A vocal that won't sit is almost never a fader problem. Fix it in five layers: compress for consistency (3–6 dB of gain reduction) so every word holds its place, give the vocal a tonal lane (clean lows, gentle 3–5 kHz presence), cut the instruments masking the 1–5 kHz range, duck competing tracks 1–2 dB with sidechain compression, and finish with pre-delayed space plus 0.5–1 dB fader rides.
- 'Floats on top or drowns' is the signature of an unprocessed dynamic range, not a wrong fader position.
- Compression consistency is the foundation: a vocal can only sit at one level if it stays at one level.
- Masking is the invisible enemy — carve 2–4 dB around 2–4 kHz from the instruments crowding the vocal, not the vocal itself.
- Subtle sidechain ducking (1–2 dB) from the vocal into busy instruments creates room you can't hear as pumping.
- The last 10% is automation: small fader rides per section beat any static setting.
Diagnosing why it won't sit
'Sitting in the mix' means the vocal reads as part of the record — connected to the instrumental, always intelligible, never pasted on top. When it won't sit, one of two failure modes is usually happening: the vocal's level is inconsistent (some words vanish, so you push the fader until other words float), or something in the arrangement occupies the vocal's frequency range and masks it whenever both play at once.
The quick test: loop the busiest chorus. If the vocal disappears only on certain words, it's a dynamics problem — start at layer one below. If the whole vocal recedes whenever the track fills up, it's a masking problem — skip to layer three. Most real mixes need some of both.
Layer 1 — Make the vocal hold still
A vocal with 12 dB of natural swing can't sit anywhere: whatever fader position works for the loud phrases loses the quiet ones. Compression narrows that swing so the whole performance lives at roughly one level — the precondition for everything else in this guide. Start with the standard recipe from our compression settings guide: 4:1, 10 ms attack, auto release, 3–6 dB of reduction, in series with a second gentle stage if the performance is wild.
Layer 2 — Give it a lane tonally
A vocal cuts through on its midrange presence, not its volume. Clean up the lows and low-mids (high-pass, mud cut) so the vocal isn't wasting energy where the instrumental owns the spectrum, and add a gentle presence boost around 3–5 kHz — the full map is in the vocal EQ cheat sheet. A touch of saturation does similar work by adding midrange harmonics that survive a dense mix.
Layer 3 — Move the furniture (masking)
Here's the counterintuitive part: when the vocal still won't cut through, the fix is usually on other tracks. Synth pads, guitars, keys, and cymbals all pour energy into 1–5 kHz — the exact range speech intelligibility depends on. While the vocal plays, cut 2–4 dB around 2–4 kHz on the loudest sustained instruments (wide Q), and high-pass anything that doesn't need its low end. The vocal gets louder without its fader moving.
This is why no vocal chain alone can fully solve 'sit in the mix': a chain processes the vocal, but masking lives in the arrangement. PhantomRack gets your vocal's own tone and dynamics to reference level — the carving of the tracks around it is the mix move that remains yours.
Layer 4 — Subtle sidechain ducking
Instead of (or on top of) static EQ carving, let the vocal itself push instruments out of the way: put a Compressor on the pad/guitar/keys bus, sidechain it from the vocal, and set just 1–2 dB of gain reduction with a fast attack and a release around 100–200 ms. When the singer sings, the bed dips imperceptibly; when they stop, it swells back. At this depth nobody hears pumping — the mix simply always has room for the voice.
Layer 5 — Space that connects, rides that finish
A totally dry vocal in a wet mix sounds pasted on — sharing space is part of sitting. Use short, pre-delayed reverb and filtered delay so the vocal inhabits the same universe as the band while staying forward (the pre-delay is what keeps it from receding).
Then the finishing layer no plugin replaces: fader automation. Ride the vocal up 0.5–1 dB in the chorus, down slightly in stripped sections, and nudge individual words the compressor missed. Twenty minutes of rides is routinely the difference between a good vocal mix and a finished one.
Frequently asked questions
Why do my vocals sound on top of the mix instead of in it?
Usually three compounding causes: no shared space (add a short pre-delayed reverb or filtered delay), inconsistent level forcing the fader too high (compress 3–6 dB), and a scooped or dull instrumental leaving the vocal tonally disconnected. Fix space first — it's the fastest win.
Why do my vocals get buried in the chorus?
The chorus adds instruments that mask 1–5 kHz. Carve 2–4 dB from the loudest sustained instruments in that range, duck them 1–2 dB via sidechain from the vocal, and automate the vocal up 0.5–1 dB in that section.
What dB level should vocals be in a mix?
There's no universal number — it depends on gain staging and genre. The workable target: the vocal is always intelligible in the busiest section without ever floating above the instrumental in the sparsest one. If no single fader position achieves both, fix dynamics and masking first.
Can a vocal preset make my vocal sit in the mix?
Partially. A well-tuned chain solves the vocal-side layers: consistency, tone, and space. Masking from other instruments and section-to-section automation live outside any vocal chain — those are mix moves only you can make.
Hear it onyour voice.
PhantomRack turns everything in this guide into a finished rack: upload a reference, download a stock-Ableton .adg tuned to it. First rack is free — no card required.