A common problem with CCD sensors is their inefficiency to accurately record gray level information in the shadow range of an image. This problem is magnified on original copy having important detail in the midtone to shadow range. Consequently, when this type of an image is scanned gray levels in the shadow range of the reproduction are not easily discerned. This study examines two areas, the recording consistency across the scanning sensor and the gray level recording for a desktop six bit scanner. The former determined how consistently density information is recorded across the sensor, while the latter determines the amount of gray level discerned by the scanner. The result from scanning a three step scale in the consistency test showed that the scanner was not capable of consistently recording across the imaging sensor. These inconsistencies were more pronounced in the medium and dark steps in the test scale reproduction. The medium step reproduced ten distinct gray levels, while the dark step reproduced four distinct gray levels. The test results from the gray level recording capability test showed that 38 out of 64 possible gray levels were recorded by the scanner. This resulted in 59% efficiency in the gray level recording. The sensitivity density range which the scanner was capable of recording corresponded to a .41 . density range. (.08 to .49) In short, this study proposed an objective method for determining scanning consistency and gray level recording capability for a desk top scanner. Although emphasis was placed on a six bit scanner, with some minor modification the suggested method can be easily adopted to other scanners having greater bit values. The review of the literature section discussed the types of scanners used in desktop publishing. These scanners have many types of imaging sensors, such as self-scanned photodiodes charged injection devices (CID). Many factors must be taken into consideration prior to scanning an image. These include, sampling rate, scanner's bit value, screen frequency and output size. Since, each factor has a direct effect on the quality of the hardcopy output, they should be carefully considered.

Library of Congress Subject Headings

Optical scanners

Publication Date


Document Type


Department, Program, or Center

School of Print Media (CIAS)


Cost, Frank

Advisor/Committee Member

Chung, Robert

Advisor/Committee Member

Layne, Charles


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: TK7882.S3G36 1994


RIT – Main Campus