CUSTOMIZABLE HEAD RELATED TRANSFER FUNCTION

V. Ralph Algazi

CIPIC, Center for Image Processing and Integrated Computing
University of California, Davis

Pierre L. Devinyi

Speech and Hearing Research Facility
V.A. Martinez Medical Center, Martinez, CA

Richard O. Duda

Department of Electrical Engineering
San Jose State University

CONTACT INFORMATION

CIPIC
University of California
Davis, CA 95616
Phone: (916) 752 8066
Fax : (408) 752 8894
Email: vralgazi@ucdavis.edu

WWW PAGE

http://info.cipic.ucdavis.edu/mspg/hearing.html

PROGRAM AREA

Other Communication Modalities

KEYWORDS

Sound localization, spatial hearing, 3-D sound, sound separation, auditory scene analysis, head-related transfer functions

PROJECT SUMMARY

In the past few years, there has been a significant increase in interest in the synthesis of three-dimensional spatial sound. In several important areas, accurately synthesized spatial sound is of great value and of growing importance: Human/computer interfaces; auditory aids for the vision impaired; virtual reality systems; acoustic displays for pilots and air-traffic controllers; teleconferencing and shared electronic workspaces and auditory display of scientific or business data. The key to generating spatial sound is the so-called Head-Related Transfer Function (HRTF). However, an individualized or custom HRTF is needed to obtain a faithful perception of spatial location. During the last decade, increasingly accurate techniques have been developed to measure individual HRTF's experimentally. The resulting HRTF's are stored as finite-impulse-response tables, indexed by a large number of different azimuth and elevation values. If head motion is taken into account, playback requires rapid interpolation between the entries in these tables. This purely empirical approach has two serious drawbacks: (a) it provides no scientific insight into the factors that control spatial hearing, and (b) it requires complex and expensive equipment for applications.

Recognizing these problems, a number of researchers have investigated ways to model the HRTF. Prior work can be summarized by noting that two basically different approaches have been used. One approach is mathematically based, typically involving the use of series expansions to approximate the measured data. The other approach is physically based, typically employing some kind of approximate solution to the wave equation.

Our research is based on the belief that the HRTF can be accurately modeled by a physically-based model employing a small number of free parameters. We also anticipate that these parameters can be adapted or customized to individual listeners by correlation with a small number of properly chosen anthropometric measurements. Based on these premises, we propose to develop and validate HRTF models using a combination of the physical and mathematical approaches. We will also perform extensive psychological experiments to validate the adequacy of the resulting models and filters. This research should provide the basis for a simple and inexpensive way to determine a computationally efficient way to synthesize individually customized three-dimensional sound.

PROJECT REFERENCES

This is a new project. We list here two papers in preparation.

V. R. Algazi, P. L. Devinyi and R. O. Duda, "Subject dependent transfer functions in spatial hearing", Proc. 1997 IEEE MWSCAS Conference. (August, 1997)

R. O. Duda and W. L. Martens, "Range-dependence of the HRTF for a spherical head," Proc. 1997 IEEE ASSP Workshop on Applications of Signal Processing to Audio and Acoustics (October, 1997).

For other references, see the related on-going project,`` A computational model for sound localization''directed by Richard O. Duda.

AREA BACKGROUND

The HRTF captures the position-dependent spectral changes that occur when a sound wave propagates from a sound source to the listener's ear drum. These spectral changes are due to diffraction of the sound wave by the torso, head, and outer ears or pinnae, and their character depends on the azimuth, elevation, and range from the listener to the source. Because the HRTF completely characterizes the acoustic information available for sound localization, it has become central to hearing-aid research, 3-D auditory computer interfaces, and any scientific study of spatial hearing (Blauert 1997).

Unfortunately, the sizes and shapes of torsos, heads and particularly the pinnae vary substantially from person to person, and thus the character of the HRTF also varies from person to person. These inter-subject variations are often quite significant, and serious localization errors (primarily front/back reversals and large elevation errors) can occur when one person hears the source through another person's HRTF.

Studies of the neural pathways from the cochlea to the auditory cortex provide inspiration for both the structure of a localization model and the kinds of signal processing that are appropriate, helping to define parameters such as filter bandwidths, response times, and compressive nonlinearities to cope with dynamic range. Studies in psychoacoustics reveal the different kinds of cues that humans use to localize sources, and human abilities to deal with echoes and reverberation. Our effort involves synthesizing this information in computational models of sound localization.

AREA REFERENCES

D. Begault, 3-D Sound for Virtual Reality and Multimedia (Academic Press, Boston, MA, 1994). An elementary but very clear presentation of 3-D audio principles and current technology.

J. Blauert, Spatial Hearing: The Psychophysics of Human Sound Localization, Revised Edition (MIT Press, Cambridge, MA , 1997). The standard reference on the psychophysics of three-dimensional hearing.

A. S. Bregman, Auditory Scene Analysis (MIT Press, Cambridge, MA, 1990). A massive description of experiments by the author and his students on the factors that influence the formation and segregation of sound streams.

S. Carlile, Virtual Auditory Space: Generation and Applications (R. G. Landes Co., Austin, TX, 1996). A lucid and valuable book of survey chapters that emphasize the physical factors that control spatial hearing.

H. L. Hawkins, T. A. McMullen, A. N. Popper and R. R. Fay, Eds., Auditory Computation, Springer-Verlag, New York, 1996. An important edited volume of papers about models of the hearing process.

J. C. Middlebrooks and D. M. Green, "Sound localization by human listeners," Annu. Rev. Psychol., Vol. 42, pp. 135-159 (1991). An excellent review of the abilities of people to localize sound. Highly recommended.

RELATED PROGRAM AREAS

Virtual Environments, Adaptive Human Interfaces, Intelligent Interactive Systems for Persons with Disabilities.

POTENTIAL RELATED PROJECTS

1. Sound localization in hearing aids. The customization of binaural hearing aids would improve spatial localization and the discrimination of sound sources by their location.

2. 3-D sound for teleconferencing. Spatial sound synthesis should allow a participant in a teleconference to place the sounds from other participants in different spatial locations, and should both improve intelligibility and reduce the cognitive load.

3. Acoustic displays for pilots and air-traffic controllers.

4. Virtual reality systems.

V. Ralph Algazi

Last modified: Tuesday Jun 17 17:16:07 PDT 1997