The inducing effect of blue in colour assimilation on a CRT screen is so strong that pale blue apparently veils the stimulus pattern (Miyamoto and Hasegawa, 1996 Perception 25 Supplement, 105). A similar tendency can be observed on painted paper, but this effect is not so marked. This led us to try colour matching between coloured papers and colours reproduced on a CRT screen. In both conditions the shapes (square) and the sizes (35 cm × 35 cm) were the same. They were displayed in the centre of a 155 cm × 155 cm gray (Munsell value 5, 22 cd cm−2) square with 10 cm wide white fringe (74 cd cm−2). The task was to adjust the colour on the CRT screen to match the coloured chip which was inside an illuminated box at a viewing distance of 180 cm in a dark room. Twenty-four colours were tested. Matched blues and other blue hues were slightly more saturated than object colours, but differences in chromaticity coordinates (u‘v’) were not statistically significant. This result is opposite to the prediction because it suggests a less bluish impression in the film-colour mode. By contrast, luminance in matched blues was lower than that of the object. This means that the brightness effect increased. An additional experiment in which the stimulus size was larger and the illuminating condition was changed to a light room produced the following results: (i) no meaningful difference in chromaticity coordinates could be seen; and (ii) the luminance of matched colours increased with size and in the lighter viewing condition. These facts show that the bluish colours on the CRT screen do not change their perceived chromaticity or spread spatially. The only difference was the subjective brightness increase in the condition when the room was dark. The strong effect of the blues in assimilation on the screen, therefore, does not depend on the colour identification process itself, but on the combined effect of luminosity and visual frequency response.
