Imagesetter output resolution in newspaper color image reproduction

Show full item record

Title: Imagesetter output resolution in newspaper color image reproduction
Author: Süsstrunk, Sabine
Abstract: The use of color in newspaper printing has significantly increased over the last few years. To compete with other news media, advertising as well as editorial design had to become more visually attractive, using more color graphics and images. The concurrent development of integrated desktop pre-press systems facilitated the integration of color, having less cost associated with new technology purchases than the traditional high end systems. The purpose of the thesis is to compare different output resolutions of an imagesetter and to find the minimal output resolution necessary for acceptable quality in newspaper color image reproduction. The speed of an imagesetter, which is dependant on the chosen output resolution, is an important factor in newspaper turn-around time. In this study, a matrix was designed containing various test targets and four different images with different image characteristics at four different output resolutions. Most imagesetters on the market today support output resolutions that approximate the four selected for the study (846, 1016, 1270, and 1693 spots per inch). The screen frequency of the images was 85 lines per inch, a common screen ruling in newspaper reproduction. The separations were output on a Linotronic 530 imagesetter. The output time varied from 18 minutes at the lowest resolution of 846 SPI to about half an hour at the highest resolution of 1693 spi. The test matrix was printed on Consolidated Newsprint on an offset newspaper Goss Community press. The print application's spatial resolving power was evaluated by exarnining Fresnel zone plate resolution targets under magnification. Due to ink spreading, no visible difference could be detected on the printed test matrix between the different output resolutions. The test matrix was also submitted to an audience for visual evaluation. The psychometric method applied was the paired comparison method, which is based on the principle that the percentage of observers preferring a stimulus over the other gives a direct indication of how the two stimuli differ. In this study, the stimuli were the images with different output resolutions. Each reproduction was paired with another image from the same type, but with another output resolution. The observers had to indicate if they preferred the left or right image of a pair. The criterion for evaluation was "better." The statistical analysis of the paired comparison method indicates that for all four images combined, there is a 95 percent confidence level that no visual difference in quality between the four resolutions could be detected by the observers. Experimental noise, specifically registration, skewed some of the individual image results. As a result, the first hypothesis-the visual quality of images reproduced in newspapers is not dependent on the imagesetter output resolution above a certain limit for a given set of print parameters-was proven to be correct. However, the second hypothesis-for offset newspaper printing on consolidated newsprint at a screen frequency of 85 lpi, the resolution limit lies between 1,000 and 1,200 SPi-has to be rejected. This study concludes that for the given print parameters, the resolution limit is 846 spi. Further studies might even indicate a lower limit. To summarize, one of the major disadvantages of electronic halftone reproductions today, namely output speed, can be optimized independently from technical considerations such as hardware and software systems. A ininimal output resolution can be found for a given print application by exarnining its characteristics, such as resolving power and print contrast. Any image output at a resolution higher than the mirtirnum does not improve the quality of the reproduction, and only slows down the production turn-around time.
Record URI: http://hdl.handle.net/1850/12640
Date: 1998-05

Files in this item

Files Size Format View
SSusstrunkThesis03-1993.pdf 15.02Mb PDF View/Open

The following license files are associated with this item:

This item appears in the following Collection(s)

Show full item record

Search RIT DML


Advanced Search

Browse