Proceedings of 18th International Congress on Acoustics ICA 2004, Kyoto, Japan; 4.-10.4.2004.

Part Va: Effect of head movements on measured head-related transfer functions


Klaus A J Riederer


Laboratory of Acoustics and Audio Signal Processing
Helsinki University of Technology (HUT)
P.O. Box 3000, FIN-02015 HUT, Finland
Tel: +358 9 451 2494; Fax +358 9 460 224
Email: Klaus.Riederer@hut.fi, URL: http://www.hut.fi/~kar




Three head movement types are common in real head-related transfer function (HRTF) measurements: e.g., tilting occurring persons with dystonia, pivoting (“nodding”) happening with sleepy subjects and rotation taking place with shaky/elderly people. Third part of the present study [Riederer, Proc. 18th Intern. Congr. Acoust. ICA 2004] considers varying head positions between silicone-blocked HRTF measurements.

A 25-year-old male was first measured at 10°-azimuth-intervals involving a small constant head tilt and then re-measured with straight head pose at 40°-azimuth-intervals. Difference-plots of 1/10-octave-smoothed system-compensated HRTFs show even a minor ~5º tilt causes moderate, ca. ±10 dB magnitude variation in HRTF measurements at ipsilateral sound incidents in all elevations (d) >6-7 kHz. The variation is less prominent at the ear to which head is tilted.

The effect of major head pivoting is compared with repetition measurements on another male. A strong head pivot (>20°) alters semi-strongly (±5..15 dB) the idiosyncratic high peaks and dips above 6 kHz, where it also interacts with varying microphone position at the blocked cavum conchae. This interaction makes these effects somewhat difficult to separate; yet the pivoting causes less smooth shifting in frequency and not affecting d=90°. Both pivoting and varying capsule position are distinguishable from altered auditory canal blockage and hairstyle.

The effect of horizontal head rotation is simply revealed by viewing the difference between dummy head measurements and the same data offset a constant q. A simulated 10° rotation shows moderate variation (±10 dB), smooth shifting in frequency of HRTF characteristics, above 2 kHz at ipsilateral (q » qright,left ±40°) sound directions d<90°. Stronger 20° rotation causes more powerful variations (±15..20 dB) >1 kHz at wider directions (d < 90°, q » qright,left ±150°. The head rotation effect resembles alterations caused by different hair/styles on HRTF characteristics.



[Work supported by Graduate School of Electronics, Telecommunication and Automation.]

Keywords: head-related transfer function (HRTF), three-dimensional (3-D) sound, azimuth, elevation, binaural cue, monaural spectral cue, acoustic measurement, repeatability, quality, cavum conchae, pinna, pinna cues, eigenfrequency, natural resonance, idiosyncratic response, dummy head, real head