Abstract
Purpose: :
To compare color contrast sensitivities estimated by using two psychophysical methods with different criteria for threshold estimation.
Methods: :
Six subjects (1 male, 6 females; 23.7 ± 1.9 years old) were studied; they were all normal trichromats with 20/20 corrected visual acuity. The subjects were monocularly tested using two methods for contrast threshold estimation: Mullen protocol (Mullen, 1985) and equiluminant stimulus protocol. Both protocols used a Gabor stimulus (12° standard deviation), generated along a red-green chromatic axis (CIE1976: red u’ = 0.288, v’= 0.480; green u’= 0.150, v’ 0.480) surrounded by the same mean chromaticity and luminance background (CIE1976: u’= 219, v’ = 0.480; 40 cd/m²). Eight spatial frequencies (0.1, 0.2, 0.4, 0.8, 1, 2, 4, and 8 cpd) color contrast (estimated as the chromaticity distance in the CIE1976 diagram) decreased according to a staircase of 12 reversals; contrast thresholds were taken as the mean contrast of the last 4 reversals. For the Mullen protocol, contrast threshold was estimated for 9 stimulus chromatic conditions and one achromatic condition. The color contrast threshold for each spatial frequency corresponded to the chromatic conditions that provided the largest difference in relation to the achromatic threshold. For the equiluminant stimulus protocol, a heterochromatic flicker photometry at 20 Hz was initially performed to find stimulus equiluminance condition before contrast threshold measurements. The chromaticity distance along the CIE1976 u’ axis was taken as the color contrast threshold for both protocols. One-way and two-way ANOVA and Bonferroni's HSD post-hoc test (α = 0.05), and Pearson’s correlation were used to compare the results obtained with the two protocols.
Results: :
There were significant statistical differences for color contrast sensitivity estimated with the Mullen protocol between low spatial frequencies (0.1-0.8 cpd) and high spatial frequencies (8 cpd) (p < 0.01). There were significant statistical differences for color contrast sensitivity estimated with the equiluminant stimulus protocol between low spatial frequencies (0.2-1 cpd) and high spatial frequencies (8 cpd) (p < 0.01). There were no statistical differences for contrast sensitivity estimated with the two methods (p > 0.05).
Conclusions: :
Spatial color contrast sensitivity estimated with two different psychophysical methods provided similar values for every spatial frequency tested. Color constancy at different luminance conditions has the potential to help the matching of color sensitivity obtained with very different psychophysical conditions.
Keywords: color vision • contrast sensitivity • color appearance/constancy