Professor Ken Frampton, foreground, works with Professor Wes Grantham to fine-tune KEMAR, a mannequin used to understand better how humans perceive the location of sound.

Just the thought of having to cross a busy traffic intersection blindfolded might understandably bring on a panic attack in most people, but it's a challenge the blind have to tackle regularly.
       The blind rely on a system of auditory cues often used in important traffic intersections, but these systems can sometimes be as confusing as they are helpful.
       The Vanderbilt School of Engineering is working with a team of universities to develop better auditory systems to guide blind pedestrians and train them to use the new intersection systems safely. The Blind Pedestrian Access to Complex Intersections project, sponsored by the National Institutes of Health, will incorporate Vanderbilt research to develop an acoustic virtual reality system that will not only deepen understanding of how people perceive sound but will interactively teach the blind to use their perception of sound to fine-tune their location as they navigate intersections.
       "There is a great deal of individual variation in how people perceive sound," says Vanderbilt acoustics specialist Kenneth D. Frampton, assistant professor of mechanical engineering. "Our technical challenge will be to develop the computer program that can create a real-time acoustic virtual reality environment that is instantaneously interactive."
       When finished, the system will allow people to practice navigating an intersection using an auditory guidance system. They will wear a headset connected to a computer that will generate sounds in response to the position and motions of their heads.
       The first step in developing this interactive system is to analyze how sound is perceived from different directions, depending on the position of the head.
       For this research, Professor Frampton and his team of psychologists associated with the Vanderbilt Wilkerson Hearing Institute are enlisting the help of KEMAR, a microphoned mannequin. With sensitive microphones in each ear that simulate the performance of the human ear, KEMAR transmits sound data to a computer system that analyzes the characteristics of the sounds received in each ear.
       "People determine location of sound by comparing the difference between what the two ears are hearing," Professor Frampton explains. "The position of the head and the structure of the cartilage around the ear affect both the volume and the timing of the sounds perceived by each ear."
       KEMAR, placed in the center of an Anechoic Chamber which absorbs ambient sound, faithfully reproduces these head positions and cartilage structure to help engineers determine the relationship of the amplitude and phase readings from both ears and the originating sound. These readings are computed through MATLAB software to determine how an audio signal should be produced in order to simulate sounds from different directions.
       "The signal processing system we're developing will enable us to present sounds that appear to be originating from any direction," Professor Frampton says. "What used to be accomplished using dozens of speakers will be achievable with only two."