COLOR PHOTOGRAPHY &19 



The various grey tints range from almost total black to almost the pui'e white of 

 the paper. A successful three-color process should permit a faithful reproduction of 

 the above wedge by means of the complementary printing pigments, yellow, magenta, 

 and cyan, taken in the correct amount and superimposed one on the other. If we 

 refer again to Fig. 4, it will be noticed that the light range represented by the straight- 

 line proportion of the curve (useful range) for this particular emulsion is just about 

 16 to 1 (density 0.4 to 1.6 or transmission 50 to 2.5 per cent). This is also the lumin- 

 osity range chosen for the gray wedge as indicated by the figures above each square. 



In order to faithfully reproduce these luminosities in the negative, it is necessary 

 to choose such an exposure as to give, upon correct development to gamma of unity, a 

 density of 0.4 to correspond to step 1 and so on to density 1.6 for step 5. If the 

 exposure factor for each one of the three-color filters is correctly chosen and the three 

 negatives are developed to exactly the same gamma, the three wedges will appear 

 identical and will have the same densities by actual measurement. Under these con- 

 ditions each successive step of the wedge from light to dark in the negative represents 

 the same increment of density, or the light transmission of each step is one-half that 

 of the previous step. 



Now let it be assumed that one of the negatives (the green-filter negative, for 

 instance) has been exposed twice as much as the others. In this negative the density 

 and transmission of the various steps will be as given in Table I (dotted curve of 

 Fig. 4): 



Table I. — Density and Transmission of a Typical Sensitometric Strip 



It is evident that in this case the ratio between steps is constant except for the last 

 step, which should have had a density of 1.9 instead of 1.75. When printing these 

 negatives, it is necessary to print the green-separation negative twice as long as the 

 other two. This procedure, however, will give three prints in which steps 1 to 4 are 

 correspondingly the same, but step 5 will be heavier in the case of the green-separation 

 negative as compared with step No. 5 of the blue and red separation. After the three 

 positives are converted to the three complementary colors and superimposed, it will 

 be found that steps 1 to 4 are of neutral gray, corresponding to the original; but step 

 No. 5 will appear reddish, being deficient in yellow and cyan. 



If the range of luminositj^ of the wedge had been 8 to 1, the reproduction would 

 have been correct all the way through in spite of the overexposure of the green- 

 separation negative. The same applies to the case of underexposure of some of the 

 coloryseparation negatives. 



Illumination of the Subject. — The considerations pertaining to the luminosity 

 range of the gray wedge apply equally well to the range between the high lights and 

 the shadows of the subject. 



While extreme range between light and shadows should be avoided for the reason 

 just discussed, altogether too much stress has been placed in the past on the desira- 

 bility of flat lighting. 



