Development of Image Telegraphy Over the Last Decade1

Arthur Korn.

The most important foundations of all methods of image telegraphy had already been laid in the prewar period, both for the transmission of photographs in black and white by wire and for their transmission by radio. The extraordinary development of radiotelegraphy in recent times has had the consequence that, in the field of image telegraphy as well, it has chiefly been the methods of radio transmission that have been improved and brought closer to practical applicability; in particular, amplification by cathode tubes has led to a great step forward in these wireless methods. But the use of amplifier tubes has also given advantages for transmission by wire which cannot be underestimated. Finally, in the field of “indirect telegraphing” of images (by means of letter telegrams or perforated tape with holes), numerous new experiments have been carried out in continuous connection with earlier experiments, clearly revealing the importance of indirect methods for certain purposes.

A survey of recent achievements may thus be divided into three parts:

1. Advances in wireless telegraphing of images.
2. The influence of amplifier technology on image telegraphy by wire.
3. Advances in indirect methods of image telegraphy.

Advances in Wireless Telegraphing of Images

Here, above all, the teleautographic method is of importance, which at first seemed applicable only to the transmission of manuscripts and drawings, but later proved applicable also to the transmission of photographs in the form of raster (screened) drawings. In order to make the advances of radio teleautography intelligible, let us first briefly recall the principle of tele-

of the autographic method in transmission by wires. The manuscript or drawing to be transmitted is applied, with electrically nonconducting ink, to a metallic film (of foil), which is wound onto a metal cylinder capable of rotating. A metal pin slides over the film, shifting, like the pin of a phonograph, in the direction of the cylinder’s axis as the latter rotates. Each time the metal point touches the conducting part of the film, a current is sent through the line to the receiver, whereas the current is interrupted when the point falls on a nonconducting part of the film, i.e., on the manuscript or drawing.

In the receiver, a second cylinder rotates synchronously with the transmitter cylinder; the receiving paper is wound onto it, and the manuscript or drawing is reproduced there. For this purpose, marks are made on the receiving paper by means of the intermittent current arriving from the transmitter. In the original electrochemical receivers, the receiving paper was soaked in a suitable solution (potassium ferrocyanide, potassium iodide, etc.), and a metal pin slid over it; currents from the line were passed through this pin into the moistened paper, the receiving cylinder, and then to ground. The paper was colored blue under the pin each time current arrived, remaining uncolored in the absence of current. In this way, with synchronous rotation of the cylinders in the transmitter and receiver, the manuscript or drawing appeared on the receiving paper as white on a blue background. The electrochemical method, as well as the electromechanical one—where marks were made on ordinary white paper by means of a writing pin pressed against the receiving paper electromechanically—cannot provide the transmission speed of the method using the newer photographic receivers, in which the line current produces marks photographically on sensitive paper or film. In particular, in this field the first place was taken by the string galvanometer introduced by me in 1906 into the receiving devices of picture telegraphy (a thin thread stretched between the poles of an electromagnet, through which the line currents pass), and also by the oscillograph (a small mirror turned electromagnetically by the line currents). Owing to the high speed of transmission by the photographic method, it became possible to transmit, by teleautography, photographs as well in the form of raster drawings.

The task now was to put this teleautographic method into practical use also in the field of radiotelegraphy. Similar experiments had already been carried out before the war, though only in a crude form. So long as rapid transmission is not at issue, the transfer of the method of wire transmission to wireless transmission seems a simple matter, and similar proposals had already been made in the earlier period of the development of radiotelegraphy. Even with a coherer it was possible, by means of an electromechanical relay, to obtain marks each time,

as the pin of the receiver, touching the conducting area on the transmitting metallic film, sent waves; when the pin fell on a nonconducting area of the film, no signs were obtained. As radio engineering developed, proposals were made to use other detectors in the receiver as well, but in all cases it was necessary to use an electromechanical relay in the receiver, converting the action of the detector into visible signs on the receiving paper.

I tried to avoid such electromechanical relays, which hindered an increase in the speed of transmission, by proposing in 1910 to make the detector in the receiver act, without the aid of a relay, directly on a sensitive string galvanometer.

In the transmitter, in the primary circuit I with self-inductance 1 and capacitance 2, undamped or weakly damped oscillations are excited; self-inductance 1 is coupled with self-inductance 3, connected between the earth and the antenna. If the transmitting station is tuned to a quite definite period of oscillation, then the oscillations can be detuned by short-circuiting part 5 of self-inductance 3. This occurs every time the pin of the reproducing telegraph 4 passes over a conducting area of the transmitting film; on the contrary, the whole self-inductance is included when the pin touches the nonconducting part of the film.

Fig. 1. Diagram for explaining the basic principle of wireless teleautography.

In the receiver there is an oscillatory circuit II, consisting of capacitance 8, self-inductance 7, and detector 9. Upon the arrival of electromagnetic waves of a definite period of oscillation, a current of definite direction passes through the detector, if a path is provided for it through a circuit connected in parallel with capacitor 8. To block the path of the oscillations through the circuit parallel to the capacitance, a self-inductance (choke coil) 10 is introduced into this latter circuit. By inserting into the parallel circuit a string galvanometer of sufficient sensitivity, we shall obtain in it a deflection whenever waves of a definite period arrive; otherwise we shall obtain no deflection, and thus, just as in wire transmission, we are able here also, by means of photographic recording of the galvanometer deflections, to transfer a drawing made on the transmitting metallic film with nonconducting ink onto photographic paper or film in the receiver.

The results obtained before the war by this method were also rather crude, since the number of signals received per second was too small; amplifier tubes at that time had not yet reached a development sufficient to send through the string galvanometer in the receiver sufficiently strong currents; moreover, in the transmitter, in order to achieve detuning by means of a sliding pin, it was necessary to resort to the aid of electromagnetic relays. The speed of such relays reached, in the best case, 60–100 characters per second, so that the speed of wireless transmission in general could amount to only one tenth or one twentieth of transmission by wires. For practical purposes this made it necessary to confine oneself to the transmission of comparatively rough drawings and manuscripts, and there could be no question of transmitting, within a practically permissible interval of time, any detailed photographs in the form of raster images. But in any case the mere possibility of wireless transmission of simple drawings already had, for certain purposes, great significance—for example, for the wireless transmission of simple military-topographic sketches, and especially from aircraft. The significance of this application became clear during the World War; however, suitable apparatus was constructed only in the last year of the war.

Cathode tubes made it possible to surpass these results (in the form they had reached by 1918) and to approach the speeds of transmission by wires and even exceed them.

The point is that cathode tubes with three electrodes (a heated cathode, an anode, and a grid) are very sensitive relays, making it possible to obtain the desired action by means of small changes in the voltage on the grid, without the participation of electromechanical relays. These tubes are used, as is well known, in radiotelegraphy and telephony as modulator tubes in the transmitters of electromagnetic waves, whereby corresponding changes in the voltage applied to the grid of the modulator tube change the intensity of the waves being sent.

The simplest application of this basic idea to wireless teleautography consists in the fact that, through the pin of the teleautograph, the operating voltage is or is not applied to the grid of the modulator tube of some generator of electrical oscillations (the transmitter need not necessarily be tube-based; any generator of undamped or weakly damped oscillations may be used), depending on whether the transmitting pin touches a conducting or nonconducting point of the transmitting film.

In my apparatus for wireless teleautography, as a development of my original method shown in Fig. 1, I used the former procedure, by which the transmitted wave of a definite period is detuned or not, depending on whether it touches ...

...the stylus in the transmitter to the conducting or nonconducting part of the metallic film. The modulator tube forms part of the oscillatory circuit in the transmitter; depending on whether voltage is applied

Fig. 2. Manuscript transmitted by wireless telegraph (teleautography).

to its grid by the teleautographic transmitter or not, its properties change, and with them the properties of the parts of the oscillatory circuit that determine the period of oscillation of the electromagnetic waves.

Fig. 3. Photograph transmitted by radio (teleautography).

Only with the elimination of the electromechanical relay in the transmitter did it become possible to use the advantages of the string galvanometer as a receiver, which also operates without a mechanical relay; the former detectors are, of course, replaced by audion and amplifying arrangements.

Thus, in recent years it has become possible to transmit by radio not only manuscripts and drawings, but also photographs in the form of halftone images in as short a time as by wire. The method developed by Jenkins in the United States differs from my device for wireless teleautography only in that my method of photographic reception is somewhat modified there. Although I do not have exact data on Jenkins’s method, from brief press reports it appears that in Jenkins’s receiving apparatus the aperture through which light falls onto the photographic film of the receiver is covered by means of a device...

in the role of the instantaneous shutter of a photographic apparatus upon the arrival of waves of a definite length. In view of the fact that the string galvanometer surpasses all such shutters in sensitivity and speed of action, it cannot be admitted that, by the Jenkins method, better results of transmission could be obtained at the same (not to mention a higher) speed. Likewise, the use of an oscillograph, encountered in the receiving devices of teleautographs of other designs [Belin, Pedersen, and others], cannot compete with the string galvanometer as the best recording instrument for teleautography by wire and without wires. Synchronization by radio presents somewhat greater difficulties than synchronization by wire. At not very great distances (for example, distances within Germany between Germany and neighboring countries), the signals regulating synchronization can be made sufficiently powerful, so that synchronism does not suffer too much from atmospheric discharges or signals from other stations; but at very great distances, for example between Europe and America, maintaining synchronism by radio would prove practically very difficult.

If in recent times some experiments have been carried out by the Marconi Company and the Radio Corporation between London and New York, in which, among other things, the transmitter employed not the teleautographic method, but the method of “photoelectric elements” (to which we shall return in § 2), then there, evidently, the matter was an experiment in which no expense was spared and in which no attention was paid to the speed of transmission; my opinion is that these experiments did not have in view the direct telegraphy of drawings between Europe and America for practical purposes. Precisely the difficulty of preserving synchronism over such great distances compels us, at least temporarily, to give preference to methods of indirect transmission of drawings.

2. The influence of amplification technique on the telegraphy of drawings by wire.

The teleautographic method by wire generally does not require amplification, since here it is easy to deal with line currents of 10–20 milliamperes without the appearance on the transmitting film of excessively disturbing sparks. For very great distances (over 1500 km), at which it would be necessary to use an increased voltage for the line currents, entailing already disturbing sparks, the teleautographic method by wire is limited in its application by the capacitance of the line, which reduces the speed of transmission. If in individual cases amplifier tubes may be of use here, nevertheless their application does not signify a decisive step forward.

The situation is different in the method of photoelectric elements. Let us briefly recall that transmission here takes place in the following way: the photograph to be transmitted, in the form of a transparent film, is wound onto a glass cylinder, which can rotate and, while rotating, slowly move along its axis. The light of a bright constant source is concentrated by means of a system of lenses onto a small portion of the image, passes through the film and the glass cylinder, and falls upon a photoelectric element, which allows a current to pass to a greater or lesser degree depending on its illumination. Thus, as the cylinder rotates, the entire image is telegraphed, and it becomes possible to send continuously through the line to the receiver currents which, in their strength, correspond to the degree of transparency of the elements of the image illuminated at a given moment. In the receiver, the varying line currents arriving from the transmitter serve to recreate the image from its elements by photographic means. The method of reception with the aid of a string galvanometer, by means of which I succeeded in making the first such transmissions, still proves to be the best. The system of threads of the string galvanometer (for example, two metallic threads with a small aluminum leaf in the middle, stretched between the poles of an electromagnet, or a single thread slightly widened in the middle) serves as a shutter for the light rays entering through a narrow opening into the receiving chamber and acting there upon photographic film or paper; the latter is wound onto a cylinder rotating synchronously with the cylinder of the transmitter. The line currents enter the galvanometer and there produce a greater or lesser deflection of the system of threads, according to the transparency of the corresponding elements of the image in the transmitter, thereby causing a greater or lesser darkening of the receiving film or paper; with synchronous motion of the transmitting and receiving cylinders, the photograph sent by the transmitter with the aid of the photoelectric element is reproduced on the receiving paper or film by photographic means.

Of the photoelectric elements, formerly only selenium preparations were used (as is known, they possess the ability to diminish their electrical resistance when their surface is illuminated), since all the other photoelectric elements gave too weak an effect. I actually succeeded in making the first transmission, approximately 20 years ago, with the aid of selenium. True, selenium does not give an especially strong effect; the currents that were sent into the line were generally less than 1 milliampere, and only the string galvanometer made it possible to obtain successful results.

The state of affairs has changed recently with the appearance of amplifier tubes. On the one hand, the selenium method has received, thanks to amplification, a considerable improvement, so that transmission was no longer subject to the possibility of distortions from the action of neighboring

THE DEVELOPMENT OF IMAGE TELEGRAPHY OVER A DECADE

wires; on the other hand, other photoelectric elements could now also be used, whose effect had previously been too small for picture telegraphy.

Of primary importance are photoelements whose action is based on the effect of light on a spark discharge or on a discharge in evacuated tubes.

After Hertz first clarified in detail the influence of ultraviolet light on a spark discharge, Hallwachs found that illuminating the cathode of an evacuated tube changes the voltage on it in a noticeable way; illumination causes the emission of negative particles from the cathode, and this emission is proportional to the intensity of the illumination. The strength of the currents obtained in this way is, to be sure, only about one thousandth of the currents of a good selenium preparation, but by employing suitable tube amplifiers it is possible to obtain currents sufficient for the purposes of practical picture telegraphy. Such photoelements, to which Elster and Geitel gave a particularly suitable form, are in one respect superior to selenium preparations: they do not possess the noticeable inertia that is characteristic of selenium. The resistance of a selenium preparation that has long been exposed to illumination and is suddenly placed in darkness assumes the larger value corresponding to darkness not instantaneously; the influence of the preceding illumination still remains for some short time. It is true that this inertia is the smaller the thinner the active layer of selenium, and my first experiments succeeded only because, on the one hand, I took selenium preparations with a very thin layer and, in addition, used the following expedient: for measuring the degree of transparency I used not one selenium preparation, but two; the second preparation was compensating, its action being opposite to the action of the main preparation. By choosing the sensitivity of the main (active) preparation to be the greatest and its inertia the least, and the compensating preparation to have a somewhat lower sensitivity and a somewhat greater inertia, it was possible, in the differential action, to weaken the inertia considerably. Nevertheless, the speed of transmission was still limited by inertia, since the reduction of inertia was achieved at the cost of reducing the strength of the line current.

Therefore the use of photoelements without inertia aroused great hopes for increasing the speed of transmission, and last year experiments between Cleveland and New York, carried out with such photoelements and with colossal amplifications (these experiments were organized by the American Telegraph and Telephone Co), gave very encouraging results.

But here it is necessary to note the following: the fact that such photoelements are applicable for the purposes of picture telegraphy is by no means

with a mortal blow to the selenium method. Indeed, selenium retains the great advantage that its effect is considerably greater, and this makes it possible to avoid the difficulty of the colossal amplification required for inertialess photoelements. On the other hand, the use of amplification makes it possible to reduce significantly the inertia of selenium devices as well. I have already pointed out that their inertia is the smaller, the thinner the active layer. This, of course, increases the resistance of the selenium device and weakens the action, but amplification eliminates this defect. Furthermore, the compensation method can be made all the more effective the weaker the illumination with which we work; and again, under weak illumination, we encounter a weakening of the action, which can be eliminated by amplification.

Fig. 4. Photograph transmitted by the selenium method (direct method).

Thus, only the future will be able to decide whether selenium devices will retain their pre-eminent position or not; perhaps, even for some purposes one type of photoelement will be used, and for others—another.

It should be mentioned that radio experiments recently carried out by the Marconi Company and the Radio Corporation between England and the United States were performed with inertialess photoelements; the small changes in voltage produced at the cathode of the photoelement act on the grid of the modulator tube controlling the waves sent out by the radio station. We have no details concerning the transmitting apparatus of the Marconi Co.; in particular, the reports of the experiments available to me do not indicate whether waves of different intensity, corresponding to the different brightness of the individual points of the image, were sent, or waves of different period, as had been proposed.

The matter of radio transmission of pictures by means of photoelements is still in its initial stage; here the teleautographic method so far still has the advantage.

Let us briefly also mention the so-called relief method. In it, clichés are used in the transmitter in which the light shades are represented by greater or lesser relief. Such clichés can be obtained by copying photographs on chromogelatin, followed by washing and tanning. As in the teleautographic method, over a cliché, wound—

…onto a rotating cylinder, a pin slides, which rises and falls depending on the relief and thereby inserts a greater or lesser resistance into the line. Such transmitters using the relief method were proposed more than 30 years ago by the Americans Eaton and Amstutz, and were developed chiefly (beginning in 1907) by the French engineer Belin. In comparison with the selenium method, the relief method had the advantage that it could operate with currents of greater strength, without having to deal with inertia. Its drawback lies in the impossibility of obtaining accurate clichés in a short time and of wrapping them around the cylinder without distortion. With carefully worked clichés Belin succeeded in obtaining good transmission by using an oscillographic photographic receiver (a small mirror, rotated electromagnetically and reflecting, according to the strength of the line current, more or less light onto the photographic film of the receiver). In view of the fact that, with the use of amplifying tubes, the strength of line currents even with photoelectric cells can be brought up to a magnitude satisfying all practical requirements, the relief method can hardly constitute serious competition to the photoelectric-cell method, however interesting the corresponding devices may be in themselves.

3. Successes of indirect methods of picture telegraphy.

We shall also consider methods in which transmitting devices are used in order first to transmit an intermediate cliché adapted for telegraphic transmission, for example, a telegraphic perforated tape or a letter telegram. These letter telegrams or perforated tapes are transmitted in the usual way, by wire or wirelessly, to the receiving point, where the picture is reproduced with the aid of the letter telegram or the perforated tape.

These methods are more troublesome and more expensive than direct methods with synchronism between transmitter and receiver, in which the picture is obtained in the receiver simultaneously with its dispatch from the transmitter; but they remain the only possible ones in those cases where direct transmission of pictures is as yet impossible or unreliable, i.e., chiefly in the case of transmission over great distances, for example, from Europe to America.

I have already described in my earlier report on the development of picture telegraphy during the period 1905–1915 a method which makes it possible, with the aid of selenium, automatically to obtain for each element of the picture a letter of the Morse alphabet or a combination of holes in a telegraphic tape, as well as methods which convert such telegrams back into a picture.

This method was improved and actually applied to transmission over very great distances. I shall briefly recall that, in this method, the currents issuing from the selenium receiver (other photoelectric cells with suitable amplification may also be used), and constantly corresponding to the brightness of the given element of the picture, instead of going directly to the receiver, serve, by means of a special translating device of the fast-acting Siemens telegraph, to print letters or to punch combinations of holes in a telegraph tape—for example, the letter “a” if the given element of the picture is very bright, the letter “z” if it is completely black, while the other letters transmit intermediate shades. In this way there is formed a rather long telegram, divided into letters and lines, or a corresponding perforated tape, in which each letter or combination of holes represents a separate element of the picture with its brightness.

Fig. 5. Photograph transmitted by the selenium method (Rome—New York—1922. Indirect method by means of a letter telegram).

The telegram is transmitted in the ordinary way to the receiving point, where for reproduction it is simplest to use a typewriter, differing from an ordinary one only in that, when the corresponding key is pressed, instead of a letter a small square or rectangle is printed, the dimensions of which are different for different letters. Thus, for example, instead of the letter “z,” corresponding to the darkest element, a rectangle is printed that fills the entire space allotted in the typewriter for a letter; for the letter “y” a somewhat smaller rectangle, and so on down to a small dot; the letter “a” will correspond to a light field, equal in size to the “space” on the typewriter.

A great difficulty lay in the use of the weak currents supplied by the selenium receiver for printing letters or for punching holes. The methods that I employed in order to get around this difficulty, and without the use of amplifier tubes, I have already described earlier; and by 1915 the method had already been developed to such an extent that good results were being obtained in the laboratory. Recently experiments have appeared even over great distances: in 1922 the Italian Ministry of the Navy carried out a series of transmissions from the radio stations of Centocelle and San Paolo near Rome, in particular between Rome and Massaua and back, between Rome and ships in the Mediterranean Sea, and from Rome via Nauen—Bar Harbor to New York.

Only very recently was amplification by tubes applied in this method in order to improve the transmitter, which constituted a major step forward. On the one hand, in this way it becomes possible to prepare telegrams or perforated tapes in the transmitter more rapidly; on the other hand, this considerably weakens the inertia of selenium preparations, as follows from § 2; finally, the use of inertia-free photoelectric cells has also become possible.

This method will be important in all those cases where the direct method presents difficulties, i.e., especially over very great distances, such as, for example, between Europe and America, from these parts of the world to Japan, etc. For the transmission of telegrams one may use either radio or cable. The promising development of high-speed telegraphy, both by cable and by radio, is of enormous importance for this method. In those cases where it is desired to transmit by means of letter telegrams or perforated tape not photographs, but manuscripts or line drawings, the transmitter employs not the photoelectric-cell method, but the teleautographic method. Here, after all, it is not a question of transmitting shades of light, but exclusively of transmitting black and white; and it is always possible to take five elements of the drawing transmitted by means of the pin of the transmitter and combine them into a single combination of perforated tape with, for example, five lines, and thereby into a single letter. In the receiver it is again simplest to use a typewriter, differing from an ordinary one only in that when a key is pressed, not a letter is printed, but a combination of five dots, corresponding precisely to the combination of white and black colors of the five elements of the drawing that were taken by the pin of the receiver and expressed by the corresponding combination of holes or else by a letter.

Fig. 6. Fingerprint transmitted by radio. (Indirect method by means of a letter telegram.)

By this method, during the past year, signatures and fingerprints were transmitted between Berlin and Rome.

These indirect methods, as has already been indicated, are important for all those cases in which direct methods are impossible or do not give reliable results, i.e., for example, over very great distances, and also, for example, where the receiving apparatus cannot be installed immovably (in this case the optics of the receiver are endangered), i.e., for reception on ships on the open sea, etc.

A few more concluding remarks concerning the successes of attempts at direct vision at a distance. The difficulties inherent in this problem, as compared with the transmission of still images in the form of photographs, are as follows:

1. The need for a further, quite exceptional increase in the speed of transmission; that which until now has taken several minutes must at most take place within one-tenth of a second, and since, for the time being, such a quantity of signals cannot be sent over a single wire (even over medium distances), while simultaneous transmission by radio over a large number of waves of different lengths is likewise a very costly complication, this necessity entails, inevitably, very large expenditures on apparatus and operation if one wishes to have genuine vision at a distance. These expenditures at present must be so great that for this reason alone one cannot yet think of practical tele-vision.

2. In sending moving images, like those obtained on the ground glass of a photographic camera, much lower light intensities are available than in transmitting photographs illuminated by intense concentrated light. Of course, amplifier technique can help here, but it is necessary to take into account that, in view of the degree of amplification already required for the transmission of photographs, one cannot go too far in this respect either.

If, notwithstanding this, experiments in tele-vision have multiplied in recent years, it should be noted that in all the experiments in tele-vision reported in the newspapers [Mihály, Jenkins, Baird, and others], only the transmission of simple figures moving in the field of view is achieved, i.e., only the transmission of pictures with a small number of elements. This, of course, bypasses the essential difficulty, the overcoming of which alone could have led to practical tele-vision. Indeed, even for the transmission of facial features, from 5,000 to 10,000 elements are necessary; for groups and landscapes, of course, still more. Interest in these primitive experiments in tele-vision is, of course, nevertheless quite legitimate, even if they have not yet provided a solution to the problem of practical tele-vision.

We may say that the telegraphy of still images has already entered the realm of practical application, but that for the transmission of moving images there are no insurmountable obstacles either.

1. Die Naturwissenschaften, 13, p. 517. 1925. ↩