Abstract
Hyperacuity performance deteriorates if targets are of opposite polarity. With grating pairs of mixed luminance and chromatic contrast, matched or unmatched in polarity, we have previously reported (Sun et al., 2005) luminance contrast polarity to determine hyperacuity thresholds, except close to isoluminance, when chromatic contrast polarity becomes important. This suggests both luminance and chromatic mechanisms compete for access to vernier mechanisms We expand these data and further investigate these mechanisms using target separation. With increasing target separation, contrast polarity effects on vernier are lost (thresholds become similar), which has been interpreted such that, with matched contrast polarity, vernier thresholds are determined with a linear filter operation, and with opposite contrast polarity local signs are used (Waugh & Levi, 1993). If luminance and chromatic mechanisms both access hyperacuity mechanisms, the interaction of target separation and contrast polarity should be the same for luminance and chromatic gratings. Psychophysical hyperacuity thresholds for sine-wave grating pairs of the same or opposite polarities for both luminance and chromatic conditions were measured as a function of target separation. Additionally, blue-yellow gratings were used with a deuteranopic observer to investigate hyperacuity performance based on S cones. We also related these results to physiological responses of macaque retinal ganglion cells. Thresholds for both luminance and chromatic gratings showed the same interaction with target separation and contrast polarity; for abutting targets thresholds were higher with targets of mixed polarity but when separated (ca. 1 deg) contrast polarity made no difference. We were able to discount other spatial cues (e.g., 2F chromatic response of MC cells) as contributing to hyperacuity with chromatic gratings. We interpret the results as evidence that both chromatic and luminance postreceptoral mechanisms have direct access to vernier mechanisms.
Meeting abstract presented at VSS 2012