A good vocal mix is not a pile of plugins. It is a sequence of decisions: remove problems, control dynamics, shape tone, place the voice in space, then check whether the lyric still feels close.
Most bad vocal mixes fail because the steps happen in the wrong order. If you add reverb before the vocal is controlled, the room exaggerates every harsh syllable. If you boost air before de-essing, the vocal gets shiny and painful. Use this order first, then break it later when you know why.
| Step | Goal | Common mistake |
|---|---|---|
| Cleanup | Remove noise, clicks, bad breaths | Editing after compression |
| Subtractive EQ | Cut mud and harsh resonances | Boosting before fixing |
| Compression | Keep words at a stable level | Crushing every transient |
| De-essing | Control S and T sounds | Making the vocal dull |
| Space | Add reverb and delay without pushing the vocal back | Too much wet signal |
Start with editing, not plugins. Cut headphone bleed before the first word, remove obvious clicks, and lower breaths that jump out. Do not delete every breath; a vocal with no breath can feel unnatural. Pull loud breaths down 3–8 dB instead.
Set the raw vocal so the average phrase is comfortably below clipping. You do not need a magic number, but leaving headroom makes compressors and saturation behave more predictably.
Automation beats rescue plugins. If one word is much louder than the rest, lower that clip before the compressor. The compressor should shape the performance, not fight one random peak.
Before making the vocal brighter, remove what blocks the lyric. A typical vocal has low rumble, low-mid boxiness, and a few narrow resonances that poke out only on certain notes.
"EQ should make the lyric easier to understand, not just make the vocal brighter."
Vocal compression is about intelligibility. The quiet endings of words need to stay audible without the loud notes jumping out. Start with a moderate ratio around 2:1 to 4:1 and aim for a few dB of gain reduction on normal phrases.
Attack controls how much of the consonant gets through. Too fast and the vocal loses life. Too slow and peaks still jump out. Release controls how quickly the compressor lets go between words. If the release is too slow, the vocal feels pinned down; if it is too fast, it can pump.
Try two compressors instead of one extreme compressor. One gentle compressor for leveling, then a second for tone or peaks, usually sounds more natural than asking one plugin to do everything.
Sibilance is not always at 8 kHz. Some voices spit around 5–6 kHz; others get sharp above 10 kHz. Find the worst S or T syllable, loop it, and tune the de-esser to that range.
The goal is not to remove every S sound. The goal is to stop the sharp syllables from feeling louder than the lyric. If the vocal suddenly sounds like it has a blanket over it, the de-esser is working too hard or listening to the wrong frequency.
After cleanup, compression, and de-essing, add tone. A small presence boost around 3–5 kHz can bring the vocal forward. A gentle high shelf around 10–16 kHz can add air. Do this late in the chain so you are enhancing a controlled vocal, not exaggerating problems.
Use the beat as context. A dark track can handle more air. A bright track may need less top end and more midrange focus so the vocal sits without fighting hats and synths.
Reverb gives the vocal a room. Delay gives it depth without washing out the words. For modern vocals, start with a short plate or room reverb and a tempo-synced delay. Filter both effects so they do not fill the low mids or hiss above the vocal.
"If the vocal disappears when the hook gets busy, the problem is usually space, not volume."
Use Music Producer Lab to practice EQ, dynamics, and ear-training moves in short interactive lessons.
Start free →Short lessons on EQ, de-essing, space, and vocal tone. Follow @musicproducerlab.
Follow @musicproducerlab