xr E 
CET OR SSI EEE PPT 
oe — ru —— ——— ORT. d s OM 
ES RS SELS BARON NE RSA n CRE EEE TONER EE SE AO A ey zt A 
Effective exposure time at a point 'd' 
Exposure at a point 'd' (1x.s) 
Illuminance at a point 'd' (1x) 
(s) (7) 
2.3 Experimental Results 
2.3.1 Illuminance Distribution 
Some examples of the relative exposure distribution curve over the 
camera picture format are shown in Figures 8 to il. The reason for 
not presenting the relative illuminance distribution is that, accord- 
ing to the results of our experiment, the relative illuminance distri- (© 
bution curve of a lens at an aperture is almost identical to the rela- 
tive exposure curve of the same lens at the same aperture. We have 
presented therefore only the exposure distribution curves. The above 
finding means that the camera shutter does not affect the exposure 
distribution because the shutter blades move exactly, or very closely, 
in the principal plane of the lens. 
In each figure, the following curves are shown: relative exposure 
distribution curves at all apertures (f-numbers) available, a relative 
exposure curve at the full aperture (the smallest f-number) with the 
A.V. filter fitted, and the cosine fourth power curve for reference. 
Systematic error due to the deviation from the cosine law of incident 
angle has been compensated for. 
Figure 8 shows the relative exposure distribution curve of camera A. 
As shown in the figure, the curves closely match the cosine fourth 
power curve regardless of the aperture settings. This means that this 
camera lens shows no vignetting over the whole picture format at all 
apertures. The improvements in the relative exposure distribution (@ 
curve obtained with the A.V. filter is that the curve closely matches 
the cosine second power curve over the whole picture format. 
Figure 9 shows the results for camera Bl. In this case, the curve for 
£/5.6 lies between the cosine third and fourth power curves. This 
means that the effective aperture of the lens when viewed from an off- 
axis point is greater than when viewed from on-axis. At £/8 and 
smaller apertures the exposure distribution within the central 80 mm 
diameter circle on the picture format is improved so that the curves 
match the cosine third power curve. However, outside this circle, 
the exposure decline has not been totally improved. With the A.V. 
filter, the exposure distribution within the above circle is improved 
so that the distribution curve lies between the cosine Second and 
third power curves. However, outside this circle, improvement has 
still not been very significant. 
Figure 10 shows the results for camera B2. The distribution curves at 
large apertures are a little worse than those obtained by camera Bl. 
(6)