The paper by F. W. Campbell and J. G. Robson explores the application of Fourier analysis to the visibility of gratings, focusing on the contrast sensitivity function of the human visual system. The authors measure the contrast thresholds for various types of gratings (sine-wave, square-wave, rectangular-wave, and saw-tooth) and compare them to the contrast-sensitivity function. They find that the contrast sensitivity for square-wave gratings is generally greater than that for sine-wave gratings, with the ratio of these sensitivities expected to be \(4/\pi\) at higher spatial frequencies. However, at lower spatial frequencies, the ratio deviates from this value. The study also examines the visibility of suprathreshold gratings, where the contrast is above the threshold required for detection. The results suggest that the visual system may have multiple independent channels, each tuned to different spatial frequencies, rather than a single peak detector mechanism. This model explains why the higher harmonic components of square-wave gratings have less effect on the threshold than expected. The findings are supported by neurophysiological evidence showing that retinal ganglion cells have band-pass characteristics with different optimum spatial frequencies.The paper by F. W. Campbell and J. G. Robson explores the application of Fourier analysis to the visibility of gratings, focusing on the contrast sensitivity function of the human visual system. The authors measure the contrast thresholds for various types of gratings (sine-wave, square-wave, rectangular-wave, and saw-tooth) and compare them to the contrast-sensitivity function. They find that the contrast sensitivity for square-wave gratings is generally greater than that for sine-wave gratings, with the ratio of these sensitivities expected to be \(4/\pi\) at higher spatial frequencies. However, at lower spatial frequencies, the ratio deviates from this value. The study also examines the visibility of suprathreshold gratings, where the contrast is above the threshold required for detection. The results suggest that the visual system may have multiple independent channels, each tuned to different spatial frequencies, rather than a single peak detector mechanism. This model explains why the higher harmonic components of square-wave gratings have less effect on the threshold than expected. The findings are supported by neurophysiological evidence showing that retinal ganglion cells have band-pass characteristics with different optimum spatial frequencies.