Printer banding which appears as nonuniform light and dark lines across a printed page have been an important image quality issue in the printing industry, and have been studied by many researchers. However, no literature has reported banding perception in different printers with different characteristics of banding. In this paper the authors develop a tool for measuring banding based on human perception. The authors describe new cross-platform experiments using ten different laser printers having different imaging characteristics. In order to focus on the impact of the printer mechanism on banding and to remove the confounding effect of the halftoning algorithm, the authors employ a specially designed line screen with all the test printers. The authors use the pulse width modulation capability of a single reference printer to match the absorptance of these printers, and to also generate extrinsic banding signals. The experimental results identify the points of subjective equality of the ten printers relative to the banding of a reference printer, and provide the basis of a method of computing banding power by considering a contrast sensitivity function. The authors results show that regardless of the different banding spectral characteristics, the contrast banding power of a given printer can be mapped to a perceptually equivalent level of contrast banding power of one reference printer with added extrinsic banding.
Zygmunt Pizlo, Yousun Bang, Jan Allebach, "Banding Assessment with Controlled Halftoning: The Ten Printer Experiment" in Journal of Imaging Science and Technology, 2006, pp 522 - 529, https://doi.org/10.2352/J.ImagingSci.Technol.(2006)50:6(522)