2 to Rev D (LN), ratio 4, ATTACK 3, RELEASE 4, with 3 to 5 dB.

Unit 2's needle will move less than you expect. That is correct: it sees a signal whose peaks are already tamed. Try swapping the revisions too; clean first and dirty second sounds noticeably different.

Vocal compression questions

How much compression is right for vocals? For most lead vocals, 3 to 6 dB of gain reduction on the loud words. If the needle never returns to zero between phrases, the vocal loses its life; back INPUT off.

What is the best attack and release for vocals? Start at ATTACK 3 and RELEASE 5 in PINCH-76. On a slow song, slow RELEASE to 4, because a fast release brings up breaths and room noise between words.

Should the de-esser go before or after the compressor? After. A fast FET compressor can make "s" sounds louder, so tame them once the level is set.

Why does my compressed vocal sound better? Often only because it is louder. Use AUTO GAIN or ☰ › Level-match A/B before you judge.