Adjustment and discrimination measurements of the precedence effect

Abstract

A simple model to summarize the precedence effect is proposed that uses a single metric to quantify the relative dominance of the initial interaural delay over the trailing interaural delay in lateralization. This model is described and then used to relate new measurements of the precedence effect made with adjustment and discrimination paradigms. In the adjustment task, subjects matched the lateral position of an acoustic pointer to the position of a composite test stimulus made up of initial and trailing binaural noise bursts. In the discrimination procedure, subjects discriminated interaural time differences in a target noise burst in the presence of another burst either trailing or preceding the target. Experimental parameters were the delay between initial and trailing stimuli and the overall level of the stimulus. The model parameters (the metric c and the variability of lateral position judgments) were estimated from the results of the matching experiment and used to predict results of the discrimination task with good success. Finally, the observed values of the metric were compared to values derived from previous studies.

Echo suppression and discrimination suppression aspects of the precedence effect

“…The combined results from Shinn-Cunningham et al (1993) and Freyman et al (1991) suggest that all three measures of precedence outlined above reflect the operation of a single underlying mechanism. If this is the case, the degree to which the lead sound dominates in a localization task should be highly correlated with the degree to which information about the echo's position is suppressed and the degree to which the echo's identity as a separate event is suppressed.…”

“…To date, there has been little systematic comparison of performance in the three tasks; two notable exceptions are the studies of Shinn-Cunningham et al (1993) and Freyman et al (1991). Shinn-Cunningham et al (1993) obtained adjustment (localization) and lag discrimination measures in the same subjects from precedence-effect stimuli (noise bursts) presented over headphones.…”

“…To date, there has been little systematic comparison of performance in the three tasks; two notable exceptions are the studies of Shinn-Cunningham et al (1993) and Freyman et al (1991). Shinn-Cunningham et al (1993) obtained adjustment (localization) and lag discrimination measures in the same subjects from precedence-effect stimuli (noise bursts) presented over headphones. They successfully modeled the results from both experiments (and from other adjustment and discrimination experiments in the literature) within a single framework that quantified the relative weights given to the lead and lag (echo) events as a function of echo delay and the "direction" of the lead and lag stimuli.…”

“…In one version ofthis task, the echo direction and echo delay are held constant, and the subject must adjust the position of the direct sound so that the apparent location of the composite image matches a designated position (e.g., midline). Such "adjustment" experiments (Shinn-Cunningham, Zurek, & Durlach, 1993;Wallach et aI., 1949;Yost & Soderquist, 1984) have employed headphones and have had subjects adjust the lateral position of images (simulated direct sound and echo) within the head. In another version of this task, Hafter, Saberi, Jensen, and Briolle (1992) had subjects point to the apparent location of the fused image in a free-field setting, such as that shown in Figure 1.…”

Applications of virtual auditory displays

“…Many researchers discount the perceptual effects of echoes on source localization, citing a phenomenon known as the "precedence effect" (whereby the location of a leading sound dominates the perception of source location; e.g., see [35]). However, there is ample evidence that echoes can cause a measurable effect on perceived source location and on discriminability of the source position, particularly when the source has a slow rise time or when its onset is masked [36,37]. For applications in which information about the location of the source is more important than the realism of the display, including echoes and reverberation may be ill advised.…”