Spatial-frequency tuning, contrast tuning, and spatial summation of suprathreshold lateral spatial interactions: grating induction and contrast–contrast

Two supra-threshold lateral interactions, the grating-induction effect,1 and the contrast–contrast effect,2,3 were compared regarding their dependence upon inducing grating spatial frequency, inducing and “victim” grating contrast, and inducing grating extent. Both effects cause “victim” gratings to be matched non-veridically. The respective magnitude of the effects was measured in common units which indexed the veridicality of contrast matches across a wide range (-0.9 to 0.9) of “victim” grating contrast. Grating induction had a low-pass, and contrast-contrast had a high-pass spatial frequency response, crossing over at ca. 1 c/d. Maximal grating induction strength exceeded that of contrast–contrast for 5 of 6 observers. Observers demonstrating strong grating induction tended to show weak contrast-contrast magnitudes, and vice versa. When inducing contrast was 0.75, the departures from veridical matching varied with “victim” grating contrast. For low frequencies (0.03125-0.125 c/d) grating induction produced a skewed unimodal departure from veridical matching, peaking at “victim” grating contrasts of 0.5. The pattern became bimodal at higher spatial frequencies (0.25-2.0 c/d) peaking at “victim” contrasts of ca. -0.3 and 0.5. Contrast–contrast caused symmetrical departures from veridicality which were consistent across spatial frequency, peaking at “victim” contrasts of ca. ±0.5. Both grating induction and contrast–contrast magnitudes decreased with reductions in inducing grating height, implying spatially extended mechanisms. With increasing inducing grating spatial frequency, however, the summation space-constant for grating induction decreased, whereas it increased for contrast–contrast.