Image Quality of Four Digital Printing Methods Compared to Offset and Flexographic Printing

摘要

Reliability of the quality ratings: About eighty percent of the variance of quality measurements for the girl image, obtained from two different groups of observers, each tested with a different measurement technique, was accounted for by differences among the 41 print samples, leaving about twenty percent of the variance in the data attributable to causes unrelated to differences among the samples; either differences between the people in the observer groups, or differences caused by the measurement technique, or both. About fifty percent of the variance of quality measurements for two different images, girl and champagne, made by the same observer group, using the same measurement technique, was accounted for by differences among the print samples, leaving about half of the variance in the data to be accounted for by other causes (Figure 3). Since the coefficient of determination of the category and Proscale ratings for the girl was about 0.8, and the coefficient of determination for Proscale ratings of the champagne and the girl was about 0.5, causes other than measurement unreliability may be needed to 0.3 of the "missing variance" in the girl-champagne proscale comparisons. The other causes include the fact, as shown in Figures 3 and 4, that for offset printing, the champagne image was preferred to the girl image on high quality paper, and the girl image was preferred to the champagne image on low-quality paper. Subjective similarity and subjective preference: Both of the subjective evaluation methods, each used with a separate group of observers, showed that not all of the perceived differences among images were strongly associated with image quality. Observers separated each printing method from every other using Proscale, but some of the printing methods were judged to have the same level of print quality. Physical measurements of raggedness and gloss level were each clearly discriminated as separate proscale dimensions, and each measure was associated with a different print method, but neither measurement variable predicted print quality. Factor analysis of the category quality ratings produced two independent factors, one of which was strongly associated with gray halftone mottle, and with rated quality (Figures 8,9 and 10). The second factor was strongly associated with color gamut area and related measures, but only weakly associated (r ～.4) with rated quality. Practical implications: Several digital print methods are rated about as good as either offset of flexographic printing on papers with little coating, while offset still gives considerably higher quality ratings on glossy coated and matt coated paper. Subjective judgments of image quality are conditional on three factors: the content of an image (color and organization of detail), the type of paper, and the printing method.