This paper compares two methods of presenting targets on CRT screens of modern radar. One method makes use of grid targets (positive signals) which appear brighter than their background, and the other employs cathode targets (negative signals) which appear darker than their background. Data were collected from three subjects on a very small target and on one of moderate size. A third target, intermediate in area, was studied, but in less detail. The results show: (1) The visibility of a cathode target is better than that of a grid target at bias levels which are relatively bright. The amount by which cathode targets are more visible is from 1 to 9 db. (2) There is evidence that the dimmest bias at which the cathode target is superior to the grid target is a function of the size of the target. (3) The optimal bias for a very small target is shifted to a brighter level by about a volt and a half when target voltage is injected at the cathode. (4) The superiority of the cathode target is due, not to any inherent advantage of dark-on-light targets, but rather to a property of the phosphor which allows the phosphor to work on a portion of its response curve favorable to human vision.
M. W. HARRIMAN and S. B. WILLIAMS, "Visibility on Cathode-Ray Tube Screens: Positive vs. Negative Signals on an Intensity Modulated Scope," J. Opt. Soc. Am. 40, 102-102 (1950)