Surface electrodes have been used for many years to pick up electrical activity that originates in the human eye or occipital cortex. In the earlier work, bright flashes of light were required to elicit responses that were large enough to measure against the ambient noise of the recording system. Recently, averaging techniques and other improvements have greatly increased the sensitivity of these methods. It is therefore no longer necessary to use the relatively meaningless light flashes, and responses can be obtained to visually significant aspects of the stimulus field. Of particular interest are specific effects due to image motion, color differences, and binocular fusion and rivalry. By the use of standard contact-lens and EEG electrodes, information on the processing of the stimulus pattern is obtained at both retinal and occipital levels of the human visual system. Microelectrode experiments with animal subjects have likewise shown that single flashes of light are of minor importance as visual stimuli, and that more significant aspects of visual patterns are particularly effective in arousing the responses of visual cells at all levels. Hence this evidence can be used as a guide to the interpretation of the human results, since these are necessarily obtained from electrodes that are remote from the sites at which they are generated. In short, the traditional methods of psychophysics are still used to provide definitive data on human vision. But the neural basis for these data is most clearly provided by microelectrode studies on the visual cells of experimental animals, together with such remotely recorded human action potentials as can be obtained from the retina and cortex.
LORRIN A. RIGGS, "Progress in the Recording of Human Retinal and Occipital Potentials," J. Opt. Soc. Am. 59, 1558-1566 (1969)