Customization of head-related transfer functions using principal components analysis in the time domain

Abstract

Pinna responses were separated from HRIRs (head-related impulse responses) of 45 subjects in the CIPIC HRTF (head-related transfer function) database and modeled as linear combinations of five basic temporal shapes (basis functions) by PCA (principal components analysis) accounting for more than 90% of the variance in the original pinna responses per each selected elevation angle in the median plane. By adjusting the weight of each basis function computed for a specific height to replace the pinna response in the KEMAR HRIR at the same height with the resulting pinna response and listening to the filtered stimuli over headphones, four subjects were able to create a set of median HRIRs that outperformed the KEMAR HRIRs in producing elevation effects in the median plane. Since the monoaural spectral features due to the pinna are strongly dependent on elevation instead of azimuth, similar elevation effects could also be generated at different azimuthal positions simply by inserting the customized median pinna responses into the KEMAR HRIRs at other azimuths and varying the ITD (interaural time difference) according to the direction as well as the size of the subject’s own head.