There are two distinct methods of using this process. If the screen elements are of a regular geometrical pattern, regularly recurring, then one may use the screen-plate pressed into contact with a separate panchromatic plate, and on development obtain a negative the image of which is broken up into the geometrical pattern. From this we can make any number of positives and by binding them up with similar screens in correct register, we shall obtain a colored result. This method is known as the separate method. But with an irregular haphazard mosaic, such as is formed in the autochrome plate, and with elements of such a fineness, it is an almost hopeless task to fit the positive to the screen-plate; therefore, the emulsion is coated on the plate itself, and, after development, the negative image is dissolved away, leaving the unexposed and undeveloped emulsion in situ. This is then developed and gives us a positive in colors. This will be easily grasped from Fig. 17A, which shows the negative image with the black silver behind the appropriate color elements; if we imagine all the white spaces there shown to be covered by undeveloped silver bromide, we have an exact picture of the developed autochrome or any other screen-plate by what is known as the combined method, that is, one in which the emulsion is coated on the screen-plate itself. If now we dissolve the primary negative image, it is clear that we still have our silver bromide under the other elements, and if this be exposed to light and developed we shall obtain the result shown in Fig. 17B, and there is no trouble in registration.
There are advantages and disadvantages in both methods. In the combined method, if we want more than one copy of a subject we must obviously make more than one exposure, whereas, with the separate process, having once obtained the matrix negative we can produce as many copies as we like. The combined method involves another operation, that is the reversal of the negative image obtained by primary development. Also, if we make a mistake and spoil the result, we have lost both screen-plate and picture; while with the separate method our only loss is the panchromatic plate, and the screen-plate is still intact. As to which is the easier process, this is merely a matter of opinion. As regards excellence of results there is not, in the writer's opinion, anything to choose between them, assuming, naturally, that the best results are obtained in both cases. Possibly for the novice the separate method is a little the easier, that is, if he can make a transparency.
In the case of screen-plate emulsions, as with all other color-sensitized ones, the characteristic preponderance of the activity of the blue and violet rays still exists, so that it is necessary to insert between the object and its image a correcting or compensating filter, which reduces the rendering of the colors by the plate to that seen by the eye. In Fig. 18 are shown two curves representing the luminosity of the spectral colors as seen by the human eye and by the plate, the former being marked V and the latter P. As will be seen, the peak of the curve of the latter is about wave-length 4500 in the deep blue, whereas the visual luminosity peak lies about wave-length 5500 in the yellow-green. By an error in making the drawing, the wave-lengths are there given in hundreds instead of thousands. An extra zero should be added to each number to make them agree with the text.
The filter that will shift the plate's vision so as to coincide with that of the retina is of a faint reddishyellow tinge and can be made with tartrazin, pheno-safranin and aesculin. The instructions for filter making given in Chapter III (Color Filters) must be rigorously followed: instead of using the above named dyes, of which the pheno-safranin has the disadvantage of rapidly fading on exposure to light, we may use the more stable combination of fast red D and filter yellow. The density of the first named dyes, that is, the quantity per square meter of filter area, is tartrazin 0.2 g, plus pheno-safranin 0.017 g, plus aesculin 4.0 g. The densities for the more stable dyes are filter yellow 0.6 g, plus fast red D 0.07 g.
It is not immaterial where the filter is placed, as it has an effect on the focus of the lens. Placed in front of the lens, it shortens the distance of the plane focused on by one-third of the thickness of the filter, but as this reduction of the object-plane is extremely small in all out-door work compared with the distance of the object it can be ignored. It is only when one reproduces objects, such as a picture, or flowers, in something like natural size that it need be taken into account. By far the best position for the filter is behind the lens; here it is protected from light, from accidental displacement and from injury, and it automatically makes its own correction in focusing. In all cases focusing should be done with the filter in position. As to the method of calculating the area of the filter for any given lens, see Chapter III (Color Filters). It is extremely important that the filter should fit close to the lens, so that no white light can creep around the edges or be reflected from the front surface to any part of the plate; for this reflected light would cause local or general fog, and consequent falsification of the colors.
As in the use of the screen-plates the glass is always turned towards the lens and the sensitive surface away from it, it is clear that this must be allowed for in focusing. The matt side of the ground glass should, therefore, be reversed in position and face outwards, and its thickness should be that of the screen-plate, if critical sharpness is required. An ingenious idea that has been put forward is to clean the emulsion from a spoiled plate and place this, with the color elements outside, in place of the usual ground glass.