Visual Search for Warm and Cool Colours

Distinctions of warm vs cool are considered fundamental aspects of colour experience, though their status as a perceptual dimension remains unclear. Using a colour search task, we examined the relative salience of warm and cool colours relative to other hues. The search task was similar to McDermott et al. (JOV 2010) and involved searching for a 0.5° circular “fruit” target with variable chromaticity, superimposed on a dense 11.6° by 20° “foliage” background of random ellipses. The ellipses’ chromaticity varied along fixed axes in the cone-opponent colour plane that corresponded approximately to warm-cool (orange to greenish-blue), blue-yellow, or the two orthogonal chromatic directions (magenta-greenish). Targets on each background spanned a fixed range of hue angles and contrasts, and could be compared across backgrounds by how far the target colour was from the background axis. Search times for the targets were faster for detecting chromatic deviations from the warm-cool or blue-yellow backgrounds than the orthogonal directions. This suggests that both the warm-cool and blue-yellow dimensions are less salient than the orthogonal directions, consistent with biases found for a single bluish-yellowish condition tested previously by McDermott et al. However, the biases trended toward stronger asymmetries for the warm-cool axis. Interestingly, there were no differences in search times for warm vs cool targets or for blue vs. yellow, even though cool/blue hues may be more associated with shading or backgrounds while yellow/warm hues with objects. Overall the results suggest that sensitivity is weaker for the warm-cool dimension than for orthogonal dimensions of colour space, possibly because of selective adaptation to warm-cool and/or blue-yellow biases in the colour environment. This is also consistent with our recent analyses showing that the warm-cool dimensions closely align with the asymmetries in sensitivity predicted by uniform colour systems (Manalansan and Webster VSS 2023).