Abschlussarbeit

From Monaural to Binaural: Developing a Perceptual Distance Metric for HRTF Evaluation

Head-related transfer functions (HRTFs) allow us to create virtual sound sources. The functions imitate the interaction of sound waves with the listener’s body, as the waves arrive at the left and right ear canal entrance, respectively. Evaluating whether a generic HRTF dataset is suitable for a specific listener, or whether an approximation of their individual HRTF set is sufficiently accurate, remains a current research question. While there are many possible metrics that quantify HRTF differences objectively, they mostly have not been properly set in relation with auditory perception. For this purpose, a listening experiment was conducted, relating several numerically calculated values to just noticeable differences between given stimulus pairs. Serving as a first step, this experiment only involved signal variation at one ear, while the other ear received an unchanged stimulus. The goal of the present thesis is to extend the experiment to changes to both ear signals, thereby developing a binaural distance metric. The work involves analyzing HRTF data (based on the previously performed study), choosing and creating suitable audio stimuli, finally conducting the listening experiment with participants, and analyzing the results.

Requirements:
Basic knowledge in acoustics and digital signal processing
Interest to design an experiment and work with participants
MATLAB skills