Table 1.

Statistical summary of the pseudo-natural stimuli

		“Full images”	“Contours”	“Textures”
Contrast:	Rms	100(0)%	71.3(2.1)%	70.0(2.2)%
Michelson		100(0)%	100(0)%	100(0)%
Sparseness:		0(0)%	49.1(3.1)%	50.9(3.1)%
Luminance histogram:
	White	50.6(2.2)%	25.0(1.8)%	24.4(2.2)%
	Gray	0(0)%	49.1(3.1)%	50.9(3.1)%
	Black	49.4(2.2)%	25.9(1.9)%	24.7(2.2)%

The table lists several statistical measurements (average and standard deviations) for the different stimulus conditions. The statistical summary of the synthetic images is similar but with even smaller differences between the images (see supplementary Table 1). Image sparseness is simply defined by the percentage of mean-luminance pixels. The table illustrates that the “contour” and “texture” conditions were similar for all low-level statistical measures. Their luminance histograms are nearly identical; therefore all statistics derived from the histogram will be similar.