Facilitation and masking of briefly presented gratings: time-course and contrast dependence.

We measured two-alternative forced-choice contrast thresholds for briefly presented sinusoidal gratings in the presence of superimposed masking gratings of various contrasts, and at a range of onset asynchronies. Facilitation (lower thresholds) occurred when the mask was simultaneous, in-phase, and near-threshold, but was abolished at asynchronies of 50 msec or more and by presenting the test grating as a brief contrast reversal instead of a pulse. We argue that facilitation requires temporal summation of responses within the same neural channels, but our results do not distinguish between transducer and uncertainty models. Masking (threshold elevation) occurred over a broader range of asynchronies, and was not abolished by test contrast reversal. Masking and facilitation probably depend on different processes with different time-courses. The occurrence of masking at asynchronies outside the range of temporal summation suggests that a static, compressive transducer does not, in general, account for masking. Brief masking and prolonged contrast adaptation are very similar in magnitude, and as a function of contrast and relative spatial frequency. Masking and adaptation may have a common origin, but differ in speed of recovery.