Conditions under which the Haas precedence effect may or may not occur.

This investigation explored the stimulus conditions of the existence region in the free field of the Haas "precedence" effect. Experienced normal-hearing adults (N:5) listened to 50-msec bursts (0.2 msec rise-fall) of broadband noise from 2 loudspeakers at 1 m distance at ear level, at +/- 20 degrees re midline. Bursts from the loudspeakers were separated by interstimulus onset intervals (ISO-Is) between 0 and 50 msec. In one condition, uncorrelated noise was produced by separate generators; in a second condition (correlated noise), the signal from one generator was split and led to each speaker. Ss classified their experience of each presentation into one of five defined categories:: (1) single non-moving sound image; (2) 2 stationary but spatially distinct sound images; (3) single sound moving from lead to lag source; (4) as (3) but motion interrupted or broken; and (5) 2 successive sound images, with no apparent motion. In addition, Ss indicated direction (L-R; R-L) of any apparent motion. At 0 msec ISOI, directional judgments with either correlated or uncorrelated bursts were at or near chance level, as expected, and with correlated noises a single image was usually (71%) experienced, presumably at an apparent location at or near the midline. But with uncorrelated noises, 2 simultaneous but distinct sound images were usually (68%) perceived. At 2 msec ISOI, a weak precedence effect was exhibited for correlated noise, where the "single" response was often (55%) made; but even here the effect of the second burst was not "suppressed" entirely since direction judgments were 83% accurate. At 4 msec ISOI, "single" responses had declined to 36% for correlated and to 6% for uncorrelated noises, while by 8+ msec, "single" responses were dominated more or less completely by perceptions of dual sources. We conclude that the precedence effect is often confined to extremely brief ISOIs, and that the parameters of stimulus duration and complexity, of rise-fall times, absolute placement and angular separation of sound sources, subject differences, and a variety of types of perceptual experiences need to be addressed experimentally before the existence region of the precedence effect is well defined or uncritically accepted; and that it is too soon to attempt to posit neurophysiological mechanisms in explaining the phenomenon.