In order to improve present sonic guidance systems used at intersections by vision impaired pedestrians travelling alone, multiple speaker systems were adopted, putting two speakers at the destination of crosswalk, instead of a single speaker. Using this system that may provide more precise navigational clue than before, the sounds of present audible traffic signals (ATS) have been analysed from the points of view of accuracy of sound localisation capability and simplicity of discrimination. Two simulated bird songs, a chirp and a cuckoo, which are widely used in Japan for ATS and brief gaussian noise sounds were presented in front of the subject. A pair of the one of three sounds was presented from the two speakers placed in front of the subjects. To assess the spatial resolution capability of the subjects, the distance between the speakers was changed systematically, and the subject reported if the sound source consisted of the one speaker or the two speakers. The results showed that the discrimination performance was the best for brief noise sounds, poor for cuckoo sounds, and bad for chirp sounds. For example, when the angle between two speakers was set at 7.6 degrees, the rate of correct answer for two consecutive sound signals of brief noise, cuckoo and cheep sounds were 86.4, 36.4 and 27.3 percent, respectively. To obtain more than 85% discrimination performance, 15.2 and 22.6 degrees were necessary for cuckoo and cheep sounds, respectively. The results of simplicity of discrimination showed that brief noise was the simplest, cuckoo sound was second. It was suggested that the sounds of present ATS were not suitable in the multiple speaker system for providing reliable navigational clues to users since very wide distance between two speakers is necessary to obtain sufficient spatial resolution.
Abstract