Object-substitution masking: The role of low-level chromatic similarity

Poster Presentation: Tuesday, May 21, 2024, 2:45 – 6:45 pm, Pavilion
Session: Object Recognition: Basic features

Ryan Lange1 (), Michael J. McNicholas1, Steven K. Shevell1; 1University of Chicago

Synopsis: Object-substitution masking (OSM) with isoluminant chromatic stimuli tested whether flanker-target similarity effects can be selective for low-level (L/M- versus S-cone) chromatic responses, or only higher-level (integrated) representations of color. Tests revealed separable flanker-target similarity effects for the two cone-isolating chromatic axes, and hence a contribution from low-level chromatic representations to OSM. Background: OSM is generally thought to mask targets by substituting or updating one object-level representation to another. Some studies, however, have found feature-level effects in OSM. For example, color-tilt conjunction targets and flankers show greater flanker-target color similarity effects on color masking than tilt masking, and vice-versa (Gellatly et al., 2006; Huang et al., 2018). Whether OSM acts only on object-level representations, or can act on lower-level representations, is relevant to the ‘microgenesis’ of conscious visual awareness. Methods and results: Subjects reported target color from a 4-probe array (non-forced choice) in an OSM protocol. Within an isoluminant plane, target and distractor chromaticities were on either the L/M- or S-cone axis; chromaticity of the single flanker was either along one axis (50% of trials) or along an intermediate chromatic direction (other 50%). The effect of flanker-target chromatic similarity on masking magnitude was tested using similarity relations at three levels: “same” (flanker and target identical); “similar” (as in “same” but with added +/- modulation along the non-target axis); or “different” (no shared modulation on the target axis). Planned statistical contrasts showed less masking in the “different” condition than in either the “same” or “similar” conditions, but not between the “same” and “similar” conditions. Conclusions: The observed flanker-target similarity effect for the target axis, but not for added modulation on the non-target axis, is consistent with OSM mediated by low-level chromatic mechanisms. This suggests that OSM may act within and across multiple neural levels.