Mixing Vocals That Sit: Five Decisions That Matter More Than Your Plugins

Ask ten engineers how to mix a vocal and you'll get ten plugin chains, each defended with religious conviction. But listen closely to great vocal mixes across genres — from sparse folk to dense electronic production — and you'll notice they share something that has nothing to do with which compressor was used. The vocal sits in a specific place, it's intelligible, and it feels like it belongs to the same world as the instruments. Getting there is mostly a series of decisions made before any processing.

The first decision is the performance and the recording. A vocal that was tracked with the singer too close to the mic will have proximity boost in the low-mids that no EQ can remove without thinning the voice. A vocal tracked in a dead-sounding closet will need artificial reverb later, and artificial reverb rarely convinces as well as a decent room. If you have any control over tracking, spend your effort there. Move the singer back a hand's width, angle the mic slightly off-axis, and record three takes even if you think you nailed the first one. Comping from three takes sounds better than any plugin chain applied to one.

The second decision is arrangement. If the vocal feels buried, the problem is often that too many instruments are occupying the same frequency range and the same rhythmic space. Before you reach for a sidechain or a dynamic EQ, try muting something. Guitars with heavy chorusing, synth pads with wide stereo spread, and cymbal-heavy drum overheads all compete with a lead vocal. Narrow them, pan them, or cut them during vocal sections. Space is not something you create with plugins; it's something you leave alone.

Compression, EQ, and the Order That Works

Start with a high-pass filter. Almost every vocal benefits from having everything below roughly 80 to 100 Hz removed, because that range contains rumble, plosives, and no useful vocal information. This single move often clarifies a muddy mix more than any other. Then address problem frequencies with narrow cuts rather than broad boosts. If the vocal sounds honky, find the offending band by sweeping a boosted narrow EQ until it sounds worst, then cut that exact spot. Broad boosts applied before you've found the problems tend to make the vocal louder without making it clearer.

Compression should serve intelligibility, not volume. A vocal with wild dynamic swings — a whispery verse and a shouted chorus — needs gentle compression in stages rather than one aggressive compressor. Two compressors each doing two or three dB of gain reduction will sound more natural than one doing six. Set the attack slow enough to let consonants through and the release fast enough that the compressor recovers between words. If you can hear the compressor working, it's working too hard.

De-essing deserves its own mention because it's the most commonly botched step. Sibilance — the harsh S and T sounds — sits somewhere between 5 and 10 kHz and varies wildly between singers and microphones. A dedicated de-esser is ideal, but a narrow EQ cut applied only to the offending syllables via automation works just as well and preserves more of the vocal's natural brightness. Don't de-ess the whole track; de-ess the moments that need it.

Effects as Placement, Not Decoration

Reverb and delay tell the listener where the vocal is standing. A short plate reverb with a fast pre-delay puts the singer close to the mic. A long hall with a slow pre-delay pushes them back into a room. Neither is better — it depends on the song. The mistake is treating reverb as a default setting rather than a spatial choice. Ask yourself where the vocal should be, then build the effect to put it there.

Sending the vocal to a shared reverb bus alongside other instruments glues the mix together, because everything appears to be in the same space. A separate, more prominent reverb just for the vocal makes it stand apart. Both approaches are valid and both are used constantly. What matters is that the choice is deliberate. When the vocal finally sits, it's rarely because of one magical plugin. It's because five small decisions — performance, arrangement, filtering, staged compression, and deliberate space — all pointed the same direction.