Tutorials Audio Engineering

EQ Cheat Sheet: Best EQ Settings for Vocals

By Infinite Creation · Intermediate · ~20 min

Some links on this page are Amazon affiliate links. If you buy through them, we earn a small commission at no extra cost to you.

A 1960s mid-century comic book style editorial illustration showing two audio engineers at a mixing desk, with a pair of studio monitors, a phone clamped to a rig, a colour grading panel with trackballs visible in the frame.

Overview

EQ is the most powerful tool in vocal processing, and the most abused. This guide gives you specific frequency targets and the reasoning behind them, so you can make decisions by ear rather than guessing. The golden rule: always cut before you boost, and cut with a narrow Q, boost with a wide one.

What You Need

Any parametric EQ plugin (FabFilter Pro-Q 4, TDR Nova free, or your DAW's built-in EQ)
A vocal recording on a separate track
Headphones or studio monitors for accurate listening

Steps

High-pass filter: remove everything below 80Hz

Below 80Hz a vocal recording mostly picks up rumble, air conditioning, mic handling noise and floor vibration rather than voice. Apply a 12dB/octave high-pass filter at 80Hz (some engineers go to 100Hz for a cleaner result). The filter does not stop dead at its setting: at 80Hz it is already 1.5dB down at 100Hz, as the test below shows. A deep voice can have useful energy close to 80Hz, so on a bass singer or a low speaking voice, raise the filter by ear and stop before the voice thins. This frees up headroom and reduces muddiness before you've done anything else. Use it on every vocal.

Cut muddiness at 200 to 400Hz

This is the "boxy" region. Room resonances, proximity effect, and certain condenser mics all pile up energy here. Boost a narrow bell, sweep it around 200 to 450Hz until you hear it get unpleasantly boxy, then cut 3 to 6dB at that frequency. The result is a voice that sounds cleaner and sits better in a mix. This cut is almost always needed on home recordings.

Add warmth at 150 to 200Hz carefully

If the voice sounds thin after the mud cut, a gentle +2dB wide shelf at 150 to 200Hz restores body without reintroducing boxiness. Use a wide Q (0.5 to 0.7) for this. You're adding a colour, not a correction. Don't do this if the voice already sounds full.

Cut harshness at 1 to 3kHz

The 1 to 3kHz range is where ear fatigue lives. Over-compressed voices, certain condenser mics, and sibilance often have a peak here that makes them tiring to listen to. Sweep a narrow bell until you find the harsh peak, then cut 2 to 4dB. A narrow dynamic EQ (one that only cuts when the signal gets loud) works better here than a static cut.

Boost presence at 3 to 5kHz

The 3 to 5kHz range is the presence zone: it adds intelligibility, forward energy, and the sense that the voice is "in the room." A gentle +2 to 3dB wide boost here makes a voice cut through a mix without needing to raise the fader. Don't overdo it, above 4dB it starts sounding harsh again.

Add air at 10 to 16kHz

A gentle high-shelf boost above 10kHz adds air. The sense of space and openness that makes a recording sound professional rather than demo-quality. Use a wide shelf, +2 to 4dB. This works best on condenser microphones that have natural high-frequency response. On dynamic mics, this range may be absent anyway.

Tuning These Settings for Spoken Voice (Podcasts, YouTube, Voiceover)

The six moves above apply to any voice, but spoken-word content has different priorities than sung vocals. It's usually mono, recorded close on a dynamic broadcast mic, and consumed on phone speakers and earbuds, so intelligibility matters more than lushness. Adjust the cheat sheet like this:

Lean on presence (3 to 5kHz), not air. Most listeners hear you on phone speakers or cheap earbuds that barely reproduce 10kHz and up. A +2 to 3dB presence nudge does far more for clarity than a high-shelf air boost, which on a dynamic mic like an SM7B or PodMic is largely empty anyway.
Expect more low-mid mud from close-miking. Podcast and voiceover setups sit close to the mic, so proximity effect piles up energy at 100 to 300Hz. High-pass a touch higher (100Hz for spoken voice) and be more aggressive with the 200 to 400Hz mud cut than you would on a sung take.
Favour dynamic EQ over static cuts. A talk track runs for an hour and your tone drifts as you lean in and out. A dynamic cut at 1 to 3kHz for harshness and a de-esser at 5 to 8kHz ride those changes far better than fixed bands.
Get it consistent before it's pretty. For dialogue, even and fatigue-free beats characterful. Keep boosts gentle and wide, and save the aggressive tone-shaping for music.
EQ is step one. Loudness is step two. After EQ, set your levels to the platform target, see the Podcast Loudness (LUFS) Cheat Sheet. For one-click cleanup of noisy or uneven voice, Adobe Podcast Enhance is worth a pass first.

Tested: what the six moves add up to

We tested the six steps on 2026-09-25 with FFmpeg 8.1.2. We set each move to the middle of the range its step gives, using FFmpeg's own EQ filters: a 12 dB per octave high-pass at 80 Hz, a 4 dB cut at 300 Hz, a 2 dB low shelf at 175 Hz, a 3 dB cut at 2 kHz, a 2.5 dB boost at 4 kHz and a 3 dB high shelf at 10 kHz. Narrow bands had a Q of 4, our choice, since the page gives no figure for them. Wide ones had 0.6 or 0.7, from step 3's range. We passed a single-sample click through each move alone and through all six in order, and read the frequency response with an FFT.

Frequency	Move aimed here	That move alone	All six together
60 Hz	High-pass at 80 Hz, 12 dB per octave	-6.2 dB	-4.3 dB
80 Hz	High-pass at 80 Hz, 12 dB per octave	-3.0 dB	-1.2 dB
100 Hz	High-pass at 80 Hz, 12 dB per octave	-1.5 dB	+0.1 dB
175 Hz	Warmth shelf, +2 dB at 175 Hz, Q 0.6	+1.0 dB	+0.6 dB
300 Hz	Mud cut, -4 dB at 300 Hz, Q 4	-4.0 dB	-3.6 dB
2 kHz	Harshness cut, -3 dB at 2 kHz, Q 4	-3.0 dB	-1.8 dB
4 kHz	Presence boost, +2.5 dB at 4 kHz, Q 0.7	+2.5 dB	+2.7 dB
10 kHz	Air shelf, +3 dB at 10 kHz	+1.5 dB	+2.1 dB
16 kHz	Air shelf, +3 dB at 10 kHz	+2.5 dB	+2.6 dB

The bells did exactly what their settings say at their own frequency. The shelves did not. Each gave half its gain at its set frequency and came near the full gain only well past it. The 2 dB warmth shelf added 1.0 dB at 175 Hz. The 3 dB air shelf added 1.5 dB at 10 kHz and 2.5 dB at 16 kHz.

Together, the moves undo parts of each other. The warmth shelf keeps rising below 175 Hz, so it lifts the rumble step 1 took out. The high-pass alone was 3.0 dB down at 80 Hz and 6.2 dB down at 60 Hz. With all six moves on, those became 1.2 and 4.3 dB, and at 100 Hz the chain read +0.1 dB, with no cut left at all. A bell in place of the shelf, at the same gain and Q, kept the high-pass 2.1 dB down at 80 Hz and gave +1.6 dB of warmth at 175 Hz, against +0.6 dB from the shelf.

The same happens higher up. The wide presence boost still added 1.2 dB at 2 kHz, so the 3 dB harshness cut there came out at 1.8 dB. Where two moves overlap, the result is their sum. So set them with all six switched on, and if your EQ draws the combined curve, read that rather than each band's gain figure.

The six moves as one FFmpeg filter, to try on a recording: ffmpeg -i vocal.wav -af "highpass=f=80, equalizer=f=300:t=q:w=4:g=-4, lowshelf=f=175:t=q:w=0.6:g=2, equalizer=f=2000:t=q:w=4:g=-3, equalizer=f=4000:t=q:w=0.7:g=2.5, highshelf=f=10000:g=3" vocal-eq.wav

Validated. Measured 2026-09-25 with FFmpeg 8.1.2 and numpy on macOS 26.6.2. These are the filters' responses, not a listening test, and we did not process a real vocal. Another EQ's bands may not have exactly these shapes at the same Q setting.

Pro Tips

EQ in context, always make your adjustments with the full mix playing, not soloed. What sounds thin in solo sounds right in a mix.
Use our EQ Frequency Reference to hear what different frequency bands sound like and diagnose common problems.
A de-esser is not a substitute for a well-placed sibilance cut. Use it after EQ, not instead of it.

Knowledge Base

What You'll Learn

EQ vocals to sit perfectly in your mix. Learn which frequencies to cut, which to boost, and how to make vocals clear without sounding harsh. Applicable to singing, speaking, and voiceover.

Vocal Frequency Map

Below 80 Hz: Rumble, handling noise and air conditioning. Removed by the high-pass in step 1.
150-200 Hz: Warmth and body. A gentle, wide boost of about 2 dB if the voice sounds thin after the mud cut (step 3). Too much sounds boomy.
200-400 Hz: Boxiness and mud. Usually cut 3 to 6 dB at the worst frequency (step 2).
1-3 kHz: Harshness and ear fatigue. Cut 2 to 4 dB with a narrow band at the peak you find (step 4).
3-5 kHz: Presence and intelligibility. A wide boost of 2 to 3 dB. Above 4 dB it turns harsh (step 5).
10-16 kHz: Air. A wide high shelf of 2 to 4 dB, mostly worth it on condenser mics (step 6).

EQ Workflow

1) High-pass at 80 Hz, or 100 Hz for spoken voice. 2) Find the mud between 200 and 400 Hz by boosting a narrow band about 10 dB and sweeping until it sounds boxy, then cut 3 to 6 dB there. 3) If the voice now sounds thin, add about 2 dB of wide warmth at 150 to 200 Hz. 4) Sweep 1 to 3 kHz the same way for a harsh peak and cut it 2 to 4 dB with a narrow band. 5) Add 2 to 3 dB of wide presence at 3 to 5 kHz. 6) Add air with a 2 to 4 dB high shelf above 10 kHz if the mic captures it. Overlapping bands add up, as the test above shows, so make the final adjustments with every move switched on.

Common Vocal EQ Moves

Sibilance (harsh S/Z sounds): Reduce 5-10 kHz with narrow Q.
Nasal tone: Reduce 800 Hz-2 kHz with narrow Q.
Thin voice: Boost 150-200 Hz gently (about 2 dB, wide), below the 200-400 Hz region cut in step 2.
Dark, muffled: Start with the presence boost at 3-5 kHz (step 5), then add air above 10 kHz (step 6).

Where This Fits

Vocal EQ is one channel's decision, and Mixing a Full Song shows how channel decisions add up to a finished mix.

FAQ

Q: Should I compress before or after EQ?
A: Personal preference. Compress first (control dynamics before shaping tone) or EQ first (shape tone on dynamic content). Experiment.

Q: What EQ settings work best for a podcast or YouTube voice?
A: Start with the six moves in this guide, then prioritize for spoken word: high-pass at 100 Hz, a firm mud cut around 200-400 Hz (close-miking exaggerates it), and a gentle +2-3 dB presence boost at 3-5 kHz so you stay intelligible on phone speakers and earbuds. Go easy on the 10 kHz+ air band. Most dynamic podcast mics roll off up there and most listeners can't hear it anyway.

Q: Do I need different EQ for a dynamic broadcast mic vs a condenser?
A: Yes. Dynamic mics (Shure SM7B, RODE PodMic) roll off the highs, so the air boost above 10 kHz does little, spend that energy on presence at 3-5 kHz instead, and expect strong proximity-effect buildup to tame at 100-300 Hz. Condensers capture more top end, so the air boost pays off, but they also pick up more room and sibilance, so a de-esser and the 1-3 kHz harshness cut matter more.