EQBase

Headphone EQ

Headphone EQ settings for Mac

Last updated: October 2, 2026

Every pair of headphones has a measured frequency response, and almost none of them are flat. A correction curve is the set of filters that pulls a specific model toward a neutral target. The pages below carry the published AutoEQ correction for each model: the curve, the exact parametric filters, and a one-click preset for EQBase.

EQBase ships 8,666 corrections for 6,033 headphone models in the app itself, so if yours is not listed here yet, search for it in the Headphones pane after installing. These pages are the ones written out in full.

Wireless and consumer

Studio and audiophile

In-ear monitors

The Headphones pane on macOS, with a model searched by name and its measured AutoEQ correction curve loaded.
Every curve on these pages is also in the app: search the model by name and the same correction loads, no file to download.

How to read a correction curve

Each page shows the correction as a curve, as a table of parametric filters, and as a ten-band version. They are three views of the same thing. The curve is the fastest to read: everything above the centre line is being boosted, everything below it is being cut, and the height tells you by how much. A dip at 3 kHz means the headphone measured loud there and the correction is pulling it back down.

The filter table is the same information in the form an equalizer accepts: a type, a centre frequency, a gain and a Q, which is how wide the filter reaches either side of its centre. A high Q is a narrow, surgical move; a low Q is a broad tonal one. The preamp line at the top is not a tone control at all, it is headroom: boosting frequencies in a signal that already peaks near full scale would clip it, so the whole curve is pulled down first.

Ten-band or parametric EQ for headphone correction?

A parametric equalizer can put each filter at any frequency, gain and Q, so it reproduces an AutoEQ correction exactly. A ten-band graphic equalizer has fixed bands an octave apart, from 31 Hz to 16 kHz, so it can only approximate one. How closely is something the catalog can answer: across all 8,666 corrections EQBase ships, the largest gap between the ten-band version and the parametric curve, from 40 Hz to 10 kHz, has a median of 3.3 dB. 42% of them stay within 3 dB and 93% within 6 dB.

The gaps are not spread evenly. For 46% of corrections the largest one falls between the 4 kHz and 8 kHz bands, where 87% of the catalog's peaking filters in that range are narrower than a band, and for 28% it falls below 31 Hz, where a low shelf keeps going under the lowest band. Each model page gives its own figure and says where its gap sits. Use the parametric filters when your equalizer accepts them. The ten-band version is for fixed-slider equalizers, EQBase's free tier among them, and EQBase Pro applies the parametric filters directly.

Why two measurements of one headphone disagree

Most models here carry several published measurements, and they rarely match exactly. That is not sloppiness. A headphone measured on a different coupler, with a different seal, on a different unit off the production line, genuinely measures differently, and the gap is usually widest above about 3 kHz where the shape of the ear canal starts to dominate. Each page names the spread on its own data rather than hiding it behind an average, because averaging other people's measurements together would produce a curve nobody actually measured.

How do I select the correct measurement and model variant for a headphone preset?

Match the variant first, then the measurement, then listen. A correction is a very good starting point, not a verdict, and these steps rule out the biggest source of error first.

  1. Find your exact variant. Where a model page lists variants, they are separate products as far as the measurement is concerned: a production revision, worn or replaced earpads, an impedance option, an ANC mode, a tuning nozzle. Pick the label that describes your own pair. Standard means the measurement was published without a revision or condition noted.
  2. Start with the measurement the picker opens on. Within each variant the measurements are listed in a fixed order, oratory1990 first, then crinacle, Rtings and the other rigs. Each one names who made it.
  3. Check where they disagree. Each model page says which band its measurements disagree on most and whether the gap is between the variants or between the rigs. A gap between variants means the variant matters more than the rig. A wide gap between rigs measuring the same variant usually comes down to the seal or the coupler, so prefer the rig whose conditions match how you wear them.
  4. Listen for a few days before switching. Ears adjust to a new balance. If the treble or the bass still sounds wrong after that, try a second measurement; in EQBase's Headphones pane every measurement of a model is listed, so switching is one click.
  5. Fix the fit before the curve. Re-seat ear tips, check the pads and adjust the headband first. A seal that leaks loses bass no correction can put back.

How to use these

macOS has no system-wide equalizer, so a correction entered in the Music app affects Music and nothing else. EQBase processes all system audio instead, which means one correction covers every app. Pick your model, download the preset, and double-click it. The free tier includes headphone correction with no time limit.

The presets carry the ten-band version of each correction, which is what the free tier applies, so what you hear matches the ten-band table on the page. If you would rather not download anything, every page lists its filters as copy-paste text in AutoEQ's own format, which most desktop parametric equalizers import on any platform.

Download EQBase free

What alternatives to manual headphone EQ provide downloadable measurement-based presets?

Instead of tuning by ear, you can start from a correction someone computed from a measurement. The sources differ in who measured the headphone and in what you get back.

Which Mac apps apply these to all system audio, and which import AutoEQ files, are compared in headphone correction software for Mac.

How these pages are checked

A page like this is only worth reading if the numbers on it are the numbers the app actually plays, so that is tested rather than assumed. The ten-band figures and the parametric filters shown here are produced by a Python port of EQBase's own DSP, and a build step compiles the shipping Swift equalizer into a reference binary, runs both over the catalog, and fails if any band or preamp disagrees by more than 1e-9 dB. In other words the published curve is identical to the app's to double-rounding, not merely close.

Every downloadable preset is then run through the app's own decoder and re-encoded. That check exists because EQBase decodes presets tolerantly: a wrong key or a bad filter name would not raise an error, it would silently default, and you would get a flat curve instead of your correction. A preset only ships if it decodes, applies, is not bypassed, is not flat, and survives the round trip.

If your headphones are not the problem

A headphone correction fixes the headphone. It cannot fix a room: if the bass changes depending on where you sit, or a speaker setup sounds different from the same track on headphones, that is the room talking. See room correction on a Mac, which measures the space with a sweep and corrects what no headphone curve reaches. If the issue is one app being louder than another rather than tone, see per-app volume control on a Mac.

Credits

Correction data from the AutoEQ project by Jaakko Pasanen, MIT licensed, with each measurement credited to the person or lab that made it.