Noise as a carrier of graphical information and a transfer function describing its effect on images

Noise in a communications channel is a random disturbance. It is additive to the signal, and the signal-to-noise ratio must be kept as high as possible. This paper is a study of systems containing strong noise that cannot or must not be made small because it serves in one way or another as a carrier of information. Systems which can be classified as modulated noise and are found in graphical communications include some forms of both silver halide photography and toner xerography, thermoplastic deformation xerography, and halftone printing. In all cases under study the separation of signal from noise is possible because the frequency spectra of signal and noise are different. Difficulties arise where the spectra overlap. It is shown that the quality loss inflicted by modulated noise on image or line copy reproductions may be treated by the modulation transfer function method and that the transfer function has a gaussian shape.