SUBJECTIVE SHARPNESS OF IMAGES 



57i 



the figure of confusion of 0.18 milliradian would be twice as large as the 

 Hmen of 0.090 milliradian, and would result in wrong judgments in 9 

 per cent of the observations, corresponding to real discrimination in 

 82 per cent of them. Likewise a change of 0.27 milliradian would result 

 in real discrimination in 96 per cent of the observations. Any change 

 larger than about three times the limen would be discriminated in 

 practically every instance, under the conditions of our experiment. 



THOUSANDS OF FIGURES OF CONFUSION PER PICTURE 



200 100 75 50 40 30 20 10 



1 I I I I I I I 



CALCULATED FREQUENCY BAND IN MEGACYCLES PER SECOND 

 10 8 6 5 4 3 2 1.0 0.8 0.6 0.5 



m -8 



i= -12 

 < 



0.5 



0.6 



0.7 0.8 0.9 



1.0 1.5 



IN MILLIRADIANS 



3.0 



3 4 5 6 



IN MINUTES OF ARC 

 SIDE OF SQUARE FIGURE OF CONFUSION 



10 



Fig. 3 — Sharpness of small-sized motion pictures at a viewing distance of 30 inches. 

 The frequency band is calculated on the basis of a 10-Inch by 7|-inch television 

 picture, 30 frames per second, with 15 per cent horizontal and 7 per cent vertical 

 blanking, under the condition of equal horizontal and vertical resolutions. 



Figures 3 and 4 show the curve of Fig. 1 replotted in terms of some 

 additional objective variables. A scale of nominal frequency band 

 width required for transmission of the image signal over a video circuit 

 has been worked out on the basis of our comparison of the out-of-focus 

 motion picture with a television image of known characteristics, to be 

 described in section 5. We see that in order to effect an increase in 



