Vocal Compressor Attack and Release Settings That Actually Work
Vocal mixing · 4 min read
Set attack slow enough to let the consonant through (10–30 ms) and release fast enough to recover before the next word (40–120 ms, or tempo-matched). Fast attack + slow release is what makes vocals sound dull and lifeless.
Starting point
10–30 MS
Attack · release 40–120 ms · 3–4 dB gain reduction
The 60-second answer
Attack 10–30 ms, release 40–120 ms, ratio 3:1, threshold set for 3–4 dB of gain reduction on the loudest words. Slower attack = more punch and clarity. Faster release = more density, but past a point it pumps.
What attack and release actually do to a voice
Attack is how quickly the compressor reaches full gain reduction once the signal crosses the threshold. Release is how long it takes to let go. On a voice, the first 5–20 milliseconds of a word is the consonant — the T, the K, the P. That transient is most of what makes a vocal sound clear and in-your-face. Clamp it and the word turns to mush; let it through and the vocal cuts without being louder.
Why fast attack sounds worse, not tighter
Producers reach for a 1 ms attack because it feels like more control. What it actually does is flatten the consonant along with everything else, so you lose the transient that made the word intelligible and keep the vowel that was already loud. The vowel is the loud part; the consonant is the intelligible part. Compress the vowel, protect the consonant.
The settings, step by step
| # | Do this |
|---|---|
| 01 |
Start with ratio 3:1 Low enough to sound natural, high enough to do something. Save 8:1 and above for the second compressor in the chain. Ratio 3:1 |
| 02 |
Set attack to 15 ms and listen to the consonants Play the busiest line. If T and K sounds feel blunted, go slower. If the vocal jumps out unevenly, go faster. Attack 10–30 ms |
| 03 |
Set release by the tempo of the song Roughly one sixteenth note. At 120 BPM that is 125 ms; at 140 BPM about 107 ms. The compressor should be back to zero before the next syllable arrives. Release 40–120 ms |
| 04 |
Lower the threshold until you see 3–4 dB on the loudest words Not average. The peaks. If the needle is pinned at 10 dB the whole time, you are using compression to fix a performance problem. Threshold: 3–4 dB GR |
| 05 |
Match the makeup gain by ear, not by number Bypass and compare at the same perceived level. Louder always sounds better for the first three seconds, and that is how bad decisions get made. Makeup: level-matched |
| 06 |
Do it twice, gently 3 dB on one compressor, 3 dB on another beats 6 dB on one. Two small moves stay invisible; one big move always sounds like a plugin. 2 x 3 dB |
The trick: set release with the meter, not your ears
Turn the release all the way down (fastest) and watch the gain reduction meter while the vocal plays. It will look like it is having a seizure — jumping back to zero between every syllable. Now slow the release down until the needle stops fully returning between words but still recovers between phrases. That is the point where the compressor is moving with the rhythm of the vocal instead of fighting it. Takes fifteen seconds and it works on every song, every voice, every compressor.
Stop guessing the chain.
The Vocal Chain Bible is 88 real vocal chains from Grammy-winning engineers — every plugin, every setting, in order, with free alternatives for each one. 5+ years, 100+ interviews, one PDF.
Get the Vocal Chain Bible — $48 →Mistakes that kill it
- Fast attack on everything. 1 ms attack on a rap vocal turns every consonant into a soft edge. You will then reach for an exciter to fix the dullness you just created.
- Release so slow the compressor never lets go. If gain reduction never returns to zero, the whole vocal is just quieter — you have built an expensive volume knob.
- Compressing a badly recorded take. If the singer moved 30 cm off the mic mid-take, ride the clip gain first. Compression is for the last 3 dB, not the first 12.
- Judging with the level jumping. Makeup gain fools everyone. Level-match before you decide anything.
Once the attack and release are right, most of the vocal mixing work is already done. Everything after that — EQ, saturation, reverb — is decoration on top of a stable level.
Keep going
FAQ
What attack time should I use for rap vocals?
Start at 10–15 ms. Rap lives on consonant clarity, so protecting the front of each word matters more than smoothing the sustain.
What is the best release time for vocals?
Tempo-matched to roughly a sixteenth note: 40–120 ms for most songs. Slower for ballads, faster for fast, dense delivery.
How much gain reduction is too much?
Anything over 6 dB on a single compressor starts to sound like an effect. If you need more, split it across two compressors or fix the level with automation first.
Who's behind this
Ivan Meshcheriakov · IvanFromRnD
5+ years ago I moved to LA with zero connections and started building from nothing — 14-hour days, meeting people and interning at studios. It took me 5+ years to meet and interview 100+ Grammy-winning engineers — then I turned it all into the Vocal Chain Bible.
Read my full story →