Efficient Computational Techniques for Evaluating Distance-Dependent Head-Related Transfer Functions

This work proposes and validates two computational tools for synthesizing distance-dependent head-related transfer function (HRTF), which is vital in spatial sound reproduction. HRTF is an anthropometric feature-dependent function that yields the direction-dependent gain of the auditory system. Even though it is subject to the distance of the auditory source, distance-dependent HRTF measurement is rare due to its high experimental cost. Numerical simulation tools can provide viable alternatives. The required computational resources and time increase exponentially with the frequencies and degree of freedom (DoF) of the simulations; still, it is faster than experimental procedures. This work proposes finite element computational solutions to measure distance-dependent HRTFs using domain truncation methods in association with frequency-dependent adaptive meshing. Two hybrid techniques to find HRTF in the entire region, employing infinite elements (IEs) and non-reflective boundary conditions (NRBCs) with near-field to far-field transformation techniques, have been implemented and analyzed. The proposed methods calculate distance-dependent HRTF in 0.2–20 kHz frequency band, with reduced computational cost and time. Additionally, the spatial resolution of the HRTF measurement has increased a 100-fold. Since locally connected finite elements are used, the near-field effects of HRTF are well incorporated, and the obtained HRTF matches well with the experimental results. The proposed tools can also calculate sufficiently accurate HRTFs even when the surface meshes are of reduced quality. The tools also possess the versatility in effortlessly integrating appropriate bioacoustic attributes (e.g., internal reflection of the middle ear walls) into HRTF numerical models, which is noteworthy.