Luminance contrast detection is optimized for the large-scale luminance texture of the surround

Contrast sensitivity functions (CSFs) have been typically measured under persistent adaptation to a uniform background. However, this is an unusual situation, considering that visual scenes in our daily lives are typically filled by objects of various luminance levels in a wide dynamic range. We investigated the interactions between the luminance range of textured adapting background and test luminance (luminance of an immediate surround) to see how the contrast sensitivities adapt to background luminance distributions. The background was a random array of one-dimensional lines, which had a maximum luminance range of approximately 1:520 000, and the test stimulus was a Gabor patch presented on an immediate uniform surround with various luminance levels. The carrier orientation of the Gabor patch was orthogonal to that of the background lines. In the results, the sensitivity was highest near the mean of background luminance, and the luminance range with the highest sensitivity became wider (i.e., the selectivity of sensitivity optimization weakened) as the background luminance range became larger. This sensitivity optimization was significant mainly in the lower spatial frequency. These results indicate that the overall contrast sensitivity is adjusted to the luminance range of the visual scene and the sensitivity optimization is more flexible in relatively lower spatial frequency.