The Study On Physical Dot Gain Of2^nd-order FM Halftone Based On Ink Spreading In All Superposition Conditions

FM screen dots, compared with AM screen dots, have many advantages in the ability ofreappearance for detail of the image, printing more bright-colored color, solving the "turtlesprofiling" problem. But there is a problem that the image shows particle structure especiallywith FM screen dots, which reduces the image quality. For the second order FM screen dots,not only the number of dots is varied by the density of target tone but also the size of dots ismodified to control noise and plugging, which improves image quality. But the physical dotgain can not be avoided and it has big effect on the accuracy of color duplication, whateverscreen dots or printing technology is used. Since ink dots are the basic unit of the colorduplication, transferring the original’s color and tones in printing industry, the modification ofthe physical dot gain can provide high prediction accuracy for FM screen dots.For FM halftone prints, this paper first analyses the two dot mixing models: dot-on-dotmixing model and random mixing model for FM halftone dots according to the superpositionof ink dots, then a spectral prediction model combined random mixing model with dot-on-dotmixing model is developed by using a weighting factor to signify the correlation proportion oftwo models. At the same time, we take into account the physical dot gain caused by inkspreading in all ink superposition conditions and modify the dot surface coverage byaccounting for the physical dot gain with the help of ink spreading curves and ink spreadingequations. Our results show that based on Yule-Nielsen modified Spectral Neugebauer model,the new prediction model combined random mixing model with dot-on-dot mixing model hashigher prediction accuracy compared with the random mixing model and dot-on-dot mixingmodel because the mean and max ΔE differences between measured and predicted values are5.18and10.52. The results also show that the mean and max ΔE differences betweenmeasured and predicted values by the new model are2.64and6.21, which is taken intoaccount the physical dot gain caused by ink spreading in all ink superposition conditions andthe modification for the dot surface coverage. The difference is smaller than that predicted bythe model that is not taken into account ink spreading. The prediction accuracy isconsiderably improved by modification of physical dot gain.