The multiwavelength excitation (MWE) method for measuring the colorimetric properties of fluorescent whitening agent (FWA)-treated specimens illuminated by a standard daylight illuminant computationally approximates not the spectral power distribution (SPD) of the illuminant but the luminescent SPD excited thereby by a weighted sum of the luminescent SPDs excited by a few different narrow-band illuminations. The weights are optimized for the actual SPDs of those illuminations and the bispectral luminescent radiance factors of typical FWA-treated paper specimens. Since the latter is invariant among instruments once provided as the common numerical data, the variations of the narrow-band SPDs give major impacts to the reproducibility of this method.The weights optimized for the varied SPDs, however, mitigate the impacts. For investigating how they impact, one basic illumination system and its 16 simple variation systems were built virtually. The basic system consists of three narrow-band LEDs and one blue-excited white LED, whereas the individual simple variation system has either the peak wavelength or spectral width of one of the four LEDs (including the blue LED in the white LED) varied. With those systems, seven FWA-treated papers with the known bispectral radiance factors were measured computationally by simulating the procedure of the MWE method. The differences in the colorimetric values measured with the simple variation systems from those with the basic system are far below the just noticeable difference, which indicates that the MWE method can be a practical solution for better reproducibility in measuring FWA-treated papers.
K E Y W O R D SFWA-treated paper, MWE method, narrow band LED, reproducibility