Abstract
The sound-induced flash illusion (Shams, Kamitani, & Shimojo, 2000, 2002) is a striking example of auditory ‘capture’. Flashes are perceived to either increase in number when paired with more beeps (‘flash fission’) or decrease when paired with fewer beeps (‘flash fusion’; Andersen, Tiippana, & Sams, 2004). In actual fact, the standard flash is comprised of two transients: an onset and offset luminance step, and so the beeps may simply be producing a bias to respond to these transients. In addition, our unisensory results have shown that naïve observers under-report the number of flashes (11.7 msec duration, SOA = 57 msec) when there are more than two. To control for these potential confounds, we developed a single transient stimulus: a 2° disc, 7° peripheral, with luminance ramped on (to 99% contrast) or off in 350 msec. With trained naïve observers, these ramped stimuli were randomly interspersed with single or multiple standard flashes in the presence of 0, 1, 2 or 3 beeps. Observers were instructed to count either the number of transients or flashes seen depending on the condition. With our single transient stimulus greater fission effects were observed than with standard flashes, but only when beeps occurred during the ramp. For standard flashes, fission and fusion were observed dependent on the relative timing of the audiovisual stimuli. Despite the fact that our single transient stimulus eliminated bias towards reporting two events, a greater number of transients were counted than with a two-transient flash. It is suggested that ambiguity in the ramp stimulus makes an associated transient more vulnerable to auditory capture. Thus, consistent with the law of inverse effectiveness, the more accurately perceived auditory stimuli dominate the visual percept.
ACNRU postgraduate scholarship (AW).