Abstract
Abstract:
Scene perception research often uses visual masks to vary processing time. However, little is known about the effects of spatial and temporal masking parameters when masking real world scenes. Studies using visual masking typically include a blank screen inter-stimulus-interval (ISI) between target and mask, which is frequently black. However, some researchers use a neutral gray ISI screen adjusted to the mean luminance of the target and mask. Doing so is presumably based on the plausible assumption that sudden changes in contrast between target, ISI, and mask, may affect masking by causing pulses of neural activity in transient response cells. The sustained-transient theory of visual masking (Breitmeyer & Ogmen, 2006) argues that activity in transient channels can interfere with processing in sustained channels. If so, we would expect that such effects should be much greater with a black ISI than an equiluminant gray ISI.
Method: Subjects performed a scene gist recognition task with either black (high contrast) or equiluminant gray (low contrast) ISI screens. Masking strength was also manipulated by using either fully phase-randomized scenes-as-masks (weak mask) or normal scenes-as-masks (strong mask), and a no-mask control condition. Target and mask duration were both 12 ms, and ISI varied from 0–84 ms (SOA 12–96 ms).
Results: Consistent with the sustained-transient theory, the high-contrast black ISI condition produced significantly stronger masking than the equiluminant gray ISI condition, but only at ISIs of 0–12 ms (SOAs 12–24 ms), and only in the weaker phase-randomized scene-as-mask condition. ISI (or blank screen) contrast with the target had no effect in the no-mask condition.
Conclusions: Contrast between ISI screen versus the target and mask affects scene masking, but only at short ISIs with moderate strength masking. If masking is very strong, or conversely very weak (long ISIs, or no-mask), there is no effect of ISI contrast.