Mixing vocals: clarity, presence and space
The EQ, compression and effects decisions that make a lead vocal sit clearly in a mix.
Short answer
A lead vocal should sit at the front of a mix — present, intelligible and dynamically consistent. Achieving this requires four steps in order: control the dynamics (gain staging and compression), shape the frequency balance (EQ), tame harshness (de-essing), and place the vocal in the room (reverb and delay). Automation is the final step that no plugin fully replaces. The common mistake is reaching for reverb before dynamics and EQ are stable — effects amplify problems as much as they add character.
Gain staging and clip gain
Before any processing, a vocal recording should hit your channel at a consistent level — roughly −18 to −12 dBFS average, with peaks no higher than −6 dBFS. This headroom lets your plugins work correctly and prevents clipping before the signal reaches a limiter at the master bus.
Use clip gain (also called region gain or pre-fader gain, depending on your DAW) to level-match individual phrases before the vocal reaches any plugin. A verse line delivered softly and a chorus phrase belted loudly should enter your compressor at roughly similar levels. This makes the compressor's job easier and prevents it from over-compressing loud phrases while failing to catch quiet ones.
Clip gain is a manual, one-time edit — it is not automation. Once set, it stays stable regardless of how you route the track.
EQ: subtractive first, additive second
EQ on a vocal typically follows a subtractive-first approach: remove problematic frequencies before boosting anything. Common EQ moves on a lead vocal:
- High-pass filter at 80–100 Hz: removes low-end rumble, HVAC noise and mic handling vibration. Roll off at 12 dB/octave; cut higher (up to 120 Hz) if the vocalist is close-miked and boominess is an issue.
- Cut at 200–400 Hz if boxy: a narrow cut (Q of 2–4) of 2–4 dB in this range removes the hollow, boxy quality common in room recordings.
- Cut at 400–800 Hz if nasal or congested: sweep until you hear the nasal quality peak, then cut narrowly.
- Boost at 2–5 kHz for presence: a gentle broad boost (Q of 0.5–1, 1–3 dB) in the upper midrange brings a vocal forward and improves intelligibility in the mix.
- Air shelf at 10–16 kHz: a gentle high-shelf boost (1–2 dB) adds openness and brightness. Use sparingly — too much creates harshness that requires more de-essing later.
The EQ basics guide covers frequency ranges and EQ types in detail. For reference curves and frequency analysis tools, iZotope's vocal EQ guide provides a practical frequency map for vocals.
Compression: control without squash
Compression reduces the dynamic range of the vocal so quiet lines are audible and loud lines do not overwhelm the mix. A well-compressed vocal sounds consistent without sounding processed.
Key compression settings for vocals:
- Ratio 2:1–4:1: moderate ratios control peaks without over-processing. Higher ratios (6:1+) are used for aggressive character compression, not transparent levelling.
- Attack 5–25 ms: a slightly slow attack lets the initial consonant transient through, keeping the vocal sounding natural and intelligible. Faster attack smooths out more but can make the vocal sound soft or lifeless.
- Release: set by ear — too fast creates pumping; too slow holds down the signal between phrases. Many engineers set release to roughly match the phrase rhythm, or use an auto-release mode.
- Threshold: aim for 4–8 dB of gain reduction on the loudest parts of the vocal. Make up the lost gain with the output or make-up gain knob.
Stacking two compressors — a transparent compressor (VCA or FET style) for peak control followed by a slower optical or tube-style compressor for character — is a standard technique in professional vocal chains. The first does the heavy work; the second adds glue and warmth.
Waves' guide to vocal compression demonstrates ratio, attack and release settings with audio examples. The compression guide covers the underlying principles.
De-essing and harshness
Sibilance — the piercing quality of "s", "sh" and "ch" consonants — becomes more noticeable after compression, which raises quieter sibilant sounds closer to the level of sustained vowels. A de-esser is a frequency-selective compressor that dynamically reduces the 5–10 kHz range only when sibilance exceeds a threshold.
To set a de-esser: find the most problematic sibilant frequency using a solo'd band sweep. Most sibilance peaks between 6 and 9 kHz, though it varies by vocalist and microphone. Set the threshold so the de-esser activates only on the harshest consonants, not continuously. Over-de-essing creates a lispy, dull quality on all "s" sounds.
If harshness persists after de-essing, check whether the issue is in the microphone choice, the recording space, or whether the EQ air shelf is boosting too aggressively. A dynamic EQ (frequency-selective compression triggered only when a band exceeds a threshold) gives finer control than a standard de-esser for complex harshness issues.
Reverb, delay and space
Effects place the vocal in an acoustic space and create a sense of depth. The order of operations matters: effects should be applied to a vocal that is already dynamically controlled and EQ'd — not as a way to hide problems.
- Short room reverb (0.3–0.8 s RT60): adds subtle space without pushing the vocal back in the mix. Useful on almost every vocal.
- Plate reverb: a classic vocal treatment — dense, bright, with a distinctive metallic quality. Common in pop, soul and rock productions.
- Longer hall or chamber reverb: creates depth and drama; best used on sustained notes or as a tail effect on longer phrases. Use pre-delay (20–50 ms) to keep the dry vocal present.
- Slap delay (60–120 ms): a single echo that thickens the vocal without obvious repetition. Common in country and rockabilly styles; also used subtly in pop.
- Tempo-synced delay: delay time set to a musical note value (eighth note, quarter note) so the echoes land on the beat. Use the note-to-frequency tool alongside a tempo/delay calculator to find the right delay time in milliseconds for your BPM. The reverb and space guide covers how to blend effects for depth.
Keep reverb and delay on send/return buses rather than inserted directly on the channel. Route multiple vocals (lead, backing) to the same reverb return for cohesion. Attack Magazine's vocal production features include detailed walkthroughs of effects routing for different genres.
Automation and ride
Automation — drawing volume changes on the vocal track over time — is the step that separates a flat, compressed vocal from one that feels performed. Even the best compressor chain leaves 4–8 dB of dynamic variation between a quiet verse line and a belted chorus. Automation handles this manually and intentionally.
Ride the vocal fader (or use volume automation lanes in your DAW) so every lyric is audible and intentional. Common automation moves:
- Bring up quiet phrases at the end of lines (where vocalists tail off).
- Pull back the first syllable of a phrase that enters too loud.
- Automate a slight level reduction on a doubled vocal during verses so it sits under the lead without disappearing.
- Ride backing vocals down in the chorus where the lead vocal needs to cut through.
Next step
Vocal mixing draws on the same principles as general mix engineering. The EQ basics guide and compression guide give the conceptual foundation for everything described here. For effects routing and depth, see the reverb and space guide. If the vocal is competing with other instruments, the problem may be in the arrangement — the arranging for small band guide explains frequency separation before the mix stage. For deeper production technique, Sound On Sound's vocal recording and mixing archive is one of the most comprehensive free resources available, with genre-specific walkthroughs. iZotope\u2019s vocal mixing guide covers spectral processing and dynamic EQ in depth.
Frequently asked questions
Thin vocals usually lack energy in the 200–400 Hz range (body) or 2–5 kHz range (presence). Boxy or nasal vocals often have a buildup around 400–800 Hz. Use a narrow parametric EQ cut in sweep mode — boost the band, move it slowly until you hear the problem frequency, then cut 2–4 dB. A high-pass filter (around 80–100 Hz) removes low-end rumble that adds muddiness without contributing musical information.
A ratio of 2:1 to 4:1 is typical for vocals. Attack time affects transients: a slower attack (10–30 ms) lets the initial consonant through before compression engages, preserving intelligibility. Release time should roughly match the rhythm of the vocal phrase. Aim for 4–8 dB of gain reduction on louder phrases — enough to control dynamics without pumping or artifacts. Stack a second compressor at lower settings rather than pushing one compressor hard.
A de-esser is a frequency-selective compressor that targets the 5–10 kHz range where sibilance ("s", "sh", "ch" sounds) concentrates. When a vocal has harsh, piercing "s" sounds that become more noticeable after compression, a de-esser reduces that specific range dynamically without affecting the rest of the vocal. Most DAWs include a de-esser plugin. Set the frequency to match the harshest sibilant peak and adjust the threshold until the harsh edge is reduced but the vocal still sounds natural.
Almost always use a send (aux return) rather than inserting reverb directly on the vocal channel. A send lets you blend the wet reverb signal with the dry vocal, adjust the return level independently, and share one reverb instance across multiple instruments for a cohesive sense of space. Use pre-delay (20–50 ms) on the reverb to keep the dry vocal present before the reverb tail arrives — this preserves clarity while still adding depth.
Vocal automation (also called "riding the vocal") means manually adjusting the fader or volume of a vocal track over time so quieter phrases match the level of louder ones. Even after compression, a lead vocal can have 6–10 dB of dynamic range between its softest and loudest lines. Automation fills in what compression cannot without over-compressing. Draw volume automation, or use clip gain handles, to bring up quiet lines and pull back occasional peaks so the vocal sits consistently in the mix.
Clashing usually means the vocal and another instrument share the same frequency range without separation. The most common conflict is between the lead vocal (presence at 2–5 kHz) and rhythm guitars or keys in the same zone. Use sidechain EQ (dynamic EQ or multiband sidechain) to duck the competing instrument slightly in the vocal frequency range only when the vocal is present — this is called "frequency ducking" or "spectral sidechain". Alternatively, EQ the rhythm guitar with a gentle cut at the vocal presence frequency. See the <a href="/guides/production/eq-basics" target="_blank" rel="noopener">EQ basics guide</a> for the underlying principles.