How to Compress Vocals: Settings, Serial Compression, and Parallel

The short answer: compress vocals in small serial stages rather than one big squash. Start with a fast compressor (1176-style) catching only the peaks with 2 to 4 dB of gain reduction, then a slower, smoother leveler (LA-2A-style) doing another 2 to 3 dB to even out the performance. A single-compressor starting point: ratio 3:1 to 4:1, attack 5 to 20 ms, release 50 to 150 ms (or auto), 3 to 6 dB of reduction on the loud phrases. Before any of it, clip-gain the wildest phrases by hand so the compressors only do fine work. Two compressors doing 3 dB each sound like control; one doing 8 dB sounds like a compressor.

Serial compression is on every lead vocal I mix for my own releases (Sony, Universal, Ultra, Dim Mak). It's the stage that makes a vocal feel close and expensive.

Starting settings

Step 0: clip gain first

The move most people skip. Before any compressor, turn down the phrases that jump out and turn up the ones that drop away, by hand, using clip gain in your DAW. Ten minutes of this means the compressor works 2 to 3 dB instead of 8, and 2 to 3 dB is where compressors sound good. It's covered in the full vocal mixing process.

Serial compression: fast, then smooth

1. Compressor 1, the peak catcher. A fast FET-style compressor (the 1176 sound, or your stock compressor with a fast attack and release). Ratio 4:1 to 8:1, catching only the loudest syllables for 2 to 4 dB. It adds attitude and keeps peaks from spiking.
2. Compressor 2, the leveler. A slower optical-style compressor (the LA-2A sound, or stock with slower settings), ratio around 3:1, 2 to 3 dB of gentle, constant reduction. It evens out the performance so every word sits at a similar level.

The order can flip (leveler first, peak catcher after); both work. What matters is that neither works hard.

Parallel compression for density

Send the vocal to a bus with a heavily compressed copy (10 dB or more of reduction), then blend it underneath the main vocal. The dry vocal keeps its natural dynamics and the crushed copy fills in the quiet details, so the vocal feels dense and upfront without sounding squashed. It's especially effective on rap and pop vocals that need to cut through dense beats.

How to tell it's too much

• Pumping: the background noise or reverb swells between words. Slower release or less reduction.
• Lost consonants: the vocal sounds dull and blunt. Slower attack lets the transients through.
• Sibilance jumps: compression brings esses up. That's why the de-esser usually comes after compression.
• Breaths got loud: turn breaths down with clip gain rather than fighting the compressor.

Always bypass-compare at matched level. Louder sounds better for a few seconds, which is how vocals end up over-compressed.

Where it fits in the chain

Corrective EQ, then compression, then tonal EQ, de-essing, saturation, and space. How to EQ vocals covers the stages on either side, and the vocal chain guide shows the whole order. Stock compressors handle all of this; the best vocal plugins list covers the paid options.

Quick answers

What are good compressor settings for vocals? Ratio 3:1 to 4:1, attack 5 to 20 ms, release 50 to 150 ms or auto, and 3 to 6 dB of gain reduction on the loud phrases. Adjust by ear from there.

How many compressors should I use on vocals? Usually two in series, each doing a little (2 to 4 dB). It sounds more natural than one compressor doing a lot.

Should I compress vocals before or after EQ? Corrective EQ before compression, tonal EQ after. Many engineers also de-ess after compression.

What is parallel compression on vocals? Blending a heavily compressed copy of the vocal under the natural one, for density and presence without flattening the dynamics.