Auditory Consistency Of Head-Related Transfer Functions Of Kemar From Different Databases
Xiaoli Zhong,Xiu Xu,Bosun Xie
2016-01-01
Abstract:Head-related transfer functions (HRTFs), which reflect the interaction between free-field sound waves and human anatomical structures (especially the head and ears), are the core of virtual auditory display and relevant applications. Due to individuality of human anatomical structures, HRTFs are individual-dependent. On the basis of different populations, some human-subject HRTF databases have been constructed. If the existing databases can be combined, then the research of individual HRTFs will be greatly promoted. However, different databases used different measurement protocols including measuring methods, equipment, and post-processing algorithms. So, the influence of different measurement protocols needs to be evaluated before the combination of multiple databases. To this end, this study used HRTFs of the same subject KEMAR (Knowles Electronics Mannequin for Acoustics Research) to isolate the influence of different measurements from subject's individuality. On the basis of KEMAR HRTFs from five databases, the auditory consistency of KEMAR HRTFs from different databases was evaluated using objective index as well as subjective experiments. First, directional transfer functions (DTFs) were extracted from the original HRTFs; then spectral deviations among the DTFs were calculated. Finally, a subjective experiment was conducted. Results show that the average spectral deviations are around 3 dB among different HRTF databases; Moreover, although there are certain timbre differences among some databases, the five KEMAR HRTF databases are basically equivalent in terms of localization performance below 12kHz.