Contrast increment detection of the square-wave third harmonic: Some support for the hypothesis of independent spatial-frequency channels

Under some conditions, spatial frequency channels seem to be independent; activity in one channel has no apparent effect on activity in any other. The evidence suggesting channel independence comes, in part, from threshold summation experiments that have shown test grating detection to be unaffected by the presence of a near-threshold grating if the frequencies of the two gratings are sufficiently different (Graham and Nachmias, 1971; Abadi and Kulikowski, 1973). On the other hand, adaptation experiments have indicated that channels can interact (Tolhurst, 1972; Stecher, Sigel and Lange, 1973). A square-wave grating’s third harmonic, for example, produces more adaptation when presented alone than when presented as part of the square-wave grating (Tolhurst, 1972; Nachmias, Sansbury, Vassilev and Weber, 1973). One mechanism that could account for such a reduction in adaptation is mutual inhibition among channels (Tolhurst, 1972). Assuming that channels do operate independently near threshold, and interact during adaptation, an important question still remains unanswered: how do channels behave when receiving suprathreshold stimulation in the absence of adaptation? The masking experiment reported here implies that, under these conditions, channels function independently. Recent experiments have shown that test grating detection can actually be facilitated by the simultaneous presence of a masking grating (Nachmias and Sansbury, 1974; Stromeyer and Klein, 1974; and Sansbury, 1974). The present experiment shows that a square-wave grating can facilitate detection of a test grating whose frequency equals that of the squarewave’s third harmonic. More importantly, this facilitation can be accounted for by assuming that channels. operate independently. The experiments employed a two-alternative temporal forced-choice procedure. The two observation intervals, delineated by tone bursts, had a duration and temporal separation of 250 msec. During the time remaining in the 4.65 set trial, the observer responded and received immediate feedback as to the correctness