Elizabeth Fehrer

Effect of Luminance Ratio on Masking

Previous studies have shown that, if brief stimulation of a limited retinal area is followed by stimulation of adjacent areas, the apparent brighrness of the first stimulus may be greatly diminished (metacontrast). The degree of such retroactive masking increases with stimulus-onset asynchrony, ( A t ) , reaching a maximum ac a At between 75 and 125 msec. This has been equally true whether the stimuli were brief light flashes (Alpern, 1953; Fehrer & Raab, 1962; Fry, 1934; Toch, 1956) or illuminated black or white paper figures on opposice colored grounds (Kolers & Rosner, 1960; Werner, 1935 ) . Recent exploratory observations of ours, however, have shown thar this relation between extent of masking and At does not obtain when the stimulus to be masked is a light flash at a luminance well below that of the masking stimuli. Earlier studies were not designed so as to yield this finding. In most, test and masking stimuli were equally luminous (e.g., Fehrer & Raab; Toch), or equally black or white (e.g., Kolers & Rosner; Werner ) . In Alperns study (and also that of Fry), luminance ratios were not held constant. Alperns procedure involved brightness matching of the test ( the variable) stimulus to a constant comparison standard. This procedure resulted in ( a ) the use of high test-to-masking-scimulus luminance ratios and also in ( b ) ratios thar varied with At. His data essentially show, for a given masking stimulus luminance level, the series of test stimulus luminances that match a given standard at each of the On explored. I t is not possible from his data to decermine the apparent brightness of a constant test stimulus as a function of nt. The purpose of the present exploratory study was to decermine the relation between At and extent of masking when the test stimulus is less luminous than the masking stimuli. The luminance of the masking stimuli was held constant throughout the experiment; test stimulus luminance was the parameter. Two measures of masking were used, accuracy of detection and paired brightness comparisons. The first experiment was designed to determine the relation berween At and accuracy of detection, i.e., discrimination of the masks alone from test-plus-masks, over the range of test flash luminances that would yield a range of accuracy of detection scores. This luminance range was found to be narrow and limited to exceedingly low values. In order to extend the