Abstract

For nonexperts to get consistently high quality results in color printing, computer power can be used to supplement the operator’s judgment, which is at present indispensable. In a method originally developed at MIT for gravure printing, input images are scanned, displayed on a TV monitor, and then edited in terms of their appearance rather than with respect to ink amounts or dot percentages. The edited appearance varibles are converted to ink amounts in a lookup table. The relationship between the original colors and the scanner output signals, as well as the relationship between the edited appearance variables and the corresponding ink amounts, must be accurately characterized. The fact that the scanner, display, and printer gamuts are usually quite different is a complication that must be addressed for good results. The gravure system used a display that predicted the output page apppearance very accurately. For desktop publishing, a less expensive method is used involving a library of reference images known to print correctly on the given printer. If the user image is adjusted to match the reference image overall, it will also print acceptably.

© 1991 Optical Society of America