Spatial hearing is worst near 1.7 kHz where head wavelength matches.
Evidence
One method to evaluate the localization resolution is to measure the JND of direction for different frequencies of tones. Figure 12.16 plots the results of such an experiment using sinewave test signals (Stevens and Newman, 1936). The best resolution is about 1◦for tones below 1 kHz coming from the front of the subject, and it deteriorates towards the sides. At around 1.7 kHz the resolution is the worst, especially for sounds from the sides, so much so that the subjects cannot make any reliable estimate at all. This is the frequency where the ILD does not depend monotonically on direction and the head size is about one wavelength, leading to maximal confusion in the ITD and ILD decoding. At higher frequencies, the resolution improves, but is never as good as for frequencies below 1 kHz. Note that the JND values do not reflect the localization accuracy, but only how much a source has to be moved for the change to be noticed. — Ville Pulkki, Matti Karjalainen, Communication Acoustics- An…, 1472
Categorization
- QUESTION: The 1.7 kHz blind spot in horizontal localization — could have design implications for spatial audio or wearable alarm placement worth chasing.
- XREF: Connects to head-related transfer function design in spatial audio rendering. — connects to: Directional hearing resolution is worst near 1.7 kHz where head size equals one wavelength