Key points are not available for this paper at this time.
Sets of pairs of coloured paint samples have been prepared, each pair showing a difference in lightness (1–2 1976 CI E LAB units), but a negligible chromaticness difference. One set consisted of neutral samples, with Y values ranging from 0. 1 to 80. Other sets consisted of red, yellow or blue samples, the range of Y values being necessarily restricted for the more saturated sets. The pairs were assessed visually, under llluminants D65 and A by a panel of 24 observers using a ratio method. The agreement between E (and U values from various colour–difference equations and me mean visual results was found to be poor. A simple new lightness scale has been found which gave much improved agreement. The same scale could be applied to the results for both illuminants. Visual judgements of lightness appear to be independent of the chromaticity of the samples. No evidence of a “crispening” effect was found. It is suggested that such an effect decreases as the size of sample increases and that the effect is much smaller than has been claimed.
Coates et al. (Wed,) studied this question.