GB2225189A - Process and device for recalibration of the colour burst signals of a colour television video signal - Google Patents

Description

1 2'4125189 Process and device for recalibration of the colour burst

signals of a colour television video signal In order to transmit images, the PAL or SECAM colour television systems use a composite video signal which comprises, behind each line synchronisation pulse, during the period of the line blanking, a pure sine wave signal termed a colour burst signal with the initial frequency and initial phase of the colour sub-carrier subsequently modulated in the useful part of the line by the chrominance signal. This colour burst signal is usedin receivers at the start of each line to synchronise a sub- carrier for demodulation of the chrominance signals.

In order for the colour burst signals to enable receivers correctly to recover, in all circumstances, the phase at the origin of the colour subcarrier, it is necessary for the bursts to conform to the specifications contained in the CCIR Report 624 particularly as regards the time of their appearance with respect to the start of the line synchronisation pulse and for them to be formed from a pure sine wave signal, without phase breakdown. However, it happens, increasingly often in practice, that the video signal is subjected, before arriving at a receiver, to multiple treatments: recording on magnetic medium. scrambling and unscrambling for encrypted television broadcasting... which degrade it particularly at the level of the colour burst signals to the point of no longer enabling certain receivers correctly to recover the colour subcarrier phase at the start of a line, and this shows itself through very unsightly fringe effects on the left edge of the screen.

These fringes appear particularly in a SECAM system with teleprojection. sets during broadcasts transmitted by means of a composite video signal encrypted by delays of fixed values displacing, under the control of a particular code, the useful part of each line with respect to the line synchronisation pulse. In fact, to produce this encryption, it is usual to select a non-delayed version of the composite video signal at the appearance of each line synchronisation pulse and to switch over at the end of a colour burst, before the start of the useful video part, to a version having the required 2 delay and these handling operations result in a lengthening of the colour burst signal by the duration of the delay with a phase jump at the switch- over responsible for the fringes which move on the left edge of the screen at the rate of the coding.

The aim of this invention is to eliminate these disadvantages by restoring both the positions of the colour burst signals with respect to the start of the line synchronisation pulse and their spectral purity.

The object of the invention is a process for recalibrating the colour burst signals of a colour television video signal consisting of:

- sampling the video signal, memorising the video signal samples in two memory planes with random addressing operating in parallel one with the writing and the other with the reading, their roles being exchanged at the end of a period of time corresponding to the duration of one line and reading of the samples in each memory plane following the chronological sequence of their writing with the exception of those appearing over a transitory period of time straddling the nominal start of the colour burst signal which are replaced, for the part preceding the nominal start of the colour burst signal, by a re-reading of immediately preceding samples and, for the part following the nominal start of the colour burst signal, by a re-reading of one or several groups of samples following one another in the chronological writing sequence, groups which are chosen from the burst, subsequently to the transitory period of time, on the end of a sequence of samples covering a whole number of periods of the sine wave signal of the burst and linking up in phase with the start of the s_. quence of samples of the burst following on from the transitory period.

A further object of the invention is a process of the above mentioned type used in the context of the encoding or the decoding of a composite video signal the colour burst signal of which is lengthened as far as the useful part and which is encrypted by delays of fixed values displacing, to suit a particular code, the useful part of each video line with respect to the line synchronisation pulse, Q 3 further consisting, during the reading of the samples in the memory planes, of lengthening during its reconstitution the said transitory period by a number of samples corresponding to the duration of the delay to be added for the encoding or the decoding.

It also relates to a device for implementing the abovementioned processes comprising an input circuit performing a digital sampling of the video signal to be recalibrated, two memory planes with random addressing memorising the digital samples delivered by the input circuit and working in parallel, one with the reading and the other with the writing, their roles being exchanged at the end of a period of time corresponding with the duration of a video line, and an output circuit restoring to analog form the digital samples coming from the reading of the memory planes. Each memory plane of this device encloses a random access memory and a pre-positionable address counter stepped by the rate of sampling of the input circuit and fitted with a pre-positioning loop enabling it to jump addresses when reading with respect to the writing sequence in order to reconfigure the signal in the transitory period of time straddling the nominal start of the colour burst signal by re-readings of samples preceding and following this transitory period of time.

The pre-positioning loop of the address counter of each memory plane advantageously comprises three read-only memories the output signals from which are individually resynchronised on the sampling rate by banks of registers, a first read-onlv memory decoding the starts of the address jumps and being addressed directly by the address counter, a second readonly memory decoding the conditions authorising address jumps, being addressed both by the data reading signalof the first read-only memory and by synchronisation signals coming from the-input circuit and delivering a control signal for pre-positioning of the address counter, and a third readonly memory decoding the destination addresses of the Jumps, being addressed in the same manner as the second read-only memory anddelivering pre-positioning addresses to the address counter.

4 Other characteristics and advantages of the invention will emerge from the following description of a mode of implementation given by way of example. This description will be given with reference to the drawing in which:

- figure 1 shows the general diagram-of a digital encoding circuit adapted to the implementation of the process for recalibration according- to the invention and to the encryption of a colour television video signal, figure 2 shows the general diagram of a decoding circuit which is adapted to the encoding circuit of figure 1 and also implements the recalibration process according to the invention, - figure 3 gives details of the constitution of memory planes used in the encoding and decoding circuits of figures 1 and 2, figures 4, 5 and 6 are block diagrams setting out the operation of the memory planes shown in figure 3, - and figures 7, 8, 9, 10, 11 and 12 are block diagrams showing in time the transformations undergone by the video signal lines in application of the recalibration process and optionally of the encoding and decoding.

The implementation of the process according to the invention will be described in the context of a system for encrypted transmission of a 625 line colour television video signal using a scrambling system consisting of offsets of the useful part of each line with respect to the line synchronisation pulse, of a value of 0, R or 2R depending on a pseudorandom coding the keys to which are known at the transmission and at the reception alike.

Figure 1 shows the general diagram of a digital encoding circuit for supplying the scrambled composite video signal use(-' during the transmission. The following may be seen:

an input circuit 100 receiving the composite video signal to be scrambled, transforming it into a sequence of digital samples and extracting from it different synchronisation signals, - a code generator 200 which is controlled by means of an operating terminal 210 and which delivers signals identifying the value of the delay to be imposed on the useful part of the video signal line being prepared, - two identical memory planes 250, 300, with their own random addressing circuit, each capable of storing a number of samples corresponding to a complete video line, the one operating in parallel with the writing and the other with the reading, their roles being exchanged at the rate of succession of the video lines, and performing the resetting of the colour burst signals and the introduction of the encoding offsets by a reading addressing different from the writing addressing, performed under the control of the input circuit synchronisation signals and of the delay identification signals of the code generator 200 and - an output circuit 400 putting back into analog form the succession of digital samples delivered by the memory planes 250, 300 and incorporating in the composite video signal information delivered by the code generator 200.

The input circuit 100 comprises a channel for processing of the video signal formed by an analog-digital converter 101 preceded by,a circuit for restitution of the d.c. Component 102 known as a clamping circuit, and an anti-overlapping low-pass filter 103, and also a channel for extraction and recalibration of the synchronisation signals composed of a synchronisation extraction circuit 104 supplying the different line synchronisation SyL, even frame synchronisation SyT, and odd frame synchronisation SyM signals, of an oscillator with phase locking loop 105 supplying the sampling frequency set on the line frequency and also a line start signal PL, of a colour burst detection circuit 106 and of a circuit for controlling the two memory planes 250, 300 when writing or reading.

The antioverlapping low-pass filter 103 limits above 6 MHz the frequency band occupied by the input video signal and prevents the energy in the frequency band above 6 MHz from being superimposed on it by the phenomena of frequency spectrum foldover occurring during the subsequent sampling.

The circuit for restitution of the d.c. component operates, as known, by realignment on the black suppression plateau immediately following the line synchronisation pulse, by means of a clamp pulse centred on this plateau which is supplied by the synchronisation 6 extraction circuit 104.

The synchronisation extraction circuit 104 performs the separation of the video modulation and of the synchronisation signals, sorts the line synchronisation pulses SyL, even frame synchronisation pulses SyT, odd frame synchronisation pulses SyM and produces the clamp pulses ic as in any television receiver. Details will not be given of its structure since this is conventional and does not form part of the invention.

The oscillator with phase locking loop 105 comprises a voltage-controlled quartz oscillator 110 fitted with a phase locking loop formed by a lowpass filter 111 with a phase comparator 112 controlling via the low-pass filter 111 the phase and frequency control input of the oscillator 110, with two frequency dividers 113, 114 placed in succession between the output from the oscillator 110 and one of the inputs of the phase comparator 112 and with a third divider 115 connecting the other input of the phase comparator 112 to the line synchronisation output of the synchronisation extraction circuit 104.

The voltage-controlled oscillator 110 synchronised on the line synchronisation signal SyL of the synchronisation extraction circuit 104 delivers towards the analog-digital converter 101 and towards the memory planes 250, 300, a point clock signal H at the sampling frequency of 17. 734 MHz higher than twice the highest frequency of the video signal (6 MHz), and equal to the 1135th harmonic of the line frequency (15625 Hz).

The frequency divider 113 divides the frequency of the oscillator 110 by 1135 to bring it back to equality with that of the line synchronisation signal SyL. The two frequency dividers 114 and 115 are dividers by two enabling symmetrical rectangular signals to be launched at the comparator 112. The signal delivered by the divider by 1135 is a very stable signal, at the line frequency. free of the dancing which can affect the line synchronisation signal SyL delivered by the synchronisation extraction circuit 104. It cannot however be used as it is, in the place of the latter since it 1 7 has,with respect to the latter.a residual phase displacement imposed by the phase comparator when the phase locking loop is at its point of equilibrium. This residual phase displacement is corrected by a digital recalibration circuit 116 constituted by a divider-by-1135 counter pre- positionable at an adjustable value, which is re-positioned by the rising fronts of the output signal from the frequency divider 113, the capacity overflow pulses of which constitute a line start signal PL synchronous with the mean position of the rising fronts of the pulses of the line synchronisation signal SyL.

The burst colour extraction circuit 106 comprises a lowpass filter 120 which is connected at the output from the clamp circuit 102 and centered on the frequency of 4.406 MHz of the subcarrier of the red colour component of the SECAM system available in the nonmodulated state during the colour burst signal corresponding to this component. This low-pass filter 120 is followed by two samplerblockers 121, 122. The first sampler 121 is triggered by the clamp pulse appearing during the black level suppression inside the colour burst signal and delivers a binary value indicating whether or not the low-pass filter 120 has isolated the frequency of the red colour component sub-carrier. The second samplerblocker 122 is triggered by the line start signal PL and memorises, over the duration of a video signal line, the value delivered by the first samplerblocker 121 which constitutes a binary signal Dr representative of the nature of the colour component of the video signal line being prepared by reading of one of the memory planes 250, 300.

Details will not be given of the code generator 200 since it does not form part of the invention. It comprises essentially a pseudo-random binary sequence generator from which are taken as outputs, towards the memory planes 250, 300, two items of binary information Cl, C2 determining the value of the delay 0, R, 2R to be assigned to the useful part of the video line during reading and an auxiliary micro-processor type circuit performing the initial loading operations of the pseudo- random binary sequence generator, the introduction of encoding keys into the frame suppression lines 310 and 622 of the video signal coming from the memory planes 250, 300 by means of a multiplexer 401 forming part of the output circuit 400 and the handling

8 of a signal transport with the terminal 210 supervising the encoding. It receives from the synchronisation extraction circilit 104 the line SyL and frame SyT, SyM synchronisation signals enabling it to identify the frame suppression lines 310 and 622.

The output circuit 400 comprises the above-mentioned multiplexer 401 having two parallel inputs one connected to the reading output of the memory planes 250, 300 and the other to an output from the code generator 200, a digital-analog converter 402 and an interpolating low-pass filter 403.

The memory planes 250 and 300 of identical structures, details of which will be given later with reference to figure 3 receive, in addition to the point clock frequency H, the line start signal PL, the burst colour signal Dr and the binary encoding information Cl, C2, complementary writing reading orders W, W taken from the terminals of a switch 351 connected to the output from a bi-.:ttable flipflop 350 actuated by the line start signal PL so as to exchange their writing reading roles at each line change of the video signal.

Figure 2 gives the general diagram of a digital decoding circuit adapted to the digital encoding circuit which has just been described. The decoding circuit repeats many elements of the encoding circuit which have been marked with the same references to which a prime is added. It differs from it essentially by the code generator which is replaced by a reverse code generator 220 and by the output circuit 400' which no longer encloses any multiplexer, which becomes pointless.

The reverse code generator 220 delivers to the memory planes 250', 300' binary decoding --'nformation C'l, V2 representing the difference with respect to 2R of the value 0, R, 2R of the encoding delay, which difference must be added to the delay assigned to the useful part of the video line during reading in the memory planes 250', 300' so that the useful parts of all the video lines have the same delay of value 2R and recompose a clear image. No more details of it will be given than of the code generator 200 since it does not form part of the invention. It comprises essentialy 9 a pseudo-random binary sequence generator, of the same type as the code generator 200, from which are taken the items of information Cl, C2 which are transformed into items of information C'l, C'2 for example by simple reversal if the information Cl, C2 encodes the value of the delay in natural binary, and an auxiliary microprocessortype circuit performing the initial loadings of the ppeudo-random binary sequence generator using local information received via a keyboard 230, and encoding keys written into the frame blanking lines 310 and 622 which it isolates in the analog video signal reaching it from the input circuit 100' by means of the line SyL and-frame S-vT, SyM synchronisation signals delivered by this same input circuit 100'.

The memory planes 250, 300, 2.50' and 3001 comprise a fast store capable of storing the 1135 digital samples of a video line which are supplied to it at the frequency of 17.735 MHz by the analogdigital converter of the input circuit and an addressing circuit permitting reading in an order differing from that of writing by optionally repeated address jumps.

Figure 3 gives details of the structure of one of these memory planes. It shows a fast store 301 having separate input and output of 8 parallel bits, interposed between two banks of synchronisation registers 302, 303 stepped by the point clock signal H and a bank of unidirectional function amplification circuits 304 having a high impedence state control buffering'its output to enable it to be placed in parallel on the output from the store of another memory plane. This fast store 301 is addressed by means of a prepositionable counter 305 stepped by the point clock signal H and fitted with a prepositioning loop comprising three read-only memories 306, 307, 308 placed in cascade with interposed banks of synchronisation registers 309, 310, 311.

The first read-only memory decodes the starting addresses of any jumps. It is directly addressed by the address counter 305 and delivers on its output a signal encoding on several parallel bits t_ne appearances of these jump start addresses on the output of the address counter 305.

The second read-onlY memory 307 decodes the conditions of an address jump and monitors the prepositioning com,..an-- of the address counter 305. To perform this function, it is addressed by the items of information Cl, C2 or C'l, C'2 on the delay to be introduced into the useful part of the video signal, by an SIC information on the SECAM or PAL system in operation, by the Dr information. indicating in the context of a SECAM system, the red or blue nature of the colour component of the video line being read in the store 301,-by the line start signal PL, by the output signal from the first read-only memory 306 and by the reading writing signal W which come to it with a delay of one sampling point via the bank of synchronisation registers 309 and also by four bits of its output signal looped back via a bank of synchronisation registers 310 enabling it to monitor repetitive jumps.

The third read only memory 308 memorises the destination addresses of the jumps and monitors the parallel loading input of the address counter 305 via a bank of synchronisation registers 311. It is addressed like the preceding memory by the items of information Cl and C2 or C'1 and C'2, SIC, Dr, PL and by the output signal from the first read-only memory 306 which reaches it with a delay of one sampling point via the bank of synchronisation registers 309.

The pulse of the line start signal PL addresses in the second read-only memory 307 a field of data written with commands for prepositioning of the address counter 305 and in the third read-only memory 308 a field of data of zero value so that the address counter 305 is re-set at zero at each start of a video line.

The writing control signal W or its complement W for the dual memory plane, when they are in the high state, put the store 3011 into writing and the bank of function amplification circuits 304 into the high output impedence state while addressing in the second read-only memory 307 a field of data devoid of any prepositioning order so that the address counter 305 is incremented regularly without making any address jumps during the writing operations of the store 301.

In the low state, the writing control signal W or it complement W for the dual memory plane put the store 301 into readinz and the bank of function amplifier circuits 304 into the llow output impedence state. In addition, in the second and third read-onl,,memories 307, 308, they address fields of data written with Pre

11 positioning orders and jump destination addresses enabling the block diagrams of figures 4 and 5 to be implemented in the case of a SECAM system and that of figure 6 in the case of a PAL system, these block diagrams making it possible, as will be seen below, to reconstruct the start of the colour burst signal without causing a phase discontinuity with the rest of the burst to appear and also, in the SECAM system, to take advantage of this reconstruction to introduce the delay necessary for the encoding or for the decoding into the video line.

The reconstruction of the start of a colour burst signal is done by disregarding the digital samples of the,video line appearing over a transitory period of time straddling the nominal start of the colour burst signal and by replacing them for the part preceding the nominal start of the colour burst signal by immediately preceding samples read once again by means of backward addressing jumps, and for the part following the nominal start of the colour burst signal, by a group of samples following one another in chronological order, chosen from the colour burst signal following the transitory period, on the end of a sequence of samples of the burst sine wave signal linking up in phase with the start of the sequence of samples of the burst sine wave signal appearing after the transitory period, the choice of the group of samples being made by forward addressing jumps and the linking up at the end of the transitory period by a backward addressing jump.

Figure 7 shows the process of reconstruction of the part of the transitory period of time preceding the nominal start of the burst. The curve a 7 represents the start of a SECAM or PAL system video line with the leading front of the line synchronisation pulse OH used as origin of the times. The start of the black level suppression appears at the end of 4. 7ps and that of the colour burst signal at the end of 5.58 ps. The cross line b 7 represents, with respect to the chosen scale of the times, the moments of sampling of the video signal by the input circuit analog- digital converter, these being numbered, starting from zero, in their chronological sequence from the origin of the times OH, which sequence is that of writing of the 12 digital samples in the store 301. The cross line c 7 represents the reading sequence of the samples in the store 301, at an interval of one video line, i.e. 64 gs later. This sequence is the same as the writing sequence up to the start of the transitory period which has been fixed at the 95th sample, i.e. sample 94, 5.30 gs from the origin of the times OH, i.e. 0.28 gs from the nominal start of a SECAM or PAL colour burst signal- The 95th to 99th samples written in the store 301, i.e. the samples 94 to 98, which have theoretically appeared before the start of the colour burstsignalare set aside and replaced during reading by two successive rereadings of the samples 91, 92 and 93. Since the sequence of writing of the samples in the store 301 also corresponds to the addresses delivered by the counter 305 because the latter is re-set on zero at the origin of the times OH this is done, when reading the store 301, by twice causing a jump from the address 93 to the address 91. In view of the delay of three samples between an address loading command and the actual loading of the latter in the address counter 305, delay imposed by the two successive stages of synchronisation registers encountered by the signals in the prepositioning loop, this is achieved, as indicated in the block diagrams of figures 4, 5 and 6, by de-coding of the address 91 followed by a command to load the address 91 which will be executed only at the moment when the counter 305 reaches the address 94, this command being repeated once by a loop from which an exit is made as soon as the jump counting variable x reaches 2.

The reconstruction of the part of the transitory period of time following the nominal start of the burst necessitates the determining of a sequence of successive samples of the burst sine wave signal which can be looped onto itself without causing any phase jump to appear when passing from the last sample to the first. The length of this sequence of samples depend of course on the frequency of this sine wave signal.

The SECAM system gives colour burst signal at the FOB frequency of 4.25 MHz for the blue colour component and FOR colour burst signal of 4.406 MHz for the red colour component whereas the PAL system gives colour burst signal at the single frequency cf 4.433 MHz.

9 13 A period from one signal at the FOB frequency of 4.25 MHz sampled at the frequency of 17.734 MHz gives a sequence of four samples the last of which is taken at 9.7 ns from the end of the period of the signal at the FOB frequency which represents in the case of looping back of the sequence a phase jump of 150. To reduce this phase jump, a multiple of this sequence plus one sample is adopted, the multiple being chosen so that the residual time is very close to the sampling period and justifies the additional sample. Since the sampling period represents about 5.8 times the residual time interval of 9.7 ns, the sixth multiple is chosen, which leads to a sequence of 25 samples lasting 1.409 gs whereas six period of the signal at the FOB frequency last 1.411 ps. The residual time difference of 2 ns represents in the case of looping back a practically negligible phase jump of 3'- A period of one signal at the FOR frequency of 4.406 MHz sampled at the frequency of 17.734 MHz gives a sequence of four samples with a residual time difference of 1.4 ns which represents a negligible phase jump of 2. 21 in case of looping back. It is therefore not necessary to look for a longer sequence.

A period of one signal at the frequency of 4.433 MHz sampled at the frequency of 17.734 MHz gives a sequence of four samples with a residual time difference of 1.1 ns which represents a negligible phase jump of 1. 70 in case of looping back. As in the preceding case, it is not necessary to look for a longer sequence.

The length of the sampling sequences at the basis of the reconstruction of the burst start being f ixed, the number of samples from these sequences actually used must be determined. The number depends on the length of theburst start included in the transitory period and also in the case of the SECAM system the burst of which extend uninterrupted as far as the useful part of the video line, on the delay 0, R, 2R to be introduced for encoding or decoding. Different cases arise therefore depending on whether the system is PAL or SECAM and whether, in the SECAM system the video line is a line of blue or red colour component the useful part of which has to be delayed by 0, R or 2R.

14 Figure 8 shows the process of reconstructing the burst start in the case of a video line of blue colour component of a SECAM system in which it is not necessary to delay the useful video part.

Curve a 8 represents the start of a video line of the SECAM system with the line synchronisation pulse the leading front OH of which serves as origine of the times, the recalibrated burst start being placed 5.58 Ls further on. The cross line b 8 represents, with respect to the chosen scale of the times, the moments of sampling of the colour burst signal start numbered in their chronological sequence from the origin of the times OH, which sequence is that of addressing at the writing of thestore 301. The cross line c 8 represents the sequence of reading of the samples in the store 301 at an interval of one video line.

In practice, it is sufficient to reconstitute the.burst start over a duration of the order of 1.4 gs, i.e. the segment comprised between samples 99 and 123 inclusive. Here, a slightly more extended reconstitution is performed affecting the segment comprised between samples 99 and 130 inclusive.

The following 25 samples, i.e. samples 131 to 155, are chosen as a reconstruction sequence. The segment to be reconstituted extending over 32 samples, a single reading of the samples of the reconstruction sequence is not sufficient. It is supplemented by the reading of samples 124 to 130 which although situated in the segment to be reconstituted, are more than a microsecond away from the nominal start of the salvo. The reconstitution of the segment comprised between samples 99 and 130 amounts then to replacing of the samples 99 to 130 with the samples 124 to 155. This is achieved, as shown in the block diagram of figure 4, by two address jumps when reading the store 301: a first forward address jump making it possible to pass from the chronological position 98 with respect to the line start OH to that of 124 followed by a second backward address jump making it possible to pass from the address 155 to the address 131. As shown in the block diagram of figure 4 the first jump is commanded as soon as the address counter reaches 93 when leaving the X first loop, which corresponds chronologically with the position 96 with respect to the line start OH, and three samples later, at the chronological position 99, brings about the loading of the counter 305 at the address 124. The second jump is commanded as soon as the counter reaches 153 for a first time, which corresponds chronologically to the position 128 with respect to the line start OH, and is executed three samples later for the chronological position 131 where the address counter is repositioned at 131.

Figure 9 shows the process of reconstructing the burst start in the case of a video line of blue colour component of a SECAM system when it is necessary to delay the useful video Dart by a delay R equal to 902 ns corresponding to ',a segment of sixteen samples.

Curve a 9 represents the start of a video line of the SECAM system with the leading front OH of the line synchronisation pulse serving as origin of the times and the start of the colour burst signal appearing 5.58 ps later. The cross line b 9 represents, with respect to the chosen scale of the times, the moments of sampling of the start of the colour burst signal numbered in their chronological sequence from the origin of the times OH, which sequence is that of addressing when writing the store 301. The cross line C9 represents the sequence of reading of the samples in the store 301, at an interval of one video line.

As previously, the reconstitution affects the segment comprised between the samples 99 and 130 inclusive and is performed by means of the following 25 samples, i.e. the samples 131 to 155 chosen as reconstruction sequence.

The segment to be reconstituted starts at the position of sample 99 and encloses the 32 positions of the suppressed samples plus 16 positions of samples representing the delay R which must affect the useful part of the video signal. Its rear part, from samples 122 to 146, is prepared by a complete re-reading of the 25 samples of the reconstruction sequence whereas its front part, from samples 99 to 1.21, is prepared by a rereading of the last 23 samples of the reconstruction sequence.

16 This is achieved, as shown in the block diagram of figure 4 by three successive address jumps when reading from store 301.

The first address jump is a forward jump making it possible to pass from the chronological position 98 with respect to the linb start OH to that of 133.

It is triggered as soon as the address counter 305 reaches 93 when leaving the first loop which corresponds chronologically with the position 96 with respect to the line start OH and becomes effective three samples later, the moment when the address counter 305 is prepositioned at 133.

The second address jump is a backward jump making it possible to pass from the address 155 to the address 131. It is triggered as soon as the address counter 305 reaches 153 and will become effective when the latter reaches 156.

The third address jump is a repetition of the second performed by a second loop from which an exit is made as soon as the jump counting variable x reaches 2.

Figure 10 shows the process of reconstructing the burst start in the case of a video line of blue colour component of a SECAM system when it is necessary to delay the useful video part by a delay of 2R, twice the preceding delay, equal to 1804 ns and corresponding to a segment of 32 samples.

Curve a 10 represents the start of a video line of the SECAM system with the leading front OH of the line synchronisation pulse serving as origin of the times.

The cross line b 10 represents, with respect to the chosen scale of the times, the moments of sampling of the colour burst start numbered in their chronological sequence from the origin of the times OH, which sequence is that of addressing at the writing of the store 301. The cross line c 10 represents the sequence of reading in the store 301, at an interval of one video line, of the burst start samples.

The reconstitution still affects the segment comprised between the samples 99 and 130 inclusive and is done by means of the following 25 samples, i.e. the samples 131 to 155, chosen as 1 1 reconstruction sequence.

The segment to be reconstituted starts at the position of the sample 99 and encloses the 32 positions of the samples suppressed plus 32 positions of samples representing the delay 2R which must affect the useful part of the video signal. Its rear part, from samples 138 to 162 inclusive, and its middle part, from samples 113 to 137 inclusive, are each prepared by a complete re-reading of the 25 samples of the reconstruction sequence whereas its front part, from samples 99 to 112 inclusive is prepared by a re-reading of the last 14 samples of the reconstruction sequence. This is achieved, as shown in the block diagram of figure 4, by four successive address jumps in reading of the store 301.

The first address jump is a forward jump making it possible to pass from the chronological position 98 with respect to the start of line OH to that of 142. It is always triggered as soon as the address counter 305 reaches 93 when leaving the first loop, which corresponds chronologically with the position 96 with respect to the start of line OH and becomes effective three samples later, at which moment the address counter 305 is prepositioned at 142.

The second address jump is a backward jump making it possible to pass from the address 155 to the address 131. It is triggered as soon as the address counter 305 reaches 153 and will become effective when the latter reaches 156.

The third and fourth address jumps are repetitions of the second performed by a second loop which is left once the jump counting variable x reaches 3.

Figure 11 shows the process of reconstructing the burst starts of the PAL system jideo lines or of a SECAM system video line of red colour component when it is not necessary to delay the useful part.

Curve a 11 represents the start of a SECAM or PAL video systew, 1 ine, the end of the burst appearing in the case of the PAL systew. 7.85 ts after the rising front OH of the line synchronisation pulse, beyond the section shown. The cross line b 11 represents, with respect to the chosen scale of the times, the moments of sampling of the 18 start of the colour burst sl numbered in their chronological sequence from the origin of the times OH, which sequence is that of addressing at the writing of the store 301. The cross line c 11 represents the sequence of reading in the store 301, at an interval of one video line, of the burst start samples.

With the frequency of 4.43 MHz of the PAL system bursts as with the frequency of 4.406 MHz of the red component bursts of the SECAM system the reconstruction sequence may, as we saw previously, be limited to four consecutive samples.

The choice is made to reconstitute theburst start segment extending between the samples 99 and 113 inclusive by using as reconstruction sequence the four following samples, i.e. samples 114, 115, 116 and 117. The reconstruction of this segment is achieved, as shown, by preparing its rear part extending from samples 102 to 113 inclusive by a triple rereading of the reconstruction sequence and by preparing its front part formed of the samples 99, 100 and 101 by a re-reading of the last three samples of the reconstruction sequence. This is achieved, as shown in the block diagrams of figures 5 and 6 by five successive address jumps in reading of the store 301.

The first address jump is a forward jump making it possible to pass from the chronological position 98 with respect to the start of line OH to that of 115. It is triggered again as soon as the address counter 305 reaches 93 when leaving the first loop which corresponds chronologically to the position 96 with respect to the start of line OH and becomes effective three samples later, at which moment the address counter 305 is prepositioned at 115.

The second address jump is a backward jump making it possible to pass from the address 117 to the address 114. It is triggered as soon as the address counter reaches 115 and will become effective three samples later when the latter reaches 118.

The third, fourth and fifth address jumps are repetitions of the second performed by a second loop from which an exit is made as soon as the jump counting variable x reaches 4. In the cases where it is necessary to delay the useful part of a video line of red colour A t 19 components of the SECAM system by a delay R of 16 samples or 2R of 32 samples, it is sufficient to modify the number of repetitions of the second loop and to make it pass, as shown in the block diagram of figure 5 either from 3 to 7 in the case of a delay R, the exit from the loop being made when the jump counting variable x reaches 8, or from 3 to 11 in the case of a delay 2R, the exit from the loop being made when the jump counting variable reaches 12. In fact each additional repetition of the second loop enables the colour burst signal to be extended by a duration of four samples without causing any disturbing phase jump.

Figure 12 shows the process of introducing a delay R of 16 samples into the useful part of a PAL system video line. Curve a 12 represents the start of a video line with the leading front OH of the line synchronisation pulse serving as origin of the times, the colour burst signal appearing 5.58 jus later over a duration of 2.25 ps and the start of the useful video 10.5 gs after. The cross line b 12 represents, with respect to the chosen scale of the times, the moments of-samplinq of the black suppression level after the end of the colour burst si&ml numbered in their chronological sequence from the origin of the times OH, which sequence is that of addressing at the writing of the store 301. The cross line c 12 shows the sequence of reading from the memory 301, at an interval of one video line, of the samples of the black suppression level at the end of a colour burst signal The introduction of the delay R between the useful video and the line synchronisation pulse is performed simply by a repetition of a 16 sample seqment situated on the black suppression level at the end of a colour burst signal between samples 159 and 174 inclusive. This is achieved, as shown in the block diagram of figure 6, by means of an address jump in reading of the store 301 making it possible to pass from the address 174 to the address 159. It is triggered when the address counter 305 reaches 172 and will become effective when the latter reaches 175.

The introduction of a delay 2R, twice the preceding delay, is performed, as shown in the block diagram of figure 6, by repeating the address jump provided for the introduction of the delay R by means of a third loop from which an exit is made as soon as the jump counting variable x reaches 2.

When a delay R or 2R has been introduced between the useful part and the line synchronisation pulse of a PAL or SECAM system video signal, the end of the useful part of a line encroaches on the clearance interval preceding the following line synchronisation pulse. To eliminate this defect, a recutting of the end of each line is performed consisting of eliminating the 16 or 32 samples bf the useful part overlapping the clearance interval. This is achieved. as shown in the block diagrams of figures 4, 5 and 6 by an address jump in reading of the store 301 making it possible to pass from the address 1091 if a delay R has been introduced or from the address 1075 if a delay 2R has been introduced to the address 1108. This jump is triggered as soon as the address counter 305 reaches 1089 or 1073 as the case may be so as to take into account the delays introduced by the two stages of synchronisation registers encountered by the signals in the prepositioning loop of the address counter 305 and becomes effective three samples later when this counter is prepositioned at 1108.

The process for recalibration of the start of a colour burst signal which has just been described is of course applicable to noncoded or differently coded colour television video signals, for example by circular offsets of the useful video part, these offsets being obtained in the course of reading of the useful video part in the memory planes by causing the address counter to make the adequate jumps.

21

Claims (1)

1/ Process for recalibration of the colour burst signals of a colour television video signal in which the video signal is sampled and its samples are stored in two memory planes of random addressing type operating in parallel one with the writing and the other with the reading, their roles being exchanged at the end of a time interval corresponding to the duration of one line, characterised in that the samples are read in each memory plane, following the chronological sequence of their writing with the exception of those appearing over a transitory period of time straddling the nominal start of the colour burst signal, which are replaced at the reading, for the part preceding the nominal start of the colour burst signal., by imediately precedinq samples and, for the part following the nominal start of the- colour burst signal by one or several groups of samples following one another in the chronological sequence of writing, which groups are selected from the colour burst signal, subsequently to the transitory period of time, on the end of a sequence of samples covering a whole number of periods of the sine wave signal of the burst and linking up in phase with the start of the sequence of samples of the burst following on from the transitory period of time.

2/ Process according to claim 1 used in the context of the encoding or of the decoding of a composite video signal the colour burst signal of which is lengthened uninterruptedly as far as the useful part and which is encrypted by delays of fixed values displacing, to suit a particular code, the useful part of each video line with respect to the line synchronisation pulse, characterised in that it further consists of lengthening during its reconstitution the said transitory period of time, by a number of samples corresponding to the duration of the delay to be added for the encoding or decoding.

3/ Device for implementation of the process according to claim 1 comprising an input circuit performing a sampling of the video signal to be recalibrated and the preparation of synchronisation signals linked to the video signal to be recalibrated, two memory planes of random addressing type memorising the samples 1 22 delivered by the input circuit and operating one at the writing and the other at the reading, their roles being exchanged at the end of an interval of time corresponding to the duration of one video line, and an output circuit putting back into analog form the samples coming from the reading of the memory planes, characterised in that each memory plane comprises a random access memory with a prepositionable address counter stepped by the rate of sampling of the input circuit and fitted with a prepositioning loop enabling it to perform,-at the reading of the random access memory address jumps with respect to the writing sequence in order to reconfigure the signal in the transitory period of time straddling the nominal start of the colour burst signal by re-readings of samples preceding and following this transitory period of time.

4/ Device according to claim 3, characteristed in that the said loop for prepositioning of the address counter comprises three read-only memories the output signals from which are individually re-synchronised by banks of registers, the first read-only memory performing the decoding of the starts of the adcA res5 jumps and being addressed by the address counter, the second read-only memory performing the decoding of the conditions for authorisations of the address jumps, being addressed by the data read from the first read-only memory and by synchronisation signals coming from the input circuit, and.delivering a command signal for prepositioning of the address counter, the third read-only memory performing the decoding of the destination addresses of the jumps, being addressed in the same manner as the second read-only memory and delivering propositioning addresses to the address counter.

5/ A process substantially as hereinbefore described with reference to Figurs I to 8, Figures I to 7 and 9, Figures I to 7 and 10, Figures 1 to 7 and 11, or Figures 1 to 7 and 13 of the accompanying drawings.

6/ A device substantially as hereinbefore described with reference to Figures 1 to 8, Figures 1 to 7 and 9, Figures 1 to 7 and 10, Figures I to 7 and 11, or Figures I to 7 and 12 of the accompanying drawings.

Published 1990 at The patent Office. State House. 65 71 High Holborn London WCIR4TP.Purtber copies maybe obtainedfrom The patent office Sales Branch. St ma7. Cra:,% Orpington. Kent BR5 3RD Printed by multiplex techniques ltd. St Mary Cray, Kent. Con 1'87