Abstract

We have measured light-adapted foveal luminance increment thresholds for 1.5 ft. (46 cm) diameter white targets by using a staircase. We wished to learn more about the mechanism that signals foveal luminance change. When identical luminance noise is placed at two spots 1.5 ft. (46 cm) to each side of the test, the threshold is more than doubled. The detection mechanism has a relatively broad field because the noise locus must be moved more than 5 ft. (1.5 m) away from the test for its effect to disappear. The mechanism may be even-symmetric but it cannot be linear, nor can scattered light explain the result, because antiphase noise modulation of the two flanking spots is at least as effective as in-phase noise. A rectifying nonlinearity is possible, but that cannot be the entire explanation because the antiphase noise is more effective than in-phase noise at low and medium separations. Similarly, in subthreshold summation tests, opposite-polarity flank pulses are at least as helpful as same-polarity flanks. One intriguing hypothesis is that the mechanism detecting our test stimulus is motion-sensitive and that the antiphase noise is effective within that mechanism because it simulates motion across the field containing the test. Indeed, subjective motion due to the noise mask is apparent to our observers.

© 1990 Optical Society of America