Headphone measurements are more complex than loudspeaker corresponding since there is no standard for doing the measurements.
The sum effects of the auditory canal and the pinnae shape as well as the exact microphone and headphone positioning and the measurement equipment to the headphone frequency response
makes the measuring most complex.
Formerly, the responses very obtained based on heavy psycho-acoustical (subjective) experiments, where the perceived level was compared to
a reference loudspeaker.
Today, the common method is to use a dummy head for the measurements. However, because of strong differences between different models, and
their weak representation of the real human pinnae anatomy, the dummy head solution is far from a generic one.
The underlying dilemma is
the complex phenomena that the high frequency sound (e.g., transverse wave motion above ca. 5 - 7 kHz) undergoes in the pinna cavities and the ear canal.
Hence, the headphone response is very sensitive to even minor 2 − 3 mm changes in the earpiece positioning, and between-subject variability within the same headphone
response is also substancial.
KAR headphone measurement method
Typically, the headphone closes the ear canal at the other end (the ear drum at the other), which causes an 'artifact' to the perceived sound. This so-called
ear canal resonance (longitudal) is the main reason for the the often observed high frequency dip around 7 kHz. It occurs at the half wavelength, which depends on the
actual dimensions of the individual auditory canal.
In my countless
scientific and
other studies I have
found out that in many cases the best applicability is obtained by doing the headphone measurements at the entrance of the blocked ear canal, i.e., using
external miniature microphones.
Headphone measurements made in the measurement room at the
Acoustics and Audio Signal Processing laboratory.
The headphones need to be replaced a number of times (3 - 10) on top of the dummy (real) head in order to reduce the effect of poor fitting on the head. This method
also gives an idea about the ergonomical design quality of the headphones. The power-averaged responses still need to be smoothed (e.g., 0.2 octave)
in order to mimic better the human auditory resolution.
The sensitivity of the headphone is defined as the SPL (dB lin.) at 1 mW nominal power at 1 kHz.
References
Since 1996 I have measured all the
headphones for the Finnish
HIFI Magazine, the sole
magazine specialized in audio and video.
I have also done
headphone measurements for other magazines, e.g.
- Tekniikan Maailma (largest technical oriented magazine in Finland, more than 120000 subscribers)