11. Processing Distortions and Observational Conditions
Emulsion shrinkage after processing has long been a problem, particularly
for films since these suffer from differential shrinkage which makes correction
difficult. The traditional way of determining this shrinkage is by measuring the
image of a grid or réseau, contact printed on to the film. The detailed investigation of
local variations over an entire 9-inch square format requires a closely spaced grid and
many measurements. Calhoun et al. ll5] have described an ingenious method using
moiré fringes, which allows quantitative information to be extracted with a minimum
of measurement. A contact print is made of a 300-line half-tone pattern. By super
imposing the negative on the original pattern, moiré fringes can be obtained. From
the separation and straightness of the fringes, size distortions can be calculated. A
more recent paper by Adelstein and Leister [161 has used this method to make a very
thorough investigation of film distortions. This paper brings out very well the
advantages of Estar film, not only because of its smaller absolute and differential
shrinkage, but also because of its smaller random displacements (5 ¡x as against 30 ¡x
for Topo base film). Estar film is also less sensitive to the method of processing,
tray and machine methods giving very similar shrinkage results.
After processing, the negatives can either be viewed directly or diapositives
can be made. In either case, the environmental conditions can introduce dimensional
changes, and the viewing or illuminating direction can alter the apparent position of
the image. Other requirements of flatness or contact in printing are the counterparts
of similar camera requirements.
Polyester base film has not only very excellent shrinkage characteristics but also
is less sensitive to environmental conditions, temperature, humidity and aging.
Calhoun et a/. [17] have measured the physical properties of a polyester film, and find
the following values for Plus X emulsion.
Polyester Topo
Thermal coefficient of linear expansion (% per 1°F.) 0 0015
Humidity coefficient of linear expansion (%per 1°R.H.) 00027
Processing. Differential shrinkage length-width (%) 0 003
Shrinkage (%) in accelerated aging test 0 03
0004
00073
0050
012
It is interesting to note that the thermal coefficient of the emulsion is almost
exactly that of aluminium, but some three times that of glass.
The angle of viewing or illumination is important as the image exists in the
emulsion in depth. Towards the edge of the format, the image-forming light is passing
obliquely through the emulsion and therefore the image will occur on an oblique
surface within the emulsion. There is some evidence that the image is not uniformly
distributed on this surface, the silver grains being closer to the surface for lower
exposures. Viewed normally to the emulsion surface, as in a comparator, the mean
image position seen will appear displaced by a small amount dependent upon the
exposure. The emulsion layer is usually some 10 ¡x in thickness, so that the apparent
position would not be expected to change by more than a few microns.
Hallert m reports some measurements made upon thin slices of emulsion, cut
normal to the surface, to determine the image position within the emulsion. Ele
finds that the change of position would not exceed 3 /x and that time of development
has little effect. Finally, Hallerfi 7 ' has made an interesting comparison between glass
384