RESUMO
PURPOSE: Chromatic variations across different copies and different editions of pseudoisochromatic tests and violation of underlying principles of construction for individual plates can influence test results. METHODS: We analysed the colorimetric characteristics of three different editions of Velhagen-Broschmann pseudoisochromatic plates (30th edition printed in 1995, 31st edition printed in 1997, 32nd edition printed in 2000). One hundred and twelve coloured dots of 18 plates were chosen from each edition. We measured RGB and CIE XYZ values using a spectrophotometer. Differences in lightness and chromaticity between corresponding dots of different editions were analysed in terms of Delta L* and Delta u'v', respectively. For each plate deviations from dichromatic confusion lines were analysed. Furthermore, we determined the relative luminance of a target compared to its background in terms of the Weber contrast. RESULTS: The mean Delta L* across editions was 2.05 (+/-1.4) and the mean Delta u'v' was 0.0078 (+/-0.0029). For two plates the deviations of targets from dichromatic confusion lines exceeded suggested values. For a number of plates, the lightness contrast between the symbol and its background was high. Comparison with psychophysical data showed that these colour plates are easily detectable by colour-deficient observers. CONCLUSIONS: Lightness and chromatic variation across the three editions was moderate except for a small number of plates perhaps due to inaccuracies in the printing process. The design of several plates should be revised according to standard principles of construction of colour deficiency tests.