20 July 2026 | 6 min read
EQ for vocals and dialogue: the four moves that do 90 percent of the work
Most vocal EQ is four decisions. Learn where they live and you can make a voice clear in under a minute.
Vocal EQ has a reputation for being mysterious, but the useful moves are few and they live in predictable places. The frequencies below assume a spoken or sung human voice recorded reasonably close.
One: high pass
Roll off below 70 to 90 Hz on most voices, higher on a lav taped to a chest. There is no useful vocal content down there, only rumble, handling and HVAC. This single filter cleans up more mixes than any other.
Two: cut the mud
Boxiness lives between 200 and 400 Hz. A wide cut of 2 to 4 dB opens a voice without thinning it. Proximity effect and small rooms both pile up here, so close mics and untreated spaces usually need it.
Three: lift presence
Intelligibility sits between 2 and 5 kHz. A gentle 2 to 3 dB shelf or wide bell here brings consonants forward. Push too hard and the voice becomes tiring within a minute of listening.
Four: control sibilance
Harsh S sounds live between 5 and 9 kHz, varying by voice. A de esser is better than a static cut, because sibilance is intermittent and a permanent notch dulls everything else. Find the frequency by sweeping a narrow boost until the S stings, then de ess there.
Order and context
- Fix the source before the EQ. A different mic or a different take beats any curve.
- Cut before you boost, and check in mono.
- EQ in context with the music and effects, not soloed.
- Match the tone of neighbouring takes before polishing any single line.
Before EQ, make sure you are working from the best channel. Split the poly WAV, audition boom against lav, and export the winner at the sample rate your session needs.