ABSTRACT
Just a few years ago the precedence effect seemed like a straightforward auditory phenomenon based on hard-wired echo suppression, but data presented in this chapter compel us to entertain the possibility that listeners’ expectations influ ence this perceptual process. The precedence effect can be simply described and
is easily produced in any laboratory equipped with sound-generating and -control ling devices. To produce the effect one sets up two loudspeakers with one output leading the other by several milliseconds. Listeners report hearing one sound, not two, with the source located at the leading site. At short delays between the original sound and the delayed sound, the latter exerts little directional influence on the perceived location of the original sound. Delayed sounds produced in the laboratory can be used to simulate the reflections that occur in an ordinary room caused by sound bouncing off the floor, ceiling, walls, and objects in the room. These reflections “color” the original sound and reinforce its loudness, but are not recognized by the listener as separate sounds. At long delays echoes are perceived as separate sounds at their true locations. Throughout this chapter we use the term echo suppression to refer to the listener’s failure to hear the echo as a separate auditory event at its true location. We are following Blauert’s definition (1983, pp. 224-225) of echo threshold as the shortest delay between lead and lag onsets at which the echo is perceived as a separate sound, no longer fused with the original sound. When delays are below echo threshold the echo continues to influence the quality and timbre of the sound, so it is incorrect to think that echo suppression means the echo is not “heard.”
