795,193. Television. BRITISH BROADCASTING CORPORATION. Jan. 4, 1956 [Jan. 4, 1955], No. 242/55. Class 40 (3). A method of generating a television signal comprises the production of at least one train of regularly-recurring pulses which are modulated independently in accordance with brightness values and the duration of such brightness values. By suitably selecting the pulse recurrence frequency a signal of smaller band-width than conventional methods allow is derived. In one embodiment (Fig. 1), utilizing two separate pulse trains, which are modulated respectively in accordance with the brightness and duration values, a video signal, Fig. 2 (a) (derived by conventional methods) is supplied to a sampler 11 operating at a frequency 2 (f) (where f is the desired reduced band-width) and supplies samples to a quantizer 13 which produces the signal, Fig. 2 (b). The latter signal is then fed to differentiator 14 producing the pulse train, Fig. 3a, the amplitudes and time positions of which represent, respectively, the changes in amplitudes of the quantized waveform and the times of occurrence of these changes. Although this pulse train contains no " redundant " information, since pulses only occur when there are changes in amplitude of more than a predetermined minimum value, the minimum bandwidth required for the transmission of this pulse train is still f because the maximum rate at which information occurs has not been changed. To reduce this rate the time spacing of the pulses is equalized in the following manner. The pulses (Fig. 3a) are firstly rectified in 15 and amplitude limited in 16, producing the train (termed " position markers ") shown in Fig. 3 (b) and these pulses are then employed to control a sawtooth wave generator 17 which produced the wave shown in Fig. 3 (c) in which the maximum amplitude (at the commencement of the flyback period) is a measure of the duration of the brightness level at the time of occurrence of the " position marker " pulse preceding that which produces the flyback. The pulse waveform (Fig. 3 (c)) is then differentiated in 18 to produce the train shown in Fig. 4 (a) (termed " position pulses ") whose amplitudes are representative of the durations of brightness levels and the " position marker " pulses (Fig. 3b) are employed to control a gate circuit 19, to which the video waveform, Fig. 2a, is supplied, to gate out pulses (termed " brightness pulses " and shown in Fig. 4 (b)), the amplitudes of which are representative of the actual brightness values. The pulse trains of Figs. 4 (a) and 4 (b) are then supplied to pulse spacing equalizer circuits 21 and 20 respectively which produce the equal spaced equivalent pulse trains shown in Figs. 5 (a) and 5 (b). One form of pulse spacing equalizer circuit is shown in Fig. 6 and comprises two storage type signal generating tubes (e.g. iconoscopes) 22, 23, the pulse train to be space-equalized (supplied via terminal 24) being applied, alternately, to the modulating electrode of each tube via a change-over switch 29 controlled by frame synchronizing pulses from terminal 32. During the storage frame period for each tube (the writing period) the horizontal scan is controlled by a time-base 26 operating at normal line rate and fed to the tube deflecting electrodes via a switch 28 also controlled by frame synchronizing pulses. During alternate frame periods switches 28 and 29 " change-over " and, whilst pulse information is being stored on one of the tubes 22, 23, the line scanning rate of the other tube is increased by switching in a fast time-base 27 which effects " reading " of the information already stored and, at the same time, a switch 31, also operated by frame synchronizing signals, connects the signal output electrode of the tube being " read " to a R.C. output circuit 30, 35. During the " reading period, when the " fast beam reaches the first stored pulse the condenser 35 is charged to a potential representative of this pulse and the voltage developed across resistor 30 is employed to control a trigger circuit 36 which arrests the fast time-base 27. At the same time, the pulses being stored on the other tube are also fed to a counter circuit 33, the output of which is a voltage dependent on the number of pulses counted in one frame period and this voltage is employed to control a pulse generator 34 which generates regularly recurring pulses at a recurrence period equal to the average time-spacing of the pulses counted by 33. The pulses from generator 34 then operate switch 37 to release the charge stored on condenser 35 to the output 38 and at the same time operate 36 to release the beam which again scans at the " fast " rate. This operation continues until all the stored information on one tube has been " read " the switches 28, 29 and 37 then " changing-over " to effect " reading " of the information stored on the other tube. In this manner the pulse spacing equalizer circuits 21, 20 (each similar to Fig. 7) convert the pulse trains of Figs. 4 (a) and 4 (b) into the equal spaced equivalent pulse trains of Figs. 5a and 5b respectively. Since the " position - pulses " (Fig. 5 (a)) indicate the duration of the brightness level represented by the preceding " brightness-pulse " (Fig. 5 (b)), the latter pulses are delayed by one recurrence period in the arrangement shown in Fig. 7 in which the pulses to be delayed are supplied, via terminal 40, to charge one of condensers 43, 44, the charge on which is passed to the output 48 on the occurrence of the next pulse which also effects change over of the switches 42, 47 and is itself stored on the other of the condensers 43, 44 until the arrival of a further pulse when the device reverts to the position shown. This delay arrangement may be employed either at the transmitter or at the receiver. The pulse trains of Figs. 5 (a) and 5 (b) (or Fig. 5 (b) as delayed by the arrangement of Fig. 7) may be transmitted by filtering each train to produce two continuous signals (of appropriate band-width) which are then modulated on to quadrature phases of a carrier wave, or, alternatively, the pulse trains of Figs. 5 (a) and 5 (b) may be time-interlaced, the " brightness-pulses " (Fig. 5 (b)) being delayed by onehalf the recurrence period and at the receiver, after separation of the trains, delaying the " brightness-pulses " by a further half recurrence period. At a receiver suitable for the reception of signals transmitted by the latter method the output of the detector 50 (Fig. 8) is supplied to a gated amplifier 51 controlled by gating pulses (from a generator 53), having a frequency f equal to the recurrence frequency of the pulses in one train, the gating pulses being supplied to amplifier 51 directly and via a delay device 55 introducing a delay of one-half the pulse recurrence period to gate out the " brightness-pulses " to a waveform integrator 57 and the " position-pulses " to a waveform integrator 59. These integrators cause each pulse to be maintained over the interval between it and the next pulse, the output from the former being supplied to the intensity modulating electrode of the picture reproducing tube 58 and that from the latter being supplied to a triggered integrator 60 which varies the rate of change of the line deflecting voltage 57, tube 58, in proportion to the output of integrator 59. In an alternative transmission system, two further pulse trains representative respectively of the sum and the difference of the " position " and " brightness " pulse trains are generated and after transmission in any desired manner the " position " and " brightness " are derived by addition and subtraction of the received signals, the arrangement giving an increased signal-to-noise ratio in comparison with other transmission arrangements. In an alternative way of carrying out the invention the video signal 65 (Fig. 9 (a)) is quantized, without sampling, to produce the waveform 66 which is then differentiated to produce a pulse train of equal amplitude but varying time-spaced pulses from which (by the arrangement described in connection with Figs. 3 to 6) a train of uniformly spaced pulses (Fig. 9 (b)) is produced. Each of these pulses indicates by its polarity an increase or decrease in brightness of one quantized step and by its amplitude the time for which the (new) brightness must be maintained. At a receiver suitable for receiving the latter type of signal the received pulses are gated out by an amplifier 67 (Fig. 10) and after limiting of both polarities in 70, are employed in an integrator 71 to produce a waveform of brightness increments which are supplied to the modulating electrode of the picture-reproducing tube 58. The gated-out pulses from 67 are also supplied via a full-wave rectifier 72 to a waveform integrator 73, the output from which controls the line deflection of tube 58 in similar manner to the elements 59, 60, of Fig. 8.