Comparison of direct and indirect perceptual head-related transfer function selection methods.

Area of Science:

• Acoustics and Psychoacoustics
• Virtual Audio Environments
• Human Perception

Background:

• Personalized head-related transfer functions (HRTFs) are crucial for realistic 3D audio rendering.
• In the absence of personalized HRTFs, selecting appropriate substitutes from databases is a common practice.
• Existing HRTF selection methods rely on localization cues, subjective evaluations, or anthropomorphic similarities.

Purpose of the Study:

• To investigate the comparability of head-related transfer function (HRTF) rankings derived from different selection methodologies.
• To determine if HRTF selection based on one metric predicts perceptual performance using another metric.

Main Methods:

• A perceptual study was conducted involving a basic source localization task.
• A subjective quality judgment method was employed.
• Eight head-related transfer functions (HRTFs) were evaluated, and rankings were generated using multiple metrics for each subject.

Main Results:

• A significant positive mean correlation was observed between specific metrics from the different selection methods.
• Head-related transfer functions (HRTFs) identified as optimal by one method received significantly above-average ratings in the other.
• The findings suggest a degree of convergence in HRTF selection outcomes across distinct evaluation approaches.

Conclusions:

• Different methods for selecting head-related transfer functions (HRTFs) can yield comparable and reliable rankings.
• The study validates the use of multiple metrics for HRTF selection, enhancing the robustness of substitute HRTF identification.
• This research contributes to improving the accuracy and effectiveness of spatial audio rendering in applications lacking personalized HRTFs.