I started down the raster graphics path in the '70s when the university computing center installed a Clevite electrostatic dot printer.
I open coded the CDC random number algorithm into a tight loop that dithered images into existence. I eventually printed pictures but for a test case I rendered one large sphere using direction-cosines as described in my analytic geometry textbook.
For photographs I borrowed time on Rossmann's optical scanner normally used for digitizing x-ray diffraction patterns. wikipedia 
My dithering software had controls for lightness and contrast but with minutes per run this was too slow to get set correctly. I cooked up a scheme to force an equal distribution of all shades which I later learned was called histogram normalization. wikipedia