LIGHT FILTERS 299 



the filter is used will be somewhat as indicated by curve C, which is the same as curve 

 B except that its ordinates have all been multiplied by the "filter factor." By 

 using the G filter and increasing the exposure by the proper filter factor, we have 

 actually increased the density of the silver deposit on the film between 565 ray. (where 

 the C curve crosses the A curve) and have decreased the density below 565 mix. 



The photographic operations taking place and illustrated graphically in Fig. 5 can, 

 perhaps, be better understood if we consider the efi'ect on the density of a developed 

 panchromatic material having the characteristics given by the third curve when a 

 photograph is made of a white or gray (nonselective absorbing) surface which is 

 illuminated by mean noon sunlight. For an exposure such that the linear region of 

 the H and D curve is employed, the density of the negative (as a function of wave- 

 length) will be given by curve A when no filter is employed. Since we are not here 

 concerned with color separation, this exposure will result in a uniformly dense negative 

 whose density we shall call Z)i. If we make another photograph of the same subject 

 with the filter but with the same exposure as in the original case, the density of the 

 second negative will be less than that of the first negative because those wavelengths 

 below 550 m. are considerably attenuated or entirely blocked from affecting the 

 photographic negative. In this case the density of the second negative will be desig- 

 nated as Di. However, if we make a third photograph of the nonselective surface 

 through the filter but this time multiply the exposure by the exposure factor of the 

 filter, the developed negative will have the same density Di as the original photograph. 

 It is assumed that the processing conditions are the same in all cases. 



It is therefore evident that the use of a filter changes the density of the developed 

 negative as a function of wavelength or color, as is immediately evident by comparing 

 curves A and B of the bottom graph of Fig. 5. From the explanation in the preceding 

 paragraph it will be seen that the exposure factor for the filter is effective in tending to 

 keep the density of the developed negative independent of the manner in which the 

 filter alters the spectral distribution of density deposit, thereby making more certain 

 that the straight-Une region of the H and D characteristic is utilized. It may there- 

 fore be said that, in a general way, the filter alters the spectral response of the negative 

 to the stimulus of the light source and object being photographed whereas the filter 

 exposure factor represents an increase in the exposure which is required in order that 

 the region of correct exposure of the H and D curve be maintained when some of the 

 light is removed by the filter. ^ 



Curve A at the bottom of the chart is determined by multiplying together for 

 each wavelength from a = 350 m/x to /3 = 712 mju the relative intensity of the sun- 

 light J\ and the sensitivity of the photosensitive material R\. Curve B is determined 

 for the same wavelength limits by multiplying together J\ from the top curve, the 

 filter transmission T\ of the middle curve, and the spectral characteristics of the 

 panchromatic material R\ of the bottom curve. Thus we have illustrated graphically, 

 the essential operations involved in determining filter factors for a given set of condi- 

 tions. Since the middle curve (representing filter-transmission characteristic) will 

 be the only one of the top three curves which will remain constant for a given filter 

 and either the top curve will change for a different light source or the third curve will 

 change for a different film emulsion, it is apparent that there is no single "filter 

 exposure factor" which applies under all conditions for a given filter; this filter 



' In this discussion it has been assumed that an ideal uniform nonselective absorbing medium has 

 been photographed. In practice this is seldom, if ever, the case. The photography of images of vary- 

 ing colors and tone depths will produce an imaged (rather than a uniformly dense) negative. In such 

 cases the results are not so simple as outlined above, and the effects of selective absorption of the filter 

 and that of the filter exposure factor cannot be considered independently of one another. Yet the 

 general conclusions given above hold true with secondary modifications. 



