When a deviant stimulus is presented within a stream of homogeneous stimuli, its duration tends to be overestimated. Two experiments investigated the effects of oddball serial position and pitch deviancy on perceived duration. In Experiment 1, the oddball method was used, in which an oddball stimulus is embedded in a series of standard stimuli and randomly positioned in each trial. In Experiment 2, the oddball position was stable and its deviancy varied from trial to trial. Musician and nonmusician participants were asked to judge whether the comparison interval was shorter or longer than the standards. The study indicates that for nonmusicians, the duration of an oddball stimulus appears longer than the repeated standard stimuli. Moreover, the oddballs occurring in later positions in the stream of stimuli are perceived to be longer than oddballs occurring in earlier positions in the stream. Also, a higher degree of oddball deviancy results in a greater dilation of perceived duration. In contrast with the results of nonmusicians, there is neither a position nor a deviancy effect with musician participants; the subjective duration remains constant. Several explanations are discussed in order to account for these group differences.
One of the features of the auditory system is its ability to efficiently process events that occur in rapid succession. The aim of the present study is to propose a new way of investigating sensitivity to auditory tempo changes. More specifically, it proposes to compare the relative sensitivity (bias) to acceleration and deceleration in both musical and monotonal conditions. Bias was measured with (1) a conventional psychophysical method known as the method of constant stimuli (MCS) and (2) a so-called method of dynamic stimuli (MDS). The latter method consists in responding with a finger press as soon as a near-continual tempo change is detected. With the MCS, there was no preference, as estimated by the point of subjective equality, between acceleration and deceleration in the monotonal condition, but there was a preference in the musical condition that indicated more facility for estimating decelerations than accelerations. The results obtained with the MDS are consistent with the MCS results, given that the response time was faster for decelerations than accelerations in the musical condition but not in the monotonal condition. We conclude that the MDS is a sensitive tool for investigating slight tempo variations.