EarEqual
Headphone calibration

EarEqual Reference Target

A technically motivated starting point for measurements made at the blocked ear-canal entrance. Version 0.1 is a preliminary universal reference, not a validated preference curve or a claim that one target is right for every listener or headphone.

v0.1 · 20 Hz to 20 kHz
Download target CSV 31-point, 1/3-octave reference · normalized to 0 dB at 1 kHz
EarEqual Reference Target v0.1 · dB re 1 kHz
20501002005001k2k5k10k20k-10-5+0+5+10 Frequency (Hz) Target level (dB)
Why a separate target?

EarEqual measures at a different acoustic reference.

Conventional headphone targets such as Harman are normally defined at the eardrum or on artificial-ear measurement systems. The Plug measures at the blocked ear-canal entrance. Those target curves therefore cannot simply be copied across unchanged.

The EarEqual Reference Target is expressed for the same kind of blocked-entrance measurement made by the Plug.

Human blocked-ear data

The acoustic backbone is the published population-average diffuse-field response measured at the blocked ear-canal entrance in real human ears. It includes average head, pinna and concha effects at that reference point.

No downstream canal target

The target does not add the entrance-to-eardrum canal transfer. When the Plug is removed for listening, your own ear canal is physically back in the playback path.

Broad listening balance

A broad low-frequency shelf and gentle spectral tilt are added as practical listening-balance terms rather than as anatomical resonances.

Construction
blocked diffuse-field average + broad bass shelf + gentle spectral tilt

For v0.1, the broad shaping is approximately a 5 dB low shelf around 105 Hz plus a −0.55 dB/octave tilt above 200 Hz. These values are a starting choice, not fitted perceptual constants.

The more individual option

Measure your own target.

If you can measure a loudspeaker or another acoustic reference with the Plug, that is the conceptually cleaner calibration. Measure the target and the headphones on the same listener, at the same reference point.

EQ = your measured target / your measured HpTF

This avoids replacing your measured blocked-ear response with a population target. The measured target can be smoothed, windowed to control room reflections, or deliberately shaped afterwards.

What v0.1 does not claim

A reference, not the final answer.

A universal blocked target still replaces individual head/pinna/concha differences with a population-average version. Also, the headphone can behave differently with the ear blocked during measurement and open during normal listening; this is most relevant for headphones that are sensitive to acoustic loading, including some closed-back designs.

Very high-frequency equalization is also sensitive to fit and placement. The 16 to 20 kHz end of v0.1 is smoothly extrapolated because the underlying blocked-ear population data end at 16 kHz.

Listening level

One target cannot sound identical at every level.

Human frequency sensitivity changes with listening level. EarEqual v0.1 is intended as a neutral starting point for normal listening levels, not as a level-independent perceptual law. At quieter playback, additional low-frequency,and sometimes high-frequency,compensation may be preferred.

For now EarEqual provides one reference curve rather than several loudness-dependent targets. Level-dependent versions can later be expressed as adjustments around the same blocked-ear reference.

References

Where the rationale comes from.

  1. Hammershøi, D. & Møller, H. (1996), Sound transmission to and within the human ear canal, JASA 100(1), 408 to 427. The work separates free-field-to-blocked-entrance transmission, pressure division, and transmission along the ear canal.
  2. ISO 11904-1, Determination of sound immission from sound sources placed close to the ear, Part 1: Technique using a microphone in a real ear (MIRE technique). The published blocked-ear diffuse-field values are used as the population reference.
  3. Harman headphone-target research provides an important comparison for broad listening preference and spectral balance, but is defined in a different measurement domain.
  4. ISO 226 equal-loudness contours motivate the explicit note that a single target is not perceptually invariant with playback level.