Robert Foster


Over the last decade, newspapers have been compelled to introduce a lot more process color to their pages. This coincides with the introduction of USA Today, but also with the fact that all forms of media, including television and magazines, have become significantly more colorful. There is considerable market pressure on newspapers to be colorful, as well. Continued audience interest and advertising dollars are at stake. This is not a trivial issue. At the same time, newspapers are being admonished to make sure that, if they do introduce color to their pages, then it should be of high quality. Low quality color is worse than no color at all. During most of this period, traditional, high end scanners comprised the only reasonable option available to address the task of generating color separation films. In recent years, with the improvement of microcomputer based technology, desktop scanning systems have become an option. The quality of the output of these systems has been suspect, however. On the other hand, it is generally acknowledged that these systems are improving. Of all the forms of lithographic printing, newspaper printing (coldset offset lithography on newsprint) is least able to take advantage of all the data that the upstream processes, particularly the color separation process, can provide. It represents the lowest level of reproduction fidelity. It has a shorter density range, requires a lower screening frequency, and is restricted to a more limited color gamut, for example, than the other forms of lithographic printing. This is due primarily to the substrate, newsprint, but also in part to news inks. Is it possible that today's desktop scanners now provide output whose quality level is sufficient for this printing process? Would readers show no preference for reproductions made from separations generated by these systems over those made from separations generated by high end scanning systems? That is the fundamental question addressed in this thesis. A common set of transparencies were separated through a high end system and a desktop system. Care was taken to prevent either system from being disadvantaged through parameter settings. The separations from both systems were stripped into a common test form, and printed on an offset newspaper press. Judges evaluated the pairs of images in a paired comparison test, indicating a preference for the high end generated image or the desktop generated image. The results indicate that readers do indeed continue to show a clear preference for images generated on a high end scanner. The reader is requested to take note of two caveats. First, this test represents a comparison of two scanning systems, not two families of scanners. The author found the two best scanners from each family that were available to him at the time of the test. The reader applies the findings of this test to the families of scanners that these two scanners represent at his or her own peril. Second, the reader is admonished to recognize that these are times of rapidly changing technology. The results of an experiment such as this could change as the technology advances.

Library of Congress Subject Headings

Color printing--Data processing; Color separation--Data processing; Desktop publishing; Newspaper publishing

Publication Date


Document Type


Department, Program, or Center

School of Print Media (CIAS)


Noga, Joseph


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: Z258 .F68 1995


RIT – Main Campus