Should I boost treble to make my voice clearer?

Not automatically. Boosting high frequencies can increase hiss, room noise, harshness, and sibilance. Try reducing competing low-mid energy or improving the microphone position before adding a small, broad presence or air boost.

Where should I cut nasal voice frequencies?

Nasal character varies by voice and microphone. Sweep a narrow band at low level, identify the smallest unpleasant area, then make a modest cut, often somewhere between 700 Hz and 1.2 kHz. Do not remove the whole range by default.

TutorialsAudio

Voice EQ Settings for Content Creators

Beginner · ~35 min

Overview

EQ can make a voice easier to understand, but a preset cannot know your microphone, room, distance, voice, or audience. Use the settings below as controlled starting points, then make small changes while listening to real words. The goal is not a shiny voice. The goal is a stable, natural voice that stays intelligible on headphones, laptops, phones, and small speakers.

What You Need

A spoken-word recording with a normal performance and no clipping
An EQ with high-pass, parametric, or shelving controls. A stock DAW EQ is enough
Headphones plus at least one small speaker or phone playback check
A compressor or loudness meter for the later workflow, not as a substitute for EQ

Starting EQ Map

Area	Starting move	Listen for
Below 60-90 Hz	High-pass filter, usually 12 or 18 dB/octave	Rumble, desk knocks, traffic, plosives
120-300 Hz	Cut 1-3 dB with a broad Q only if needed	Mud, wool, proximity buildup
350-800 Hz	Small cut only when the voice sounds boxy	Cardboard, room box, hollow tone
700 Hz-1.2 kHz	Find and cut the narrow problem area gently	Nasal or honky tone
2-4.5 kHz	Broad boost of 1-3 dB, or cut a competing harsh spot	Consonant clarity and forwardness
5-9 kHz	Prefer dynamic EQ or a de-esser over a permanent boost	S, T, SH harshness and listening fatigue
10-16 kHz	Optional shelf of 0.5-2 dB after noise checks	Air, breath, hiss, room noise

Steps

Fix position and recording level first

Put the microphone close enough to favour the voice over the room, usually 10 to 20 cm for a close spoken-word sound, and aim it slightly off-axis when plosives or harsh consonants are a problem. Keep the distance consistent. Set the input gain so normal speech has headroom and loud words do not clip. EQ cannot restore a clipped waveform or remove a room that is louder than the voice.

Remove rumble with a high-pass filter

Start around 70 to 90 Hz for a typical podcast or YouTube voice and listen while the speaker says words with low vowels. Raise the cutoff until rumble and plosive energy reduce, then back it down if the voice begins to lose weight. A deeper voice, a lavalier, or a thin recording may need a lower cutoff. A noisy desk or handheld mic may need a higher one.

Use a gentle slope first. A steep filter can sound clean but unnatural if it removes too much of the lower voice. Bypass the filter at matched loudness so a louder version does not trick you into preferring it.

Cut mud and boxiness, not the personality

Make a broad, small cut in the 150 to 300 Hz region when the voice feels thick or cloudy. If the sound feels like it was recorded in a small box, test 350 to 800 Hz. Use the narrowest band only to locate the problem, then widen the Q and reduce the amount. A 1 to 3 dB cut is a safer starting point than a dramatic scoop.

Room reflections can make several bands sound wrong at once. If a large cut is required to make the voice usable, move the microphone, add absorption, or reduce the room before adding more EQ.

Find nasal tone with a focused sweep

Nasal or honky character often sits somewhere around 700 Hz to 1.2 kHz, but the exact point changes with the speaker and microphone. Temporarily boost a narrow band by a few dB while speaking a normal sentence, find the unpleasant tone, then turn that band into a modest cut. If the voice loses identity or becomes hollow, the cut is too large.

Add presence only when the words need it

For a voice that is clean but sits behind music or a video bed, try a broad 1 to 3 dB boost around 2 to 4.5 kHz. Use the smallest boost that makes consonants easier to follow. Presence is not the same as loudness: match the bypassed and processed versions before judging the change.

If presence makes the voice brittle, try cutting a competing instrument or reducing low-mid buildup first. A clear voice usually comes from space in the mix, not from continually boosting the voice.

Control sibilance dynamically

Do not use a permanent high-frequency cut to solve occasional S sounds. A de-esser or dynamic EQ can reduce 5 to 9 kHz only when the sibilance crosses a threshold, leaving the rest of the voice open. Start with 2 to 4 dB of reduction and listen for lisping. If the whole recording is harsh, inspect the microphone angle, room reflections, and compression before applying more de-essing.

Add air only after checking noise

A gentle shelf above 10 kHz can add openness to a clean recording, but it also raises hiss, computer fans, room tone, and mouth noise. Skip the air boost when the recording is noisy or already bright. Content creators often get a more trustworthy result by leaving the top end alone and improving the microphone position.

Compress and check the full voice chain

EQ changes what a compressor reacts to, so make broad corrective EQ decisions before compression, then use compression for level control. Recheck the EQ after compression because makeup gain can reveal hiss and sibilance. Finish with your loudness workflow, then listen to the whole episode rather than judging one isolated phrase.

Three Safe Starting Presets

These are not universal presets. Apply one, level-match it, and adjust by ear.

Natural podcast voice: high-pass at 75 Hz, broad -1.5 dB around 220 Hz if muddy, no presence boost until music is added, dynamic de-esser around 6.5 kHz only when needed.
Thin webcam voice: high-pass at 60 Hz, avoid cutting the low mids, try a broad +1 dB around 150-220 Hz if the recording is clean, then a restrained +1 dB around 3 kHz for intelligibility.
Bright USB microphone in an untreated room: high-pass around 80 Hz, cut the specific boxy or harsh band by 1-2 dB, skip the air shelf, and use light dynamic control above 5 kHz instead of a broad treble boost.

Pro Tips

Make cuts before boosts. Removing rumble or mud often creates clarity without making the voice aggressive.
Use bypass at matched loudness. The ear usually prefers the louder version even when it is less natural.
Do not copy a male/female EQ chart as a rule. Voice, microphone, distance, room, and delivery style matter more than gender labels.
Check voice under the actual music, intro, ads, and sound effects. A solo voice can sound perfect and still disappear in the episode.
Keep a dry or lightly processed backup. Future platform, translation, caption, or mix requirements may need a different treatment.

Knowledge Base

What EQ Can and Cannot Fix

EQ changes the balance of frequencies. It cannot remove every reflection, repair clipping, restore missing high end, replace a poor microphone position, or make a distant voice sound like a close microphone. Use it as one part of a chain that starts with the room and recording.

Listen for the problem, then name the move

Rumble: high-pass filter and better isolation.
Mud: small low-mid cut and less proximity buildup.
Boxiness: targeted mid cut plus room treatment.
Nasal tone: focused, modest cut around the actual problem.
Buried words: reduce competing music or add restrained presence.
Harsh S sounds: dynamic de-essing, microphone angle, and controlled compression.
Hiss: fix gain and noise at the source before adding air.

Where This Fits

This guide covers one specific part of mixing and colour. The wider picture. The full mix workflow from gain staging and balance through bus processing and knowing when to stop, is in Mixing a Full Song: From Rough Levels to Final Bounce, which frames the discipline as a whole and links out to the detailed guides underneath it, including this one. If you are starting from scratch rather than solving a specific problem, read that first and come back here.

FAQ

Q: What is the best EQ setting for a podcast voice?
A: Start with a gentle high-pass around 70 to 90 Hz, then make only the smallest cut or boost needed for that recording. There is no reliable voice preset that works unchanged on every microphone and room.

Q: Should I EQ before or after compression?
A: Correct broad problems before compression, then recheck after compression. You may need a small post-compressor de-esser or tone adjustment, but avoid stacking large moves.

Q: How much EQ is too much?
A: When listeners notice the processing before they notice the message, or when bypass sounds more natural at matched level, reduce the move. If you need more than a few dB across several bands, fix the recording environment.