This laboratory routinely measures the HRTFs of subjects involved in our psychophysical projects. The methods have been fully described in the literature where we have also demonstrated that our recording techniques provide a very accurate record of the HRTFs (Pralong and Carlile, 1994).
The measurement procedure is delicate, requiring considerable skill and a very well controlled acoustic environment. We originally developed these techniques at the University of Oxford (U.K.) but since 1994, the laboratory has moved to the University of Sydney (Australia). The recordings we are making in our new laboratory are of an extremely high quality and in addition to measuring HRTFs we can now also measure the transfer functions of headphones with a high degree of reproducibility (Carlile and Pralong, 1995, Pralong and Carlile, 1995).
We now also test the fidelity of virtual auditory space generated using these HRTFs by comparing the localisation accuracy of each subject under normal free field conditions with their localisation accuracy for targets presented in virtual auditory space. To date we have found that localization accuracy under both conditions is almost identical. Emphasis has been placed on a good control of the signal delivered in VAS using in-ear tube phones instead of circumaural headphones and on controlling non-auditory factors in the free field and VAS environments. Our preliminary results show that under these conditions the rate of front-back confusions in VAS probably does not exceed the rate measured in the free field. A front-back confusion is where a subject correctly identifies the location of a source with respect to the midline but is confused about whether the sound is from the front or the back
It is relatively easy to take recordings from the distal portions of the outer ears of a human subject. The difficulty is knowing if they represent a good record of the HRTFs of the subject. If the user of a VAS display is interested in generating high fidelity VAS for research or in mission critical applications, then particular attention must be paid to the quality of the HRTF and to how well they match the HRTFs of the operator (Wenzel et al., 1993). This laboratory provides a measurement service where the HRTFs of individuals involved as subjects in VAS research can be recorded. As with all of our HRTF libraries the fidelity of the library is determined objectively by comparing the localisation performance of subject under both free field and VAS conditions. This kind of careful psychophysical control experiment is the only way in which the fidelity of a simulation can be guaranteed.
If you are interested in taking advantage of this service please contact Dr Simon Carlile.
Carlile and Pralong (1995) "Recording the head related transfer functions (HRTF): the essential link in high fidelity three dimensional audio displays", In Workshop on Integration of sound with visual displays. Proceedings of the 3rd International meeting of the TTCP Subgroup U, Technical Panal 7 (Air Operations), Defence Science and Technology Organisation. Department of Defence. Australia (In Press)
Pralong and Carlile (1994)"Measuring the human head-related transfer functions: construction and calibration of a miniature "in ear" recording system," J. Acoust. Soc. Am. 95, 3435-3444.
Pralong and Carlile (1995) "The role of individualized headphone calibration for the generation of high fidelity virtual auditory space." (submitted to J. Acoust. Soc. Am.)
Wenzel, E.M., Arruda, M., Kistler, D.J. and Wightman, F.L. (1993) "Localization using nonindividualized head-related transfer functions." J. Acoust. Soc. Am,. 94: 111-123.
Return to the index.