CHAPTER XIX

FILTERS

=The Function of Filters in Aerial Photography.=—The use of color screens or filters has been very common in ordinary landscape photography, for the purpose of securing approximately correct renderings of the brightnesses of colored objects. Plates of the non-color-sensitive type have their maximum of sensitiveness in the blue of the spectrum (Fig. 105) and in consequence blue skies photograph as white, while other colors are likewise reproduced on a totally wrong scale. Filters for correct brightness rendering are calculated for a given color sensitive plate so that the resultant reaction to the light of the spectrum copies the sensitiveness of the eye, which is greatest in the yellow-green. Such filters for use with the common orthochromatic plates are of a general yellow color.

Filters for aerial work are meant to serve quite a different purpose. Correct tone or color rendering is of quite secondary importance to another use of filters, namely, to cut or pierce aerial haze. It is quite a matter of accident that the same general color of filter is called for both to give correct color rendering and to pierce aerial haze, namely, _yellow_. Yet on closer analysis it is found that quite different types of yellow filter are demanded, spectroscopically considered.

Figure 106 (K_{1} and K_{2}) shows the spectral transmission curves of the Wratten K_{1} and K_{2} filters, intended for correct color rendering with orthochromatic plates. The absorption increases _gradually_ toward the blue. In the same figure is shown on an arbitrary scale the spectroscopic character of typical haze illumination, increasing in brightness inversely as the fourth power of the wave-length, that is, with great rapidity in the blue and violet. It is evident from this that a much more abrupt absorption than that of the K_{1} or K_{2} filter is desirable, because in the green of the spectrum the haze light is comparatively weak, and more will be lost by any absorption in this region through decreasing useful photographic action than will be gained by cutting out the haze. This latter consideration is important. The use of any filter means an increase of exposure; the use of yellow filters multiplies it several times. Careful experiment has shown that no filter of depth less than K 1½, to use the Wratten filters as a basis for discussion, are of real value in haze piercing. The _filter ratio_, or ratio of exposures with and without filter, is 4.7 for the K 1½ with the Cramer Isonon plate—a figure which shows the importance of securing the necessary haze-piercing character with the minimum absorption of useful photographic light.

=Practical Filters.=—Since the character of the absorption of the “K” filters is not all that could be desired, new filters, both of dyed gelatin and of glass, have been produced. The glass, a Corning product having a very sharp-cut absorption, has not yet been produced on a commercial scale with the high transparency in green, yellow and red that selected samples have shown. The United States Air Service has adopted filters of a new dye, called the EK, from the name of the company in whose laboratory it was produced. These filters are standardized in two depths of staining, called the “Aero No. 1” and “Aero No. 2.” Their spectral transmission curves appear in Fig. 106, along with those of certain darker filters useful only with panchromatic plates for exceptionally heavy haze. The characteristic of these Aero filters is their great transparency through all the spectrum except the blue, whereby the greatest haze-cutting action is attained together with a low filter factor. The filter factors of the Aero No. 1 and No. 2 with Cramer Isonon plates are 3 and 5, respectively.

=Effects Secured by the Use of Filters.=—The efficiency of yellow filters for haze-cutting is best shown by photographs taken at high altitudes with filters and without. Such illustrations are given in Figs. 107 and 108, where the first photograph is one taken at 10,000 feet without a filter, the second taken at the same altitude under the same conditions, but with an orange filter. Both are on panchromatic plates, and it will be seen that even with these plates the filter makes all the difference between a useless and a useful picture. But it must be clearly understood that the difference here lies between a plate sensitive chiefly in the blue and violet, and a plate affected only by the yellow, orange and red. The difference is not between what the eye sees and what a plate with a filter sees, as is sometimes supposed. As shown in Fig. 108, a filter enables the plate to photograph through the haze between clouds, but not through the clouds themselves. In general, no filter and plate combination which is feasible for aerial exposures is capable of showing more than the eye can see if yellow or orange goggles are worn. To do this it would be necessary for the photographic action to take place by deep red or infra-red light, which would demand exposures now out of the question.

Filters are almost always necessary in photographing from high altitudes or in making distant obliques. At times, particularly after a heavy rain, the air is clear enough so that filters may be dispensed with. Clearing weather was therefore chosen whenever possible for making obliques of the battle front.

=Filters for the Photographic Detection of Camouflage.=—In the photographic as in the visual detection of camouflage, the problem is to differentiate colors which ordinarily look alike, but which are actually of different color composition. Particularly important are the differences between natural foliage greens and the paints used to simulate them. If these differ in their reflection spectra, a proper choice of filter will show up the two greens as markedly different. Two kinds of difference may be produced; either the two colors may be changed in relative brightness, or they may be altered in hue. Thus foliage green, due to its possessing a reflection band in the red of the spectrum, which is absent in most pigments, may be made to appear _red_ while the camouflage remains green or turns black. Filters which cause changes of color are of course of no use for photographic detection of camouflage, since the photographic image is colorless. Brightness differences are alone available.

Those same filters which have been worked out primarily for producing brightness differences in visual detection of camouflage could be used photographically, provided the plates employed were color sensitive, and were as well screened to imitate the sensibility of the eye. But the most useful visual filters are those causing color differences to appear; more than this, the visual camouflage detection filters as a class have low light transmissions, so that their usefulness in photography is doubtful. Little work has actually been done with camouflage detection filters for photography. Yet in spite of this photography has been of real service in this form of detective work. Its utility for the purpose comes from the fact that the natural sensitiveness of the plate to blue, violet and invisible ultra-violet acts to extend the range of the spectrum in which differences between identical and merely visually matched colors may be picked up. Consequently the plain unscreened plate has proved a very efficient camouflage detector—so efficient in fact that all camouflage materials have had to be subjected to a photographic test before acceptance. Fig. 171 shows how an ordinary photograph reveals the unnatural character of the camouflage over a battery.

=Methods of Mounting and Using Filters.=—The most primitive way of mounting a gelatin filter is to cut a disc from a sheet of dyed gelatin and insert it between the components of the lens. For this purpose the gelatin must be perfectly flat, which is insured by its method of preparation and test. One disadvantage of this method is that the filter can be inserted and removed only upon the ground. It is less satisfactory the larger the diameter of the lens, and the wastage of filters due to insertion and removal is apt to be high. The camera should be refocussed after filters of this kind are inserted.

Glass filters, ground optically true, or gelatin filters, mounted between optically flat glass plates, are the most convenient and satisfactory. They may be mounted in circular cells to screw or attach by bayonet catches to the front of the lens. Or they may be mounted in rectangular frames to slide into transverse grooves in the camera body. Fig. 44 shows the mount of this latter form adopted in the larger United States Air Service cameras. This is particularly convenient if it is desired to insert or change the filter while in the air—a practice not generally considered feasible in war work with the photographically inexperienced observer, but likely to be common with the employment of skilled photographers for peace-time aerial photography.

German cameras are reported in which the glass filter is carried behind the lens, on a lever which also carries a clear glass plate of the same thickness, to be thrown in when no filter is needed, thus maintaining the focus. The performance of the lens will be impaired by this scheme, unless it is specially calculated to offset the effect of the glass introduced in the path of the rays behind the lens—optically true glass has no effect on definition if placed in front of the lens. Glass filters may also be placed in close contact with the plate or film, in which case they must be much larger, but do not need to be of as good optical quality.

=Self-screening Plates.=—Mention must be made of a quite different mode of realizing the filter idea, a method available where the sensitive plate is always to be used with a filter. This is to incorporate a yellow dye in the gelatin of the plate itself. The dye must be one which has no direct chemical effect on the plate, but which acts simply as a coloring agent for the gelatin. “Self-screening” plates, as they are called, have been produced by the use of the dye called “filter yellow” and have found some use in orthochromatic photography. They effect a useful saving of light through the elimination of the reflection losses at the surfaces of glass and gelatin filters. The filtering action of the dye in the plate is somewhat different from its ordinary one, since the deeper portions of the sensitive film are subject to greater action than the surface, and this tends to diminish contrast.