Compression keeps a vocal at a steady level so every word reaches the listener: the quiet syllables come up, the loud ones come down, and the voice stops jumping in and out of the beat. The settings below are starting points. The final word is always your ears and the gain-reduction meter.
- Ratio: 2:1–4:1 for singing, 4:1–6:1 for rap
- Attack 3–30 ms, release 40–150 ms as a start
- Aim for 3–6 dB of gain reduction on loud phrases
- Two compressors taking a little each sound more natural than one doing everything
- Match the output level before you judge it
What each control does
- Threshold: the level where compression starts. Lower threshold, more compression.
- Ratio: how hard it pushes down above the threshold. At 4:1, 4 dB over becomes 1 dB over.
- Attack: how quickly it reacts. Fast attack catches peaks but can dull consonants; slower attack lets the start of each word through, which keeps rap punchy.
- Release: how quickly it lets go. Too fast sounds jumpy or distorted; too slow and it's still compressing the next word.
- Makeup gain / output: brings the level back up after compression. Match it to the bypassed level to compare honestly.
Vocal compression cheat sheet
| Source | Ratio | Attack | Release | Gain reduction |
|---|---|---|---|---|
| Rap lead | 4:1 – 6:1 | Fast–medium, 3–10 ms | Fast–medium, 40–100 ms | 4–8 dB total |
| Sung lead | 2:1 – 4:1 | Medium, 10–30 ms | Medium, 60–150 ms | 3–6 dB total |
| Adlibs / backing | 4:1 – 8:1 | Fast, 1–5 ms | Fast, 30–80 ms | 6–10 dB |
| Vocal bus (glue) | 1.5:1 – 2:1 | Slow, 20–30 ms | Auto or 100–300 ms | 1–2 dB |
For a rap lead, the "total" gain reduction is ideally split across two compressors, as below.
How to set a vocal compressor in five steps
- Start with a 4:1 ratio, medium attack (around 10 ms) and medium release (around 80 ms).
- Lower the threshold until the loudest phrases show 4–6 dB of gain reduction and quiet phrases show little or none.
- Adjust attack: slower if consonants lose their bite, faster if loud syllables still jump out.
- Adjust release so the meter falls back between words, roughly in time with the song.
- Set the output so bypassing doesn't change the volume, then decide if it's actually better.
Serial compression: why two compressors beat one
Asking one compressor for 10 dB of gain reduction usually sounds squashed. Splitting the job sounds more natural, and it's what most pro vocal chains do:
- First, a fast peak-catcher (1176-style, FET): quick attack and release, grabbing 3–5 dB off the loudest moments.
- Then a smooth leveller (LA-2A-style, optical): slower and program-dependent, evening out another 2–3 dB across the phrase.
Some chains add a third, very light stage at the end, or a multiband compressor that controls only the boomy low mids or harsh upper mids. The chains in our Ultimate Vocal Chain mostly stack compressors this way: an LA-2A or LA-3A-style stage, a multiband, and a final general-purpose compressor.
1176 vs LA-2A style for vocals
- 1176 (FET): fast, aggressive, adds bite and energy. Great on rap and on adlibs that need to stay in your face.
- LA-2A (optical): slow, smooth, gentle. Great on sung leads, or second in line after an 1176.
- LA-3A: a solid-state cousin of the LA-2A, a little faster and brighter. A good middle ground for pop and R&B.
Where compression goes in the chain
After pitch correction and corrective EQ, before tonal boosts, saturation and effects. Put a de-esser before it if the "s" sounds keep triggering heavy compression. The full order, with a real chain broken down, is in vocal chain order explained.
Signs of too much compression
- Breaths and mouth noise become as loud as the words.
- The vocal sounds flat and lifeless, with no difference between verse and hook.
- Audible "pumping": the level ducks and swells in a rhythm.
- Sibilance gets louder. Compression brings the quiet, hissy parts up too.
If you hear these, raise the threshold or split the work across two compressors.
FAQ
What is a good compression ratio for vocals?
2:1 to 4:1 for most singing, 4:1 to 6:1 for rap. Higher ratios are for peak control or effect.
How much should I compress vocals?
Around 3–6 dB of gain reduction on the loudest phrases is a common starting point. Rap can take more, ideally split across stages.
Should I compress vocals while recording?
Only lightly, if at all. You can always add compression later but you can't remove it from a recording. Monitor through a compressor if it helps the performance, and record the clean signal.
Fast or slow attack for rap vocals?
Usually medium-fast (3–10 ms). Too fast removes the punch from consonants. Start slower and speed it up until the level is under control.
