ABSTRACT
The objective of using green-noise is to combine the maximum dispersion attributes of blue-noise with that of clustering of amplitude modulated halftone patterns, and to do so at varying degrees. The motivation for using green-noise and green-noise halftoning algorithms is to produce patterns with adjustable coarseness that can be tuned to the reliability of a given printer to produce dots consistently. In a reliable printing device, green-noise would be tuned to produce patterns composed of small clusters having low halftone visibility, but in an unreliable device, green-noise would be tuned to produce large clusters that have high visibility but print consistently. Although error-diffusion is a good generator of blue-noise, the nature of green-noise to cluster pixels makes error-diffusion inappropriate. The chapter looks at modifications to R. Levien error-diffusion with output-dependent feedback that may improve the results offered by Levien’s arrangement of two hysteresis and two error filter weights.
