Error reduction in two-dimensional pulse-area modulation, with application to computer-generated transparencies

The paper deals with the analysis of computer-generated half-tone transparencies that are realised as a regular array of area-modulated unit-height pulses and with the help of which we want to generate [via low-pass filtering] band-limited space functions by optical means. The mathematical basis for such transparencies is, of course, the well-known sampling theorem [1], which says that a band-limited func-tion y(x), say, [with x a two-dimensional spatial column vector] can be generated by properly low-pass filtering a regular array of Dirac functions whose weights are proportional to the required sample values yey(Xm) [with X the sampling matrix and m=(mi,m2)t an integer-valued column vector; the superscript t denotes transposition].