RESEARCH OEPARfMfMT
TELEVISION SIGNAL STORAGE USING IMAGE ICONOSCOPE
Report Mo, T~070
( 1958/22)
GcF, "ewell, A.I'o1,LEEo PcHoCc Legate Thi. Report i. the propert1 of the Briti.h BroadcastiDg CorporatioD aDd aa1 Dot be reproduoed in an1 form without the writteD perai •• ioD of the Corporation. Report No. T-070
TELEVISION SIGNAL STORAGE USING IMAGE ICONOSCOPE
Section Title Page
SUMMARY 0 0 • • • • • 0 0 • 0 C • • " • 0 0 • • 0 • 0 • , • co, 0 •• 1
1 1 INTRODUCTION .. • '" f) 0 ,., .. .. • • ., • c .. '" .. '" " () • C '" ~ r 2 DESCRIPTION OF THE SYSTEM 1 1 2.1- General. 0 0 2.2. Production of the Charge Pattern 2 2 2.3. "Reading off" the Stored Signal 0 3 2.4. Inherent Advantages of the System 3 3 PRACTICAL IMPERFECTIONS IN THE SYSTEM. 0 ••• 0 , • 0 • 0 0 •• 3.1. The Non-uniform Effect of the Potential Gradient between the Final Anode and the Mosaic Elements 3 3.2. "Whi te-crushing" • • • • • • • • • • • 3 4 3.3. Spurious Charge Patterns • e. • 3.4. Shadow Effects due to "Over-Writing" 4 4 3.5. "Black-crushing" . 0 • • • • • , 5 3.6. Possible Remedy for these Defects 3.7. Possible Damage to Photocathode • 5 5 4 DESCRIPTION OF EXPERIMENTAL SYSTEM • • 0 • • • • • • • • • • • • • • • I 8 I 5 RESULTS OF INVESTIGATION .. .. " Cl " " 0 0 ,~ • '" e .. '" e @ '" • .. ., .. " .. ( J 8 6 CONCLUSIONS 0 , • • • 0 , 0 • • 0 , • 0 0 9 7 REFERENCES () <, " 0 ::) 0 " " .. (., [} 0 f) .. " 0 " ...... G " Cl " .. 0 0 0 0 " Report No. T-070 August 1958 (1958/22) TELEVISION SIGNAL STORAGE USING IMAGE ICONOSCOPE SUMMARY This report describes an investigation of a system of television storage using the image section of an image-iconoscope camera tube. The results show that the inherent structural limitations of the image iconoscope prevent the system from having operational value. A system of this kind would probably be capable of further development, however, if a special storage tube were produced. 1. INTRODUCTION The ability to delay a television signal by the precise time duration of one field without appreciable deterioration would greatly facilitate operations such as standards conversion, telerecording and bandwidth compression. A system of storage, utilizing the afterglow of a fluorescent phosphor', 1,2 has been successfully developed for use in telerecording. Another system, which makes use of a charge pattern to control the passage of an electron beam through a fine mesh grid,3 has been developed for the storage of television pictures for periods up to many hours. Both these systems have particular applications, but neither is convenient for delaying a television waveform for one field period. This report describes an investigation into a system4 of field storage, devised by A.V. Lord, which uses the mosaic of an image iconoscope in order to store a charge pattern. 2. DESCRIPTION OF TEE SYSTEM 2.1. General The method consists of scanning the mosaic of an image-iconoscope tube with an electron beam the intensity of which is proportional to the instantaneous magnitude of the video signal to be stored. By this means, a single field is reproduced as a charge pattern on the mosaic. This pattern is reconverted to a voltage waveform by exploring it with an electron beam of constant current and taking the output voltage at the signal plate of the tube. If it is required to delay the complete video signal, it is necessary to operate two such channels, so that one is storing a field while the other is reproducing the previous field. In the particular system investigated, no use is made of the conventional electron gun of the image iconoscope; the electron beam used for "writing on" the charge pattern and, later, for "reading" it off, is obtained from the photocathode. 2 Alternate fields, say the even fields, of the picture to be stored are applied to produce a picture on the screen of a conventional cathode-ray scanning tube. An optical system is used to produce an image of the picture on the photocathode of the image iconoscope, the image section of which functions in the normal manner so that an electrical charge pattern corresponding with signal to be stored is formed on the mosaic. During the period of the odd fields, the picture is "read off" by repeating the process but with a plain "white" raster in place of picture. The mosaic is thus scanned by an electron beam of uniform current which is sufficient to stabilize the mosaic, element by element, up to the potential of the final anode. The current necessary to achieve this stabilization flows in the signal-~late circuit and con stitutes the output signal. 2.2. Production of the Charge Pattern Before the charge pattern is written on to the mosaic, it is necessary to ensure that the majority of secondary electrons emitted by the mosaic are collected by the final anode. For this purpose the mosaic is established at a potential which is negative with respect to that of the final anode. This is achieved in the following manner. In the field blanking interval preceding the field to be stored, the photo: cathode is uniformly illuminated by means of a cathode-ray flash tube and an optical system for a period of up to one millisecond. Simultaneously, the final anode is rendered negative by some twenty or thirty volts. The mosaic is flooded by electrons from the illuminated photocathode and, being unable to lose secondary electrons, acquires a uniform negative charge. Eventually this charge would equal the negative potential on the final anode and stabilization would occur. In practice, however, the magnitude of the electron beam from the photocathode and the capacity of the mosaic are such that in the one millisecond duration of the pulse and flash the mosaic potential changes by only some three to six volts. When the pulse ceases and the final anode returns to zero potential, the mosaic is left charged negatively with respect to it. Following the establishment of this potential difference the picture modulated scan commences. The optical image on the photocathode gives rise to a beam of electrons modulated in intensity in accordance with the picture information and this electron image is, in turn, focused on the mosaic. The primary photo-electrons constituting the scanning beam cause secondary electrons to be emitted by the mosaic, the number depending upon the secondary emission ratio for the mosaic surface and the energy of the primary electrons; for the con di tions under which the image iconoscope is used the ratio is approximately five to one. Following the light flash and final anode pulse, already mentioned, the mosaic is left negatively charged. The effect of bombardment by the intensity modulated electron beam is to cause each element of the mosaic to lose electrons by secondary emission in proportion to the instantaneous intensity of the electron beam. If the secondary electrons are all collected by the final anode, the charge pattern so formed on the mosaic is exactly representative of the picture to be stored. 2.3. "Reading off" the Stored Signal Following the writing-on process, the optical scan using the cathode-ray 3 tube is repeated with no intensity modulil,tion. 'rhe raster is focused on the photo- cathode as before, and gives rise to a primary electron beam of constant intensity (except for line blanking pulses), and this, in turn, is focused upon the mosaic. As the scanning beam passes over each picture element, secondary electrons are again released and are collected by the final anode. The process continues during bombardment of each element until sufficient electrons have been lost for the element to reach final-anode potential. At this point, stabilization occurs and any additional secondary electrons which are released return to the element being bombarded or to other more positive areas of the mosaic. The number of electrons leaving each element depends upon the charge on that element, so that the number of electrons leaving the mosaic varies in accordance with the picture signal. This causes a corresponding current to flow in the signal plate and load resistor, thus producing the output voltage waveform. 2.4. Inherent Advantages of the System This system has the inherent advantage that the same scanning waveform is used for both writing on and reading off the stored pattern; thus, non-linearity of the scan waveform will not cause any geometric distortion of the stored picture. Similarly, any geometric distortion produced by either the optical or electronic lens systems,5 e.g. pincushion distortion, will not cause distortion of the stored picture. The collection of secondary electrons by the final anode will, if complete, prevent the spurious charge pattern defects which normally occur with image-iconoscope camera tubes. Theile and Townsend6 have shown that some improvement in this respect has been obtained in a similar application of the image iconoscope. 3. PRACTICAL IMPERFECTIONS IN THE SYSTEM 3.1. The Non-uniform Effect of the Potential Gradient between the Final Anode and the Mosaic Elements The foregoing describes the ideal behaviour of the image iconoscope used for signal storage; in practice this is not fully achi.eved due to unsuitable features of the tube structure. The final anode consists baSically of a cylinder co-axial with, and normal to, the mosaic. Thus the potential gradient between any element of the mosaic and the final anode is a function of the distance between the element and the mosaic centre. One result of this non-uniformity is that secondary electrons emitted from the mosaic near to its periphery are more likely to be collected by the final anode than are secondaries emitted from areas near the mosaic centre and which may pass close to positively charged areas of the mosaic in the course of their journey towards the anode. 3.2. "White-crushing" The more positively charged areas of the mosaic will attract the greater --- 4 proportion of those secondary electrons which do not reach the final anode during the wr i ting-on of the charge pattern. The acquisition of these electrons will reduce the positive charge on those parts of the paiTiTern representative of the white parts of the picture being stored. This defect becomes more apparent as the charge intensity of the stored pattern is increased. 3.3. Spurious Charge Patterns The normal "tilt and bend" phenomenon is present to some extent because the collection of secondary electrons by the final anode is incomplete. Another spurious pattern, which is more difficult to correct, takes the form of a diffuse dark ring concentric with the mosaic. The distance that a secondary electron will travel before being recaptured by a positively charged area of the mosaic depends on the relative potentials over the mosaic surface and on the initial velocity of the electron. The maximum potential difference between "black" and "white" areas of the charge pattern is in practice not more than one or two volts. The majority of electrons have initial velocities in excess of this, so they will not in general be recaptured immediately. The final anode has a relative potential of three to six volts, thus en~ouraging a radial course for the electrons. The central area of the mosaic will suffer least from the acquisition of free secondary electrons and the surrounding areas will suffer to an increasing extent with increasing distance from the centre unti~near the mosaic periphery, the final-anode potential takes control and attracts most of the remaining free secondaries. It is this which results in the stored picture having a super imposed dark ring which tends to decrease in outside diameter as the poiTential of the final anode is increased relative to the potential of the charged mosaic and to the initial velocity of the secondaries. 3.4. Shadow Effects due to "Over-Writing" Another form of spurious patterning appears as white "shadows" above and radially ouiTwards from the outer edges of black areas in iThe upper half of the picture. This is due to the fact that these areas suffer less from "white-crushing" than most of the white areas. Electrons which have traversed an area having 3ero charge and then pass over a positively charged area will be attracted towards the mosaic surface but approach in a curved trajectory. Therefore few electrons will be captured by the positively charged strip immediately beyond an uncharged area. The effect of the potential of this strip is to cause secondaries to be captured by positively charged areas further along the electron path. 3.5. "Black-crushing" In addi iTion to the "white-crushing" effects described above, which occur during the writing process, a second type of amplitude distortion occurs during the reading process. This gives rise to "black-crushing" of iThe stored picture. Ideally, as each picture element is read off, iThe appropriate elemeniT of the mosaic is bombarded wiiTh a beam of primary eleciTrons and releases some five secondaries for each primary received until the positive charge on that element has been increased so that its pOiTential is equal to that of the final anode, when 5 all further secondaries released should return to the element, except for a sufficient number to compensate for the primaries being received. The current flowing in the capaci ty formed by the element of the mosaic and the signal plate is then proportional to the difference in potential between the initially charged element and the finally discharged or stabilized element. In practice, if the initial velocity of a second&y electron is less than the difference in potential between emitting element and a neighbouring element, which has been stabili zed at the final anode potential, it will not escape from the mosaic. Thus, although the element being read may be di scharged to the final~anode potential, the resulting current flowing in the signal plate will be less than proportional to the change of charge because other parts of the mosaic will acquire an opposite charge. This effect is greatest in the uncharged areas of the stored pattern because of the greater potential difference between the element about to be read and the last one to have been read. Electrons emitted from these areas require greater velocities, in order to escape from the mosaic, than electrons emitted from positively charged areas; thus a smaller pro portion of electrons escapes from the relatively uncharged areas representative of black and dark grey shades of the picture. The resulting compression of grey scale in the region near black becomes more serious the greater the potential difference between the final anode and the uncharged areas of the mosaic. 3.6. Possible Remedy for these Defects If the final anode were connected to a fine mesh, or grid immediately in front of (on the photocathode side) the mosaic, the non~uniformity of the potential gradient would be removed and the effective potential gradient between each element of the mosaic and the final anode would be greatly increased. This would assist secondary electrons to escape from the mosaic> thereby preventing the spurious patterning and the "gamma" distortion described above. 3.7. Possible Damage to Photocathode The photocathode current for optimum operation may be in the order of 20 to 30 microamps uniformly distributed over the surface for a time duration of from 700 microseconds to 1"0 millisecond; during the read and write periods, a current of up to O~ 5 microamps may be emitted from each part of the photocathode in turn. It is possible, with some image-iconoscope tube8, to damage the photocathode, (appar ently by overheating), and to produce a patch of low emissive surface in the centre of the photocathode. This results in a diffuse white spot near the centre of the stored picture. It was found possible to prevent this damage by blowing cool air on to the outer surface of the glass plate, on the inner surface of which is deposited the photocathode. If a special camera tube were developed for this storage application~ it might be that a more dense de~osit could be used for the photocathode in order to improve the thermal conductivity at the expense of photosensitivity. 4. DESCRIPTION OF EXPERIMENTAL SYSTEM A schematic diagram of the experiment is shown in Fig. 1. The camera ... 6 Display on scanner; 25 cl. gate, line alternate frames of lilt and bend picture and plain mixi"9. blanking roster of controll a b le mixing and black brightness. level clamp. Ima9~ iconoscop~. Field Line 25c/. Line sync. sync. SLr sync. 25 cls light flashes of controllable duration .intenslty and phase. Field Line Negative Video. sync. sync. going lSe/. pulses. Po.itive going 2Sc. ulse •. Divider and pulse gen. Com lete blankin Mixed s nes. Fig. I - Block schematic diagram of apparatus tube is a Photicon in which only the image section is used, the electron gun being discarded and its scanning coils being disconnected to prevent undesirable induction effects. The potential between the final anode and the photocathode is 1'0 kV and the mean current is in the order of 0'1 to O' 2 fJ- A. The lens is a Taylor, Taylor and Hobson f/1'4 lens used with an aperture of f/2'0; the lens causes a loss of upper-frequency detail amounting to less than 3'0 dB at 3'0 Mc/s. The cathode-ray tube used for the production of the writing and reading scans is an E.M.I. type R5161 tube with a zinc-oxide phosphor producing a green yellow light. It iE' used with an 18 kV e.h. t. supply and, with a raster of uniform brightness, produces an illumination at the photocathode of the Photicon of 10 foot lamberts. A 45° dichroic mirror capable of green transmission and blue-red reflec tion was interposed between the cathode-ray tube and the lens. This mirror enabled the photocathode to be flooded, when required, by the flash light source -- a Ferranti type CL61 cathode-ray tube used with a 20 kV e.h.t. supply and producing a blue light. A double condenser lens was used between this light source and the 45° mirror to ensure maximum light collect ion by the photocathode. Fig. 2 shows the waveforms relevant to this system. 7 Writ. period Read period. 2 3 -~ -t ..-900...... I I , I • I I I : I 4 r------4 ~----.------lI ~ 7ao;...s 6qu.s-i!=f'--60fL1 5 _Time. Fig. 2 - Waveforms used in the system 1 Video waveform (fields) at input to scanner video amp. 2 Video waveform as applied to scanner C.R.T. 3 Waveform applied to final anode of image iconoscope ~ Waveform on grid of light-flash C.R.T. 5 Output from signal plate across 100 k resistor The photocathode current during the flash period amounts to between 20 and 30 fJ-A for a period of between 700 fJ-sec and 1'0 millisec. Assuming that the mosaic-to-signal-plate capacity is 5000 .10 - 12 F, and that the entire mosaic is used, the charqe obtained in the pulse-and-flash period gives rise to a potential of between 3'0 and 4'0 volts. In point of fact, since only about half the mosaic is used, and because only one field of an interlaced picture is stored at one time, the charge may be as high as 6'0 to 8'0 volts. It was found that only half the mosaic is used, by noting that reduction of the vertical scan amplitude to one half its normal value could be effected without loss of definition or intensity of the stored picture. The charge pattern left by the writing process appears to have a value of peak potential of 2'0 to 3'0 volts for optimum operation. 8 5. RESULTS OF INVESTIGATION The photo~raphs in Fi~s. 3 and 4 show respectively the input pictures and stored pictures for a test card and a sin~le frame of a motion picture film. The defects described in Section 3 are apparent to Some extent in these photo~raphs althou~h some de~radation has occurred in the photo~raphic process. The photo~raphs were taken when the apparatus had been adjusted to produce the best conditions of si~al-to-noise ratio consistent with the avoidance of excessive spurious patternin~. No attempt was made to correct the "~ammall of the stored picture as it was felt that the other defects were in their own ri~ht sufficiently serious to mar the resultin~ picture. The optimum adjustment resulted in an amplitude/fre~uency characteristic which was 6 dB down at 2'5 Mc/s and 8'5 dB down at 3'0 Mc/s (relative to the response below 1'0 Mc/s), with a ratio of picture si~nal to r.m.s. noise of 32 dB. E~uali zation of the fre~uency response could be achieved at the expense of a decreased si~al-to-noise ratio. It was found that these optimum conditions re~uired critical and continuous adjustment if the picture bein~ stored had a continually ~aryin~ mean bri~htness (as in the case of newsreel films, etc.). The expected independence from scan non-linearity was found to be very marked and this effect, added to the fact that it was found unnecessary to superimpose exactly the write and read scans, results in a remarkable freedom from the extreme precision normally re~uired in the en~ineerin~ of this type of apparatus. The apparent lack of exact superposition of the successive scans must be due to the readin~ beam of electrons bein~ pulled into coincidence with the track of the writin~ beam by the char~e potential left by the latter. As an experiment the successive interlaced scans of a 405-linesystem were used for writing and readin~. This arrangement showed no disadvanta~e over the use of superimposed identical scans, presumably because the interlace was not exact and "beam pullin~n to the nearer of the two adjacent lines of the char~e pattern was unambi~uous. When the normal spacin~ between successive lines was increased by 50%, by increasin~ the field scan amplitude, uncertainty of the "beam pullin~" occurred and details of the stored picture jumped erratically by one line; further spacing of the lines resulted in a deterioration of the intensity and detail of the stored picture. 6. CONCLUSIONS The system of storage investigated suffered from defects resultin~ from the unsuitability of the internal structure of the image-iconoscope tube. These defects are such that the system is ~uite unsuitable for operational use. If, however, a special tube were developed for this application,~it appears that a system of storage mi~ht be d,eveloped which would have remarkable freedom from geometric distortion without the re~uirement of hi~h precision of construction or adjustment. 9 7. REFERENCES 1. "Considerations sur le fonctionnement des vidigraphes", Angel, Y., L'Onde Electrique, Vol. 34, No. 333, December 1954. 2. Report on Visit to Milan, December 1955, "An Appraisal of the French 'Stored Field' Method of Telerecording", Maurice, R.D.A. and Rainger, P., B.B.C. Engineering Division Report, March 1956. 3. "The Recording Storage Tube", Hergenrother, R.C. and Gardner, B.C., Proc. r.R.E., Vol. 38, No. 7, July 1950. 4. "Improvements in and Relating to Electronic Information Storage", Lord A.V., British Patent Application No. 1126/56. 5. "Improvements in or Relating to Television Transmitting Apparatus", Weighton and Pye Ltd., British Patent Specification 656,069. 6. "Improvements in Image Ieonoscopes by Pulse Biasing the Storage Surface", Theile, R., and Townsend, F.H., Proe. I.R.E., Vol. 40, No. 2, February1952. MY Fig. 3(A) - Input picture of test card Fig. 3(8) - Stored picture of test card Fig. ~(A) - Input picture (one frame of film) Fig. ~(B) - Stored picture of Fig. ~(A)