Finite difference computation of head-related transfer function for human hearing.

Area of Science:

• Acoustics
• Computational Physics
• Signal Processing

Background:

• Head-related transfer function (HRTF) modeling is crucial for spatial audio.
• Understanding HRTF variations due to head geometry and environment is essential.
• Existing models may lack the detail to capture complex acoustic interactions.

Purpose of the Study:

• To develop and validate a 3D finite-difference time domain (FDTD) model for HRTF simulation.
• To investigate the influence of head geometry and environmental factors on HRTF.
• To provide a computational tool for advancing spatial audio research.

Main Methods:

• Development of a 3D FDTD numerical model simulating acoustic wave propagation.
• Implementation of perfectly matched layers (PML) for boundary absorption.
• Utilization of an external source with scattered-field/total-field formulation to optimize computation.
• Validation against analytical solutions for a spherical head model.

Main Results:

• The 3D FDTD model accurately simulates acoustic wave interactions with head models.
• Validation against analytical solutions confirms the model's reliability.
• Simulations show HRTF differences between spherical and realistic head models up to 7 kHz.
• The model predicts HRTF changes for a realistic head near a wall.

Conclusions:

• The developed 3D FDTD model is a validated and effective tool for HRTF prediction.
• This model can accurately represent the effects of head geometry and environmental acoustics.
• The tool holds significant potential for enhancing spatial audio applications and research.