Abstract
The concept of the MTF is applicable to linear systems only and therefore can only be applied to radiographic screen-film systems if the output signal of the system is expressed in terms not of the photographic density of the film but in terms of the radiation intensity responsible for it. Both Rohler and Doi have pointed out that for double-coated X-ray film exposed with screens the conversion from photographic density to illumination cannot, for small images, be made using the normal (large area) characteristic curve of the emulsion. Rossmann and Sanderson, and Lubberts suggested that the errors involved in using the normal characteristic curve for the emulsion are within the experimental error of measurement. We examined by computer simulation the validity of the linearization process for an actual double-coated film system compared with that of a true linear model. It is concluded that the errors introduced by linearization of the screen-film system using a large area characteristic curve are within, or of the order of, experimental error. Under extreme conditions the error due to linearization exceeds experimental error.