Examining the influence of controlled HRTF-variation on various distance metrics
* Presenting author
Abstract:
Binaural hearing is a complex process that is significantly influenced by physical attributes of the listener. These dependencies on the shape of the torso, head, and pinna can be summarized in digital filters, so-called head-related transfer functions (HRTFs). Highly individual HRTFs are crucial for a realistic listening experience. However, in many applications like listening experiments or virtual acoustics, non-individual HRTFs are used due to the great effort the measurement of an individual filter entails. Depending on the degree of dissimilarity between this generic and the individual filter, the plausability of the acoustic scene can be negatively impacted.To predict errors or to justify HRTF deviations, meaningful similarity or distance criteria are required. Due to their dependency on direction and frequency, comparing two HRTFs is not an intuitive task. Although numerous distance criterions have been developed, there have been few comparative studies on their attributes.In this presentation, the behaviour and characteristics of well-established distance metrics are examined regarding their range of values, as well as spatial and frequency distribution. An investigation of very subtle differences is enabled by controlled HRTF variation. Two methods used in that matter are the principal component analysis (PCA), as well as frequency scaling.