After taking the photograph, the plate is developed and reversed and placed in the same position relative to the grating, the eyes of the observer being placed at 0,, and Each eye will then only see the bands of image correspond ing with it, and the observer will see the object in relief, without the use of any viewing instru ment, since the grating is attached to the stereogram and is thus part of it.
This stereoscopic process, suggested in 1896 by A. Berthier, was put into practice by F. E. Ives in 1903, and perfected in 1906 by E. Estanave. These two authors have pointed out many other applications of the same principle.' As in the case of anaglvphs, the elements of the object registered in coincidence appear in the plane of the stereogram, the other elements appearing either in front or behind, according to the relative positions of their corresponding points.
In practice, "parallax-stereograms" are not taken directly, as we have said, but printed successively from the negatives of an ordinary stereoscopic pair, the grating only being used in the reproduction by means of a triple-body camera ; between the copying of the two pic tures the lens is moved through a distance equal to the separation of the eyes, perpendicularly to the bands of the grating.
It may be noted that if the observer places the right eye at 0 the left eye in the symmetrical position of 0, relative to 0,, he will see a pseudoscopic image. It would be the same for an observer correctly placed relative to the image if the grating were displaced through a distance equal to the width of the open bands. In all intermediate positions, or if the observer is not at the required distance from the stereo gram, each eye sees a little of each of the images, and the result is very confused.
835. Integral Photography. Under the name of integral photography, G. l.ippmann in iqo8 suggested a method the embodiment of which in practice was attempted in 1925 by E. Estanave.
A celluloid sheet of convenient thickness is embossed on its two faces with a large number of hemispherical projections, in such a way that the converging system formed by each element of the celluloid face when facing the subject gives an image on the emulsion covering the opposite hemispherical element. Without the use of any other optical system (except a mirror for the optical reversal of the image which is necessary, although not included in the original description), one obtains a very large number of minute images of the subject opposite which the film is exposed. After development, reversal, washing, and drying, the film examined from any viewpoint will show a complete image of the subject photographed, each point of the image (or, at least, each small element—almost a point) being supplied by one of the elementary images. Binocular vision of this gives the stereoscopic effect, with this peculiarity, that as the observer moves in front of the system, parts of the reconstructed object become masked and then unmasked by other parts, just as they would in moving about in front of the object itself, whereas in viewing a stereoscopic pair it is always the same objects in the background which are masked by the objects in the foreground, whatever the deformation of the reconstructed object, as the eyes are moved in front of the pair.
Prompted by this principle and by the parallax stereograms, C. W. Kanolt (1918) then G. Bessiere (1925) and J. de Lassus St. Genies (1932) have produced some very curious photographs, which, seen from almost any viewpoints, give a. perfect stereoscopic sensation with relative dis placements of the objects situated at various dis tances when the observer moved from side to side.
To obtain these pen-stereoscopic photographs the sensitive plate can be exposed, for example, a certain distance behind a vertical grating (the openings being 1/400 in. wide and the lines r (4 in. wide). The lens, fitted with a reversing prism and a diaphragm in the form of a vertical slit, is given a sideway movement during expo sure (in the case of same-size reproduction) or a rotation about an axis situated behind the plate when reproducing on a reduced scale, the posi tion of the sharp image being just behind the plane of the plate. The plate thus records be hind each opening of the grating, which acts as an elongated pinhole, a complete image of the object, retaining the vertical dimensions but compressing considerably the transverse dimen sions. The lens only plays a small part in the formation of these anamorphoses, obtained in such a way that each point of the object corre sponds to a single point in each elementary image. Once the negative is obtained, as many prints may be taken as are desired, and these are viewed by fixing a grating in front of them identi cal with the one used in making the negative.
836. Photo-stereo-synthesis. In spite of the fact that photo-stereo-synthesis (L. Lumiere, 1920) does not involve any stereoscopic prin ciples, the powerful sensation of relief given merits their inclusion among the various methods of stereoscopic photography.
The subject (generally a head and shoulders portrait) is photographed six times in such a way that the focus of each image is confined to a single plane of the subject.' The camera is moved nearer to the subject after each exposure, so that a new " section " is recorded on the same scale of reproduction. The six negatives thus obtained are printed very lightly on trans parency plates (the sum of the six densities only just making up the density of a normal trans parency). These plates are superimposed, their thicknesses representing (on the scale of the image) the thickness of each of the sections of the subject. The whole is illuminated from behind by a diffusing screen, and, on viewing from the one possible position, a sensation of relief is obtained, the part of each image in focus being the only effective part as far as the eyes are concerned.
This method has been applied with success to photomicrography (F. Bastin, 1921) and to radiography (E. Pohl, 1930).