GB2262859A - Letterbox television signal with chrominance helper signal - Google Patents

Abstract

In a letterbox-type video signal, in which a main signal is sent on the lines forming the central three-quarters of the picture, additional helper information is transmitted in the upper and lower borders of the picture. This helper information includes chrominance components that provide enhanced horizontal detail, for vertical frequency components up to a small proportion of the vertical detail carried in the main signal. At the encoder the chrominance components are vertically sub-sampled (16) to produce one line from every six input lines, and are horizontally stretched (18) and split (20) so that each input line extends onto two lines. At the decoder the operations are reversed by retiming onto one line and horizontal compression (42) and vertical super-sampling or interpolation (44). The extra chrominance detail can then be added (46) to the main signal chrominance components to enhance the horizontal detail.

Description

IMPROVEMENTS RELATING TO TELEVISION BACKGROUND OF THE INVENTION This invention relates to color television systems and in particular to improved color video encoders and decoders for use in color television systems.

More particularly the invention is concerned with composite color television systems in which the main signal is conveyed using fewer than the total number of available active television lines.

An example of such a system is one which uses a 16:9 aspect ratio.

A 16:9 aspect ratio signal can be displayed on a special 16:9 wide aspect ratio receiver, or can be displayed on an existing conventional 4:3 aspect ratio receiver with what is known as a "letterbox display, see, for example, IBC 1992 Amsterdam, IEE Conference Publication No. 358, July 1992, pages 203 to 207, and other papers in the same publication.

In a letterbox display, illustrated in Figure 1, the wide aspect ratio image is displayed with full horizontal width in the vertically central part of the screen, and the top and bottom of the picture are blanked off to black. This subjective effect is sometimes seen already when wide-screen cinematographic films are transmitted by broadcast television. In a 625/50/2:1 standard PAL transmission system, it has already been proposed to transmit the signal on the middle 432 active lines of each picture, which are then displayed on a 4:3 aspect ratio receiver in letterbox format.

This leaves 144 spare lines, namely 72 at each of the top and bottom of the picture. These portions are referred to as the borders.

The letterbox system achieves compatible broadcasting of wide screen pictures by having black bars above or below the wide screen picture. However, the enhanced wide screen picture now occupies less of the transmission channel capacity than it did in the original picture, and it is necessary to transmit additional helper information during the black bar or border periods, so that the finally decoded enhanced picture on a 16:9 receiver has at least as good a resolution as a 4:3 picture would have done. Several methods of using the black bar period have been proposed to provide a helper for the luminance component,including that described in our International Patent Application No. PCT/GB92/01988 (published under the number W093/ on 1993) claiming priority from United Kingdom Patent Application No. 9123004.5.

Within enhanced composite terrestrial transmission systems, much attention has been paid to achieving a good horizontal and vertical luminance response. However, we have appreciated that, particularly with PAL System B and G transmissions, with NTSC and with SECAM, the limiting factor for any overall improvement in picture quality in a letterbox system is the overall horizontal resolution of the chrominance component.

This may be seen from Figure 2, which is a representation of the frequencies transmitted in the letterbox portion of the picture by the chrominance component of the signal. The plot is of vertical frequency measured in cycles per active picture height c/aph, against horizontal frequency measured in MHz (proportional to cycles per active picture width). The scales are chosen relative to each other so as to be substantially isotropic, that is to say that subjectively the same image quality is represented by equal distances along the horizontal and vertical axes, and are correctly scaled for a 16:9 aspect ratio.

Figure 2 thus shows the two-dimensional chrominance bandwidth for both PAL systems B and G which have a 5.0 Hz overall video bandwidth, on the one hand, and PAL system I which has a 5.5 MHz overall video bandwidth, on the other. In each case the maximum vertical frequency that can be displayed on a single field is 54 cycles per active picture height (equivalent to 108 cycles per picture height for a picture of two interlaced fields). Systems B and G provide a horizontal bandwidth of about 30 cycles per picture width and system I provides a horizontal bandwidth of about 60 cycles per picture width. For PAL systems B and G, the maximum horizontal frequency is just above 0.5 MHz, and for PAL system I, the maximum horizontal frequency is just above 1.0 MHz. These boundaries are not well defined and this is illustrated by the hatched areas marked on the figure.

The extra 144 lines available in the borders could be used to contain information which would assist the 432 lines of the main picture in the letterbox. There are one-third as many lines in the borders as there are in the letterbox. Thus the vertical definition could be increased for example by associating each border line with three of the letterbox lines. In this way the vertical definition could be increased to 72 c/aph, as indicated by the dashed lines on the figure.

It will be seen from Figure 2, however, that this is a rather pointless complication, because the vertical frequency resolution, as plotted on these isotropic co-ordinates, is already so much better than the horizontal resolution that further improvement would be virtually unnoticeable. This is particularly so given the sensitivity of the eye to vertical lines in the picture, giving rise to horizontal detail. This is particularly the case with PAL systems B and G, but is still true for PAL system I. Thus the capacity of the chrominance part of the composite signal spectrum in the borders has remained unused in proposals hitherto considered.

More generally the invention is applicable not only to letterbox type PAL or similar signals, but to any system in which the video signal is formed by fields of scanned lines and which includes separate luminanace and chrominance components, with the active picture occupying only some of the scanned lines of each field forming a main group of lines, and in which an auxiliary group of lines which does not carry the active picture is used to carry helper information.

SUMMARY OF THE INVENTION The invention in its various aspects is defined in the independent claims appended to this description, to which reference should now be made. Advantageous features of the invention are set forth in the appendant claims.

Various preferred embodiments of the invention are described in more detail below in the context of a PAL system letterbox signal based on a 625/50/2:1 conventional PAL signal with 4:3 aspect ratio.

The lines forming the central three-quarters of the picture can be used to transmit a wide aspect ratio picture of 16:9 aspect ratio.

Additional helper information can then be transmitted on the lines forming the borders at the top and bottom of the picture.

As will be described in relation of the embodiments illustrated, at the encoder the chrominance components are vertically sub-sampled to produce one line from every six input lines. The signal is also horizontally stretched so that each input line extends onto two lines. In this way information derived from the main letterbox signal can fill the chrominance capacity of the borders and be modulated onto the color subcarrier for transmission through a normal PAL transmission system.

At the decoder the operations effected to generate the chrominance signal in the borders are reversed by demodulation, retiming each two transmitted lines onto a single line, and vertical super-sampling or interpolation to spread each line of information onto six lines of the main signal. The extra chrominance detail can then be added to the main signal chrominance components to enhance the horizontal detail.

This processing enables the transmission in the chrominance helper information of components that provide substantial enhanced horizontal detail, for those vertical components up to a small proportion, in this case one-sixth, of the vertical detail carried in the main signal. This produces a considerable increase in subjective image quality, for relatively little extra complexity.

The derivation of the chrominance border signal as described gives rise to some duplication. Components of low horizontal frequency and low vertical frequency would be carried in both the main chrominance signal and in the chrominance helper signal in the borders. This is unnecessary, and to reduce the signal level of the chrominance helper signal it is preferred to pass the signals to be transmitted in the chrominance helper signal through a horizontal high-pass filter at the encoder. At the decoder, any residual duplication is removed by subjecting the received chrominance components in the main signal to the same vertical sub-sampling and horizontal high-pass filtering as is used at the encoder, and subtracting the resultant from the chrominance border signals after they have been horizontally compressed but before the vertical super-sampling.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described in more detail, by way of example, with reference to the accompanying drawings, in which: Figure 1 (referred to above) illustrates a letterbox display on a conventional 4:3 aspect ratio screen; Figure 2 (referred to above) is a plot of vertical frequency against horizontal frequency showing the bandpass characteristics of the chrominance component in the main part of a letterbox signal; Figure 3 is a block schematic diagram of a first encoder embodying the invention; Figure 4 is a block schematic diagram of a first decoder embodying the invention; Figure 5 is a plot similar to Figure 2 showing the bandpass characteristics of the coder of Figure 3 with a PAL system B or G signal; Figure 6 is a plot similar to Figure 5 for a PAL system I signal;; Figure 7 is a diagrammatic illustration of how the chrominance helper lines are derived from the chrominance letterbox lines; Figure 8 is a plot similar to Figure 5 showing how a preferred receiver/decoder uses the signal transmitted by the coder of Figure 3, for a PAL system B or G signal; Figure 9 is a plot similar to Figure 6 showing how the preferred receiver/decoder uses the signal transmitted by the coder of Figure 3, for a PAL system I signal; Figure 10 is a block circuit diagram of a second encoder embodying the invention, showing the chrominance border signal path at (a) and the chrominance letterbox signal path at (b); Figure 11 is a block circuit diagram of a second decoder embodying the invention;; Figures 12 to 16 are graphs showing the characteristics of the horizontal filters HFl, HF2 and HF3 and of the vertical filters VF1 and VF2, respectively, used in the second embodiment of Figures 10 and 11; Figure 17 is a block circuit diagram of the encoder of a third embodiment of the invention; and Figure 18 is a block circuit diagram of the decoder of the third embodiment of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As discussed above, the bandwidth that would be occupied by the chrominance component in the border regions has not hitherto been used. We have appreciated that this bandwidth can be used to achieve a significant improvement in the horizontal chrominance resolution. The additional chrominance helper information can be broadcast compatibly with existing receivers, but is only used to enhance the display of specially equipped receivers.

Systems to date have not used this capacity since it is only normally of narrow horizontal resolution capacity and of limited vertical resolution. No method has been devised of ensuring that enhancements to horizontal resolution can still be achieved. Such enhancement is provided by the present system yet is nevertheless independent of and not affected by any incidental additional curtailment or high frequency loss or dispersion at the upper boundary frequency of the channel bandwidth, which might occur when the signal is transmitted and received. Such impairments are entirely outside the control of the system designer or receiver manufacturer. Further complications have arisen in using the chrominance band during the black bars (the borders) and maintaining it separate from any modulated luminance signal which might also be transmitted during the same black bars.

The basic system operation is illustrated in Figures 3 and 4. Briefly, in parallel with the normal color chrominance circuits, color signals are stretched horizontally and filtered and vertically sub-sampled before being supplied to a buffer store so that the additional chrominance color information is only conveyed during the black bar intervals. Upon reception, color signals in the black bars are separated from the luminance signals, time compressed and vertically interpolated and re-timed to become coincident with the main signal. A filter is preferably used to cross-fade between the main signal chrominance and the chrominance transmitted in the black bars on the basis of vertical, vertical/temporal or vertical/horizontal/temporal frequencies.

In this way enhanced chrominance resolution can be achieved in a composite television system, with the chrominance bandwidth typically being doubled, by applying time expansion techniques, and modulating the composite color channel with this information so that it is transmitted during the black bars.

Referring in more detail to Figures 3 and 4, Figure 3 shows the additional circuitry at the encoder (transmitter) in a letterbox PAL color television system and Figure 4 shows the additional circuitry at the decoder (receiver). In the encoder 10 it is assumed that during the border portions of the signal a luminance helper signal is generated, for example using the method described in our above-mentioned application No. PCT/GB92/01988. Such a luminance helper signal is received at an input 12. At an input 14 is received a chrominance signal. This can take any appropriate form but is here assumed to comprise U and V on alternate lines.

It could however be U+V on one line and U-V on the next, or could comprise I and Q alternately for example. This signal is then subjected to three processing operations. These can take place in any order.

Considered conceptually, the first of these processing operations is provided by a filter and vertical subsampler 16.

This comprises a low-pass filter with a cut-off of one sixth of the vertical frequency passband of the main chrominance signal in the letterbox. Thus with a vertical frequency of 54 c/aph (cyles per active picture height) as shown in Figure 2, the filter has a cut-off frequency of 9 c/aph. The sub-sampling reduces the number of lines per picture from 432 to 72, i.e. by a factor of six. That is, each six input lines are used to make one output line. This again is a vertical filter function using a defined vertical aperture and thus both the filtering and sub-sampling functions can be achieved by the same circuit 16.

Having now reduced the number of lines by a factor of six, the lines are horizontally stretched by a factor of two in a circuit 18. That is to say if it is assumed that the signal is a digital or sampled analogue signal, for convenience in explanation, the number of samples on each line is doubled. Of course, this does not increase the amount of information made available but nevertheless it enables more information to be conveyed subsequently by the transmission channel. This arises because the third operation, conducted in a retiming store 20, is to rearrange each line of the signal onto two lines, now making 144 lines in all.

The three processing operations conducted in circuits 16, 18 and 20 can in principle take place in any order and it is instrumentally convenient to place them in the order indicated, namely with the horizontal stretching first in circuit 18, then the vertical filtering 16, and finally the retiming 20.

The resultant signal is applied during the borders to a circuit 22 which combines the luminance helper signal at input 12 with the output of the retiming store 20. A selector switch 24 then operates so that during the letterbox part of the signal it accepts for transmission the normal main composite signal received at a terminal 26, but during the black bars in the borders it takes instead the helper output of the circuit 22. The switch 24 provides an output 28 which can be transmitted. Alternatively, the signal may be recorded on a recorder and the expression 11transmission and similar terms is intended to encompass such recording as is wellknown to those skilled in the art.

The effect of the processing described on the signal is illustrated in Figure 5 for a PAL system B or G signal. This is similar to Figure 2. The main signal in the letterbox has a chrominance bandwidth as shown by the rectangle M, extending to 54 c/aph vertically, and roundabout 0.5 MHz horizontally, this boundary being somewhat fuzzy due to the characteristics of the transmission channel. During the borders a signal is derived which has only one sixth of the vertical bandwidth. However, due to the horizontal expansion, it is able to provide twice the effective horizontal bandwidth. This is shown by the rectangle B. It will be apparent that the combination of rectangles M and B together provide a much more satisfactory overall bandpass characteristic, as the resultant bandwidth is much more nearly isotropic.It is true that diagonal frequencies are not transmitted as well as either vertical or horizontal frequencies, but these arise less in practice and the eye is not so sensitive to them.

Figure 6 is a similar diagram for PAL system I. Here again the overall bandwidth transmitted is much improved by the inclusion in rectangle B of the higher horizontal frequencies at low vertical frequencies. The system enables horizontal chrominance signal components of up to about 2 MHz to be transmitted even though the channel bandwidth cuts off at around 1 MHz.

The manner in which the circuits 16, 18, 20 operate can be seen from Figure 7. This figure shows seven lines A to G of the chrominance signal at input 12 in Figure 3 and illustrates the samples on each line as dots. The samples along the line are numbered: 1, 2, 3, 4, 5, 6 ..... 462, 463, 464, 465, 466, 467, 468...

Also shown are two output lines M and N such as would be transmitted as the chrominance signal component during the border regions. The operations have the following effects. The subsampling first reduces the corresponding samples on six successive input lines to a single output line. Thus samples Al.. Fl are combined to form sample M1. Sample M2 is an interpolated sample, achieved by the horizontal stretching or expansion. Output sample M3 is derived from input samples A2...F2. This continues with the even samples being additional or interpolated samples, and the odd samples on line M being derived directly from lines A to F.

Half-way along the input lines A to F the end of line M will have been reached. Thus samples A462...F462 are combined to form the first sample of the next output line, i.e. sample N1. Sample N2 is again an interpolated sample, sample N3 is derived from A463...F463, and so on. This representation in Figure 7 is very diagrammatic and the actual functions may and usually will be more complicated than this but it serves to illustrate the operations that are going on in principle. Furthermore, the representation of Figure 7 is given in terms of sampled signals but the principles are applicable to all forms of signals, including analogue signals as well as fully-digital signals. The notion of samples is used in Figure 7 simply to illustrate the processes that are taking place.

It will be seen from the foregoing that the important operations involved are as follows. First, there are fewer border or auxiliary lines than there are lines in the letterbox or main signal, by a factor n which in this case is equal to 3. These border lines do not carry the active picture but instead carry chrominance helper information. The lines of chrominance helper information are derived from the chrominance component in the main part of the signal using two operations. One of these is to expand horizontally each line so that chrominance information from a line of the letterbox is extended over more than one line of the border.

This extension may be by a factor r, where r equals 2 in the present case. The other operation is to reduce the information vertically.

This is achieved by sub-sampling the chrominance component in the letterbox to produce a smaller number of output chrominance lines in the borders, the subsampling being by a factor which equals n times r, which in this case is 3 times 2 which equals 6. These two operations may take place in either order and indeed in theory could take place simultaneously by means of a two-dimensional filter.

The output lines thus obtained are then timed to occur in sequence in the border regions by means of a buffer store.

The extension of one line of information over 2, or more generally r, lines, effectively reduces the horizontal frequency of the chrominance components by a corresponding factor for the purposes of transmission as the chrominance helper signal. Thus in the embodiment of Figures 3 and 4, assuming PAL systems B and G, a horizontal frequency component of say 0.9 MHz is halved for the purposes of transmission to 0.45 MHz. This frequency is now within the bandwidth that can be carried by the conventional PAL distribution system. At the receiver a frequency doubling takes place by joining two received lines together into a single line.

The information that took two line periods to transmit is now compressed into a single line period again.

The factors n and r can take other values. For example, r could be 3. This would lead to three times greater horizontal definition over one ninth of the vertical frequency band. In principle, the values n and r could be non-integral, though the processing is easier if they are integral, or at least are a ratio of small integers in particular with a denominator not larger than about 10.

Using the receiver/decoders to be described, a great improvement in quality can be obtained, because the system adds the important frequencies that are noticed by the viewer. The sublective increase in chrominance resolution quality with a PAL system I signal can be of the order of 60 to 70%. This substantial improvement is obtained with quite a small increase in system complexity. Although described in terms of the PAL system, the principles are applicable to all other conventional television systems, including the NTSC and SECAM systems. In particular the improvement can be applied to all letterbox-type proposals.

It may be noted here that the numerical values for the vertical detail given in cycles per active picture height are those appropriate to a normal interlaced video signal. Some equipment can alternatively operate in a film mode, in which a non-interlaced signal is produced. This arises because the standard 50 Hz field rate is almost exactly double the standard film frame rate of 24 Hz.

In generating video from film the film is run slightly fast at 25 Hz. The third embodiment shown in Figures 17 and 18 gives an example of film mode operation, where the vertical detail in the letterbox is 108 c/aph and the added detail in the borders is 18 c/aph. Non-interlaced signals are used which are turned into interlaced signals for transmission. The principles so far as the present invention are entirely the same as for normal signal sources, examples of which are seen in the first and second embodiments of Figures 3 and 4 and Figures 10 and 11 respectively.

Figure 4 illustrates corresponding enhanced receiver/decoder with 16:9 aspect ratio. The decoder 30 shown has an input 32 which receives a signal corresponding to that provided at 28 in Figure 3.

This may come from a broadcast television channel or a video recorder for example. A conventional main luminance path 34 carries the luminance component during the main or letterbox part of the signal A conventional main chrominance path 36 carries the chrominance component during the main or letterbox part of the signal. A separator 38 separates-the luminance and chrominance components during the border regions. The luminance helper signal is processed as appropriate, e.g. as in our above-mentioned application and combined with the output of path 34 in an adder 40 to provide the luminance output 41. The details of the luminance helper channel are not shown.

The chrominance component supplied by separator 38 during the black bars is applied to a retiming and horizontal compression circuit 42, which reverses the effects of the circuits 18,20 in Figure 3. The output of circuit 42 is applied to circuit 44 which reverses the sub-sampling effect of circuit 16 in Figure 3, by providing a vertical interpolation, that is to spread the effect of each sample across six vertically aligned samples of the main signal. The chrominance helper information thus obtained is then combined with the output of the main chrominance path 36 in a cross-fade filter 46 to provide an enhanced color output 48. The reason that a cross-fade filter is used will become apparent subsequently.

It is seen from Figures 5 and 6 that part of the chrominance information is duplicated in the main and border signals. The duplicated part comprises those components of vertical frequency below 9 c/aph, and horizontal frequency, in the case of PAL systems B and G (Figure 5), below 0.5 MHz. The receiver could then be designed to take these components either from the main signal or from the border signal. What it in fact does is illustrated by Figures 8 and 9.

In practice it is better, as a general principle, if the very low frequency information comes from the main signal. The very low frequency information is comprised in the part of the plot which is nearest to the origin. It is better for this to come from the main signal because there are problems at low frequencies due to the need to re-join the two helper lines, which it will be recalled relate to the two halves of a single horizontal scan. Referring to Figure 8, therefore, it is seen that the frequencies below about 0.4 MHz all come from the main signal. The border signal is used to boost image quality by adding in components with vertical frequency below 9 c/aph and horizontal frequency between 0.5 MHz and 1 MHz.

There is a transition range, shown cross-hatched on Figure 8, where the receiver changes over from using the main signal to using the border signal. This transition can not in practice take place abruptly, so it is best designed as a cross-fade. This is the reason why the combining circuit 46 in Figure 4 is implemented as a cross-fader. The transition or cross-fade range is located below the shaded region in Figure 8 where the horizontal bandwidth starts to tail off, so that it is not affected by the channel capacity of the particular transmission channel being used.

With a PAL system I receiver it is likewise possible to use up to about 0.9 MHz from the main channel, even at vertical frequencies from 0 to 9 c/aph, and to use the border signals to add just the signals with horizontal frequencies of 1.0 to 2.0 MHz and vertical frequencies below 9 c/aph. However, we prefer not to do that, but rather to use the same cross-fade range of 0.4 to 0.5 MHz, as shown in Figure 9. The reason for this is simply that a single piece of equipment can then be designed to accept both PAL systems B/G and system I signals. The only difference is in the overall chrominance bandwidth, which is normally defined by the channel characteristics in any event.

If the receivers are to operate in accordance with Figures 8 and 9, then there is no need for the border signal to include those frequencies which the receivers all derive from the main or letterbox signal. Thus at the transmitter/encoder, the border signal processing circuitry can include a high-pass filter to remove those portions of the border signal corresponding to horizontal frequencies of less than 0.4 NHz as these frequencies are never used in the receivers.

The circuit of Figure 10 will now be described. This shows in a little more detail the relevant circuits of a second encoder embodying the invention. The upper part of the figure shows at (a) the signal processing circuitry 100 used to generate the lines to form the chrominance components in the borders. The U component is received at a terminal 102 and applied in succession to : a vertical filter 104 having a filter function VFl1 a sub-sampler 106 having a sub-sampling function S1, a horizontal filter 108 having a filter function HF1, a super-sampler 110 having a super-sampling function S2, and a horizontal filter 112 having a filter function HF2. The output of the horizontal filter 112 is then applied to a line splitter 114 which re-arranges each line applied to it as two output lines.The resultant lines are applied both directly and through a one-line delay 116 to a selector switch 118.

Similar processing takes place in circuitry 120 on the V signal received at an input 122. The processed U and V signals are then supplied to an adder 124 and a subtractor 126, the outputs of which are selected by a selector switch 128 to provide an output 130. The switch operates so as to select the U+V signal from adder 124 on one line and the U-V signal from subtractor 126 on the next.

These are termed the plus (P) and minus (M) signals.

At (b) in Figure 10 is shown the processing applied to the U and V signals during the main part of the signal, in the letterbox.

The circuitry 150 includes an input 152 for receiving the U signal, which is applied to a horizontal super-sampler 154 with function S3 and thence to a horizontal filter 156 having a filter function HF3.

The V signal received at an input 162 receives similar processing in circuitry 160. The combining circuitry is the same as used at (a) and has the same reference numerals. The output is then provided at 170.

The functional operation of the circuit of Figure 10 will now be described in more detail with reference to Figures 12 to 15 of the drawings and Tables 1 to 4 appended to this description.

It is assumed that the inputs both to the upper part (a) of the figure showing the helper processing for the borders and the lower part (b) showing the main processing for the letterbox are U and V signals each sampled at 216 lines per picture, 50 Hz field rate with 2:1 interlace, and 461 samples per line. This corresponds to a signal which is line locked at twice the PAL subcarrier frequency. It is represented as a 216/50/2:1 #461 signal.

The functions VF1, HF1, HF2, HF3 and also VF2 which will be considered below are all implemented by tranversal filters. The structure of the transversal filter is well known and does not need to be illustrated for those skilled in the art. Very briefly, it comprises a series of delays providing a series of taps, and multipliers to multiply the signal at each tap by a respective coefficient. The outputs of the multipliers are then summed. The taps are numbered plus and minus from a centre tap numbered zero and preferably the coefficients are symmetrical about the centre tap.

To provide a horizontal filter the delays are one-sample delays whereas to provide a vertical filter the delays are one-line delays.

For example, the ouput of the vertical filter VF1 is seen from Table 4 to be made up (considering the first four decimal places of the coefficient values only) of the following, where In (n) is the input value at tap n: Out(n) = 0.1663 * In(n) + 0.1580 * [In (n-l) + In (n+1)] + 0.1347 * [In (n-2) + In (n+2)] + The asterisks indicate multiplication, and in this case of a vertical filter the numbers n, n-l etc. are line numbers.

The effects of the transversal filters are to provide filter functions. In the case of simple horizontal and vertical filters these functions can be plotted on graphs as shown in Figures 12 to 16.

Thus reverting to Figure 10, vertical filter 104 with function VF1 is a vertical filter with a nominal cut-off frequency of 9 c/aph, as may be seen from Figure 15. The coefficients giving this function are set out in Table 4 below. This filter simply removes unwanted components. The vertical sub-sampler 106 has a function S1 which is a 6:1 vertical sub-sampling operation. In its simplest form, it simply discards all lines except lines 1, 7, 13, 19 etc. in each field; the conditioning provided by vertical filter 104 means that this simple sampling operation is adequate. In any event is produces a 36/50/2:1 #461 signal with just 36 lines.

This signal is applied to the horizontal filter 108 having function HF1. This is a high-pass filter with a nominal cut-off of around 0.5 MHz as shown in Figure 12. The filter coefficients are given in Table 1 below. The cut frequency of 0.5 MHz is just inside the difference between the color subcarrier frequency of 4.43 NHz and the PAL system B/G upper cut-off frequency of 5 MHz. It would be preferable to lower the cut-off frequency of the filter, but this would increase the number of taps and hence the cost.

The high-pass filtered signal is now applied to the horizontal super-sampler 110 which has a function 52. This provides a 4:1 horizontal super-sampling operation, inserting zeros to produce a 36/50/2:1 21844 signal. That is to say the pixels numbered 1, 2, 3 etc. in the input become pixels numbered 1, 5, 9 etc. in the output. This provides an input to the filter 112 having a horizontal filter function HF2 illustrated in Figure 13 and the transversal filter coefficients for which are given in Table 2 below. This filter provides an interpolation function for the horizontal up-conversion, so that the zero values inserted in super-sampler 110 are replaced by actual signal values.

After filtering in the filter 112, the 1844-sample lines are split into two lines of 922 samples by the line splitter 114. The samples are thus rearranged as a 72/50/2:1 #922 signal.

The U and V helper signals both of 72/50/2:1 #922 format are then converted to a 144/50/2:1 P/M signal, by repeating each line, and forming the sum and difference in circuits 124, 126. This signal is then split into two groups of 72 lines to be transmitted during the black-bar or border periods, above and below the letterbox.

Some consequential processing to the main signal during the letterbox is desirable as shown at (b) in Figure 10. Here the super-sampler 154 with function S3 provides a 2:1 horizontal super-sampling or up-conversion to produce a 216/50/2:1 #922 signal, by the insertion of a zero sample between each pair of samples.

The low pass filter 156 with function HF3 then interpolates to produce a stream of samples in which the zero samples are replaced by interpolated samples. The coefficients for filter 156 are given in Table 3 below, and the filter response is shown in Figure 14.

After this filtering, the two 216/50/2:1 U and V signals are converted to 432/50/2:1 P/M format by delay 116 and adder 124 and subtractor 126, in similar manner to the signal in the borders.

The resultant 170 is then used to form the letterbox chrominance component on the central 432 active lines of each picture, i.e. the central 216 active lines of each field.

As shown in Figure 10 there has been a 4:1 up-sampling in the horizontal direction, rather than just a 2:1 up-sampling as might have been expected. In theory this would quadruple the horizontal detail, but in fact the broadcast (or recording) channel capacity is not sufficient to carry the high frequencies that result. Only about half the frequency band can in fact be transmitted. The spread of the information over two lines rather than one is the crucial step which provides for a doubling of the horizontal detail. A 2:1 up-sampling would be sufficient.

The 4:1 up-sampling in sampler 110 is used because 922 samples per line are required by the phase-segregated clean PAL system with which the encoder is used, as will be described below.

The decoder 200 of Figure 11 will now be described. An input 202 receives the separated border P/M signal and applies it by means of a switch 204 directly and through a delay 206 to a summer 208 and a differencer 210. The output of the summer 208 now represents the U signal and is applied to a line-joining circuit 212 and then in turn to a horizontal filter 214 which has the same function HF2 as the horizontal interpolation filter 112 at the coder, a horizontal sub-sampler 216, and a subtractor 218. The subtractor receives a second signal from the letterbox circuitry, as described below. The output of the subtractor is applied through a vertical super-sampler 220 having a function S5 and then a vertical filter 222 having a function VF2. The output of filter 222 is then added to the main signal in an adder 224.

The main P/M signal transmitted during the letterbox is received at an input 232. The signal is applied by means of a switch 234 directly and through a delay 236 to a summer 238 and a differencer 240. The output of the summer 238 now represents the U signal and is applied in turn to a horizontal filter 242 having the same function HF3 as the interpolating filter 156 at the coder, and a horizontal sub-sampler 244. The output of the horizontal sub-sampler 244 provides a second input to the adder 224, the output 248 of which is the enhanced U output signal.The output of the horizontal sub-sampler 244 is also applied through a vertical filter 246 having the same filter function VF1 as the filter 104 in Figure 10 to a vertical sub-sampler 250 having the same function S1 as the sub-sampler 106 in Figure 10 and a horizontal high-pass filter 252 having the same function HF1 as the filter 108 in Figure 10. The output of filter 252 is applied to the inverting or subtractive input of subtractor 218.

The outputs of the subtractors 210 and 240 comprise the V component for the borders and the letterbox respectively. These are applied to similar processing circuits 260, 262 to those used for the U signal, and their outputs are added in an adder 264 to provide an enhanced V output 266. The circuit 262 provides a signal to the subtractor in circuit 260 similarly to the signal to the subtractor 218 in the U channel. The circuits for the V signal are not illustrated or described in detail as they are identical to the circuits for the U channel.

The operation of the decoder of Figure 11 will now be described. The signal applied to the input 202 during the borders is of 144/50/2:1 format and the signal applied to the input 232 during the letterbox is of 432/50/2:1 format. In each case there are nominally 922 samples per line though the transmission link may have reduced the effective bandwidth quite considerably. The signals are in P/M format with U+V and U-V on alternate lines.

Each signal is converted to U and V by adding and subtracting pairs of adjacent lines.

The letterbox U and V signals are each applied to horizontal filter 242 which pre-filters the signal before a 2:1 horizontal sub-sampling in the sub-sampler 244. This gives a signal in 216/50!2:1 format with 461 samples per line, which is applied as the main input to adder 224.

The border signals at the outputs of circuits 208, 210 are each of 72/50/2:1 format. The line-joining circuit 212 now joins pairs of lines horizontally to produce a 36/50/2:1 signal with double the number of samples on each line, i.e. 1844 samples per line. The horizontal filter 214 filters these long lines with the same interpolation filter as used at the encoder, and the sub-sampler 216 provides a 4:1 horizontal sub-sampling to reduce the number of samples to 461 samples per line.

Now, the helper signal from sub-sampler 216 contains some frequencies that are also present in the main letterbox signal.

These are principally the frequencies with vertical frequency below 9 c/aph and horizontal frequencies in the range about 0.4 to 0.5 MHz. It is desirable to remove this duplication. This is achieved by extracting these signal components from the main letterbox channel and subtracting them from the helper signal derived from the borders.

To this end the main channel signal is applied through vertical filter 246, vertical sub-sampler 250, and horizontal high-pass filter 252, to the subtractive input of subtractor 218.

The circuits 246, 250 and 252 subject the main letterbox signal to the same processing that was applied in the transmitter in the generation of the border signal. By subtracting the resultant signal from the border signal the desired cross-fade between the main and border signal contributions is achieved.

The output of the subtractor 218 is applied through the vertical super-sampler 220 which provides a 6:1 super-sampling, and vertical filter 222, to spread the information from one of the long lines over six picture lines. The super-sampler gives a 216/50/2:1 signal, with active data on lines 1, 7, 13 etc. of each field. The filter 222 is used to interpolate the other lines.

The coefficients used for filter 222 are given in Table 5 below and its frequency response is shown in Figure 16.

It would be possible for the subtractor 218 to be located after the super-sampler 220 and vertical filter 222, in which case a second similar super-sampler and vertical filter would be included between the horizontal filter 252 and the subtractor 218. The subtractor 218 and adder 224 could then be constituted by a single combining circuit. The arrangement illustrated is however more economical on components.

The chrominance signal which is generated in accordance with the embodiments described is suitable for transmission by a phase-segregated transmission such as that proposed for the luminance signal in our earlier application No. PCT/GB92/01988. A preferred phase-segregated transmission system is that known as the Weston Clean PAL coding system and is described in United Kingdom Patent Specifications GB-A-2 044 577 and GB-A-2 113 037. The operation is discussed in more detail in "A Compatible Improved PAL System", EBU Review, February 1986.

In the Weston system, the luminance signal is sampled at 2fsc (twice the color subcarrier frequency) and the chrominance signal is in P/M form and is sampled at fsc. By use of this system, decoding and recoding can take place without any cross-interference between the luminance and chrominance components.

The present system may therefore exploit the ability of Weston Clean PAL based systems, or may include horizontal filters to separate any signals modulated on to the color subcarrier for luminance enhancement purposes from the band which is used for chrominance signals. The method is therefore suitable for extending the horizontal chrominance resolution in a wide range of enhanced television systems based on composite signal coding methods. Also, since it is not necessary to maintain compatibility during the black bars, the method used in the black bars could be entirely different from that used for the main signal e.g. a PAL-like system could be used in the black bars with SECAM, NTSC or MAC transmissions for the main signal.

The present system is also suitable for use in conjunction with other transmission or recording methods, where the method involves the main part of the picture occupying less than the total number of lines available which the signal could occupy. For instance, if a wide screen picture is recorded using a VHS or SVHS video recorder, the chrominance resolution of the main part of the signal may be augmented by signals carried via the chrominance channel during the interval which is outside that associated with the active picture, namely during the blanking or framing period.

Thus if the coder of Figure 10 and decoder of Figure 11 are used in a Weston Clean PAL phase-segregation system, the output of the coder of Figure 10 is applied to the chrominance input of a Weston PAL assembler. At the receiver, a Weston PAL splitter provides a chrominance output which is demodulated and used by the decoder of Figure 11.

As noted above, it is in order to supply a 922 samples per line signal for the Weston phase segregated Clean PAL system that a 4:1 horizontal up-sampling is used at the encoder. If this were not a requirement, the encoder of Figure 10 and the decoder of Figure 11 could be simplified as follows: (i) the circuits which operate on the main letterbox signal could be omitted, these are 154 and 156 at the encoder and 242 and 244 at the decoder; (ii) the sampler 110 can become a 2:1 up-sampler; and (iii) the filter function HF3 of Figure 14 and Table 3 is then used in place of the filter function HF2.

By way of illustration, the minimum sample rates that could support the frequencies of interest in PAL system I are 222 samples per line at the encoder input and 111 samples per line at the encoder output.

A third embodiment of the invention will now be described with reference to Figures 17 and 18. This is designed specifically for use with the Weston Clean PAL system and the luminance helper system of our above-mentioned International Patent Application No. PCT/GB92/01988. Again, the system provides a 2:1 horizontal frequency enhancement to the color so as to achieve in excess of 1 MHz horizontal chrominance resolution for PAL system B (5 MHz overall bandwidth) or 2 MHz resolution for system I (5.5 MHz overall bandwidth). It is also assumed that the chrominance signal has been processed as necessary to achieve a good resolution as described in the above patent application No. PCT/GB92/01988.

The main features of the basic system described in our earlier application are that a wide screen picture of high resolution is coded so that the luminance signal occupies only 432 of the 576 active lines, and signals above 4.5 MHz are separately conveyed above and below the active picture using a band up to approximately 4.5 MHz. The chrominance component for the main part of the signal is coded to form a Weston Clean PAL composite signal with the luminance frequencies occupying the band below about 4.1 MHz. A combing band exists between about 4.1 and 4.7 MHz in which chrominance and luminance overlap, and signals above this band are modulated chrominance. Horizontal, vertical and temporal filters are used in pre-filtering the luminance and chrominance signals at the coder, and in post-filtering to reconstruct a good signal for display at the decoder.The chrominance capacity above the luminance band within the border can now be used to convey useful information to improve the horizontal chrominance detail.

The coder 300 shown in Figure 3 has an input 302 which receives the main chrominance signal. This comprises U and V components of 432 line-sequential color signals. The U and V signals are applied to identical U and V signal processing circuits of which only the U channel is shown in detail. The signals are first applied to a vertical low-pass filter 304 with function F1 which cuts off at 18 cycles per picture height to restrict the vertical resolution of the signal. In this instance it is assumed that non-interlaced or progressively-scanned signals are received at the input 302, so the vertical frequency is numerically doubled compared with the previous embodiments. The output of filter 304 is applied to a sampler 306 with sampling function S1 which generates 36 sequential lines from the 432 lines received at the input.This is a 6:1 re-sampling or sub-sampling operation. A temporal filter 308 having a filter function F2 receives the output of the sub-sampler and applies it to an interlace sampler 310 with sampling function S2. This provides an output in 36-line interlaced form, which is applied to a horizontal high-pass filter 312 with filter function F3. This removes low horizontal chrominance frequencies, including any DC component, from the signal. (Note that the functions S1 etc. do not correspond to those of the preceding embodiment.) The V component passes through similar processing circuitry 314, and the outputs of circuits 312 and 314 are applied to a multiplexing store 316 having a function M1. This reads in the similarly-processed U and V signals and stretches them horizontally so that each television line occupies two lines of the chrominance in the border.In this way, the chrominance channels of the border above and below the main signal have applied the multiplex chrominance information in groups of four consecutive modulated lines as follows: U left hand of line, U right hand of line, V left hand of line, V right hand of line.

The output of the multiplex store 316 is applied to a chrominance modulator 318 where they are modulated onto a subcarrier of the color subcarrier frequency fsc. The modulated signal is applied to a filter 320 which is a square-root anti-symmetric high-pass filter having filter function F4.

The luminance helper signal is received at an input 322 and applied to an equivalent square-root anti-symmetric low-pass filter 324. This luminance border signal is then added to the modulated chrominance border signal in an adder 326 to provide an output 328 for transmission or recording in the blank signal lines.

In the corresponding decoder 330 shown in Figure 18 the luminance and chrominance border signals are received at an input 332. They are separated using a square-root anti-symmetric high-pass filter 334 for the chrominance component and a square-root anti-symmetric low-pass filter 306 for the luminance component.

The luminance helper signal is then processed in accordance with our application No. PCT/GB92/01988. The separated chrominance signal is applied to a demodulator 338 where it is demodulated from its fsc subcarrier and applied to a demultiplex store 340. The demultiplex store 340 derives 36 lines of interlaced U and V detail information respectively and provides the necessary retiming. The U component is applied to an interlace-to-sequential filter 342 having a function F8. This is a temporal low-pass filter and provides a 36 line progressively scanned (i.e. non-interlaced) signal. This signal is applied to a vertical interpolation filter 344 with filter function F9, which generates a signal equivalent to the full 432 line progressive main chrominance signal.The vertical low-pass filter 344 removes alias frequencies, and provides a vertical cross-fade between the chrominance information derived from the main signal and the chrominance detail derived from the border.

Chrominance signals from the main or letterbox part of the received signal are received at an input 346 and applied to a horizontal high-pass filter 348 having a filter function F3 which is identical to the filter function used by the filter 312 to remove low frequencies from the border in the coder 300 of Figure 17. The output of filter 348 is applied to a vertical low-pass filter 350 having a filter function F9 which is identical to that used in the filter 344 for the border signal to define the cross-fade. A subtractor 352 receives the main chrominance signal at input 346 and subtracts from it the output of the filters 348, 350. In this way it is ensured that the action of the filters 348, 350 complements the action of the equivalent filters in the path of the chrominance helper signal.The signals from the main path and from the chrominance helper path are then added in an adder 354 to provide an enhanced output 356 for display.

As with the preceding embodiments, the chrominance channel derived by the third embodiment of Figures 17 and 18 enables improved horizontal luminance resolution for a range of vertical frequencies substantially less than that of the main signal chain.

In a system where the horizontal chrominance resolution is extended by a factor of 2 and the black bars occupy 25% of the total number of active television lines available, this horizontal resolution enhancement is available for 1/6 of the vertical chrominance frequencies which are broadcast during the main signal intervals.

If the horizontal resolution enhancement factor was to be 3:1, the enhancement could be applied for vertical frequencies up to 1/9 of the total vertical frequency range of the main signal channel. The exchange of vertical and horizontal frequency for this enhancement derives from the time expansion of chrominance signals added during the black bars. With no time expansion the black bars would be able to convey 1/3 of the number of lines of information compared to the main signal channel. In general, the horizontal frequency resolution achieved by the present method will be a maximum of r times that of the normal channel and this will apply for 1/3r times the normal vertical frequency range of the chrominance signal in the main channel.

By use of this system, chrominance signals at low horizontal and vertical frequencies will be available in the receiver, having been conveyed both as part of the main signal and as part of the border signal. The chrominance signals are combined in the receiver in such a way that there is a smooth combination of signals within this shared band. The most convenient method of achieving this is by vertical filtering which effects a cross-fade between the two signals as vertical frequency increases. This cross-fade can be such that all low vertical frequency chrominance signals are received via the black bar signal, where greater horizontal resolution is available, and all higher vertical frequencies, which cannot be transmitted via the black bar system, are received via the main color signal channel.However, it can be additionally beneficial to prevent very low horizontal chrominance signals from being modulated during the black bar interval, as was described in relation to the previous embodiment. This should avoid any possible cross-talk between this and other signals modulated onto the color carrier. It also enables the method to be applied to systems where little or no separation between color and luminance modulated signals is available.

The various circuits can be re-arranged or modified in numerous ways and in particular the filters can be re-arranged or coalesced. The various filter functions are only shown separately for ease of explanation. Also, the individual circuits are shown as discrete hardware units, but they could be implemented in software. In this case the schematic circuit diagrams should be regarded as being functional flow diagrams.

TABLE 1 Coefficients of filter HF1 Coefficient Coefficient position value (x) -22 0.00109555 -21 0.00141241 -20 0.00158850 -19 0.00153592 -18 0.00116099 -17 0.00037076 -16 -0.00091999 -15 -0.00278165 -14 -0.00526216 -13 -0.00838092 -12 -0.01212377 -11 -0.01644009 -10 -0.02124169 -9 -0.02640499 -8 -0.03177487 -7 -0.03717174 -6 -0.04240001 -5 -0.04725853 -4 -0.05155139 -3 -0.05509969 -2 -0.05775129 -1 -0.05939059 0 0.93914199 1 -0.05939059 2 -0.05775129 3 -0.05509969 4 -0.05155139 5 -0.04725853 6 -0.04240001 7 -0.03717174 8 -0.03177487 9 -0.02640499 10 -0.02124169 11 -0.01644009 12 -0.01212377 13 -0.00838092 14 -0.00526216 15 -0.00278165 16 -0.00091999 17 0.00037076 18 0.00116099 19 0.00153592 20 0.00158850 21 0.00141241 22 0.00109555 23 0.00000000 TABLE 2 Coefficients of filter HF2 Coefficient Coefficient position value (x) -15 -0.00036693 -14 -0.00526178 -13 -0.00516391 -12 0.00000000 -11 0.00906980 -10 0.01650757 -9 0.01484835 -8 0.00000000 -7 -0.02368355 -6 -0.04248899 -5 -0.03864074 -4 0.00000000 -3 0.07106531 -2 0.15537941 -1 0.22373545 0 0.24999994 1 0.22373545 2 0.15537941 3 0.07106531 4 0.00000000 5 -0.03864074 6 -0.04348899 7 -0.02368355 8 0.00000000 9 0.01484835 10 0.01650757 11 0.00906980 12 0.00000000 13 -0.00516391 14 -0.00526178 15 -0.00036693 TABLE 3 Coefficients of filter HF3 Coefficient Coefficient position value (x) -7 -0.01031983 -6 0.00000000 -5 0.03601265 -4 0.00000000 -3 -0.08735895 -2 0.00000000 -1 0.31166613 0 0.50000000 1 0.31166613 2 0.00000000 3 -0.08735895 4 0.00000000 5 0.03601265 6 0.00000000 7 -0.01031983 TABLE 4 Coefficients of filter VF1 Coefficient Coefficient position value (y) -17 0.00110280 -16 0.00292240 -15 0.00481629 -14 0.00569180 -13 0.00434630 -12 0.00000000 -11 -0.00714764 -10 -0.01556273 -9 -0.02242264 -8 -0.02415188 -7 -0.01736889 -6 0.00000000 -5 0.02778703 -4 0.06316689 -3 0.10096560 -2 0.13468528 -1 0.15800282 0 0.16633309 1 0.15800282 2 0.13468528 3 0.10096560 4 0.06316689 5 0.02778703 6 0.00000000 7 -0.01736889 8 -0.02415188 9 -0.02242264 10 -0.01556273 11 -0.00714764 12 0.00000000 13 0.00434630 14 0.00569180 15 0.00481629 16 0.00292240 17 0.00110280 TABLE 5 Coefficients of filter VF2 Coefficient Coefficient position value (y) -17 0.00110280 -16 0.00292240 -15 0.00481629 -14 0.00569180 -13 0.00434630 -12 0.00000000 -11 -0.00714764 -10 -0.01556273 -9 -0.02242264 -8 -0.02415188 -7 -0.01736889 -6 0.00000000 -5 0.02778703 -4 0.06316689 -3 0.10096560 -2 0.13468528 -1 0.15800282 0 0.16633309 1 0.15800282 2 0.13468528 3 0.10096560 4 0.06316689 5 0.02778703 6 0.00000000 7 -0.01736889 8 -0.02415188 9 -0.02242264 10 -0.01556273 11 -0.00714764 12 0.00000000 13 0.00434630 14 0.00569180 15 0.00481629 16 0.00292240 17 0.00110280

Claims (67)

1. CLAIMS 1. A method of transmitting a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines which does not carry the active picture is used to carry helper information, there being n times as many lines in the main group as there are in the auxiliary group, and the auxiliary line group carrying as its chrominance component a chrominance helper signal; the method comprising deriving the chrominance helper signal from the chrominance component in the main line group to form output lines by the following steps, taken in either order or simultaneously: (a) horizontally extending each line by a factor r so that the chrominance signal from a line of the main group is extended over r output lines; and (b) vertically sub-sampling the chrominance component of the lines in the main line group to produce a smaller number of output lines, which is smaller than the number of lines in the auxiliary group by at least a factor r times n.
2. 2. A method according to claim 1, including the step of low-pass filtering the signal with a vertical low-pass filter.
3. 3. A method according to claim 1, including the step of high-pass filtering the signal with a horizontal high-pass filter.
4. 4. A method according to claim 1, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
5. 5. A method according to claim 1, in which the auxiliary group of lines comprises lines in the video blanking interval.
6. 6. A method according to claim 1, in which the factor n is three.
7. 7. A method according to claim 1, in which the factor r is two.
8. 8. A method according to claim 1, in which the said steps are applied separately to two color component signals U and V, and further comprising combining the U and V outputs into the format U+V and U-V on alternate lines.
9. 9. Apparatus for transmitting a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines which does not carry the active picture is used to carry helper information, there being n times as many lines in the main group as there are in the auxiliary group, and the auxiliary line group carrying as its chrominance component a chrominance helper signal; the apparatus comprising means for deriving the chrominance helper signal from the chrominance component in the main line group to form output lines by the following steps, taken in either order or simultaneously: (a) horizontally extending each line by a factor r so that the chrominance signal from a line of the main group is extended over r output lines; and (b) vertically sub-sampling the chrominance component of the lines in the main line group to produce a smaller number of output lines, which is smaller than the number of lines in the auxiliary group by at least a factor r times n.
10. 10. Apparatus according to claim 9, further comprising vertical low-pass filter means for low-pass filtering the chrominance signal.
11. 11. Apparatus according to claim 9, further comprising horizontal high-pass filter means for high-pass filtering the chrominance signal.
12. 12. Apparatus according to claim 9, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
13. 13. Apparatus according to claim 9, in which the auxiliary group of lines comprises lines in the video blanking interval.
14. 14. Apparatus according to claim 9, in which the factor n is three.
15. 15. Apparatus according to claim 9, in which the factor r is two.
16. 16. Apparatus according to claim 9, including separate deriving means for processing two color component signals U and V respectively, and further comprising means for combining the U and V outputs into the format U+V and U-V on alternate lines.
17. 17. A method of transmitting a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines is used to carry helper information, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the chrominance helper signal consisting of output lines derived from the chrominance component in the main line group, the chrominance component output in the main group comprising frequency components up to a maximum horizontal frequency and a maximum vertical frequency; and the method comprising deriving from the main line group frequency components of horizontal frequency higher than the said maximum horizontal frequency for vertical frequenices which are low relative to the said maximum vertical frequency, to form the chrominance helper signal.
18. 18. A method according to claim 17, in which the chrominance helper signal is derived from the main line group by reducing the effective horizontal frequencies of the chrominance components by a factor r for transmission as the chrominance helper signal.
19. 19. A method according to claim 17, in which the frequency components from the main line group used in deriving the chrominance helper signal do not include at least some of the components of low horizontal frequency.
20. 20. A method according to claim 17, in which the active picture corresponds to a picture of aspect ratio 4:3 and the main line group corresponds to a picture of aspect ratio 16:9.
21. 21. A method according to claim 17, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
22. 22. Apparatus for transmitting a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines is used to carry helper information, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the chrominance helper signal consisting of output lines derived from the chrominance component in the main line group, the chrominance component output in the main group comprising frequency components up to a maximum horizontal frequency and a maximum vertical frequency; and the apparatus comprising means for deriving from the main line group frequency components of horizontal frequency higher than the said maximum horizontal frequency for vertical frequencies which are low relative to the said maximum vertical frequency, to form the chrominance helper signal.
23. 23. Apparatus according to claim 22, in which the means for deriving the chrominance helper signal from the main line group reduces the effective horizontal frequencies of the chrominance components by a factor r for transmission as the chrominance helper signal.
24. 24. Apparatus according to claim 22, in which the means for deriving the chrominance helper signal from the main line group excludes at least some of the components of low horizontal frequency.
25. 25. Apparatus according to claim 22, in which the active picture corresponds to a picture of aspect ratio 4:3 and the main line group corresponds to a picture of aspect ratio 16:9.
26. 26. Apparatus according to claim 22, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
27. 27. A method of receiving a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines which does not carry the active picture is used to carry helper information, there being n times as many lines in the main group as there are in the auxiliary group, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the method comprising processing a received chrominance helper signal by the following steps (a) and (b) taken in either order or simultaneously: (a) horizontally compressing each line by a factor r so that the chrominance signal from r lines is compressed onto one output line; and (b) vertically super-sampling the chrominance component of the lines in the received helper signal to produce a larger number of output lines, which is larger than the number of lines in the auxiliary group by at least a factor r times n; and subsequently adding the thus-processed chrominance helper signal to the chrominance component received in the main line group.
28. 28. A method according to claim 27, including the step of vertically sub-sampling the chrominance component of the lines in the received main line group to produce a smaller number of lines, which is smaller than the number of lines in the auxiliary group by at least a factor of r times n,- and subtracting the resultant from the horizontally compressed received chrominance helper signal.
29. 29. A method according to claim 28, including the step of filtering the vertically sub-sampled signal with a horizontal high-pass filter before the subtraction.
30. 30. A method according to claim 28, in which the subtraction takes place on the horizontally compressed received chrominance helper signal before the vertical super-sampling.
31. 31. A method according to claim 27, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
32. 32. A method according to claim 27, in which the auxiliary group of lines comprises lines in the video blanking interval.
33. 33. A method according to claim 27, in which the factor n is three.
34. 34. A method according to claim 27, in which the factor r is two.
35. 35. A method according to claim 27, in which the received chrominance helper signal is in the format U+V and U-V on alternate lines, and further comprising as an initial step deriving separated U and V signals from the U+V and U-V signals.
36. 36. Apparatus for receiving a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines which does not carry the active picture is used to carry helper information, there being n times as many lines in the main group as there are in the auxiliary group, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the apparatus comprising means for processing a received chrominance helper signal by the following steps (a) and (b) taken in either order or simultaneously:: (a) horizontally compressing each line by a factor r so that the chrominance signal from r lines is compressed onto one output line, and (b) vertically super-sampling the chrominance component of the lines in the received helper signal to produce a larger number of output lines, which is larger than the number of lines in the auxiliary group by at least a factor r times n; and means for subsequently adding the thus-processed chrominance helper signal to the chrominance component received in the main line group.
37. 37. Apparatus according to claim 36, further comprising means coupled to receive the chrominance component of the main line group for vertically sub-sampling the chrominance component of the lines in the received main line group to produce a smaller number of lines, which is smaller than the number of lines in the auxiliary group by at least a factor of r times n, and means for subtracting the resultant from the horizontally compressed chrominance helper signal.
38. 38. Apparatus according to claim 37, further comprising means coupled between the vertical sub-sampling means and the subtracting means for filtering the vertically sub-sampled signal with a horizontal high-pass filter function.
39. 39. Apparatus according to claim 37, in which the processing means comprises horizontal-compression means coupled to receive the received chrominance helper signal and vertical super-sampling means coupled between the horizontal-compression means and the adding means.
40. 40. Apparatus according to claim 39, in which the subtracting means is located between the horizontal-compression means and the vertical super-sampling means.
41. 41. Apparatus according to claim 36, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
42. 42. Apparatus according to claim 36, in which the auxiliary group of lines comprises lines in the video blanking interval.
43. 43. Apparatus according to claim 36, in which the factor n is three.
44. 44. Apparatus according to claim 36, in which the factor r is two.
45. 45. Apparatus according to claim 36, further comprising means for receiving the chrominance helper signal in the format U+V and U-V on alternate lines and for deriving therefrom separated U and V signals for application to respective processing means.
46. 46. A method of receiving a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines is used to carry helper information, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the method comprising processing the received chrominance helper signal by increasing the effective horizontal frequency of the chrominance components by a factor r, and adding the thus-processed chrominance helper signal to the chrominance component received in the main line group to provide a chrominance output signal.
47. 47. A method according to claim 46, in which the adding step comprises a cross-fade function between the chrominance frequency components received through the main line group and the chrominance frequency components derived from the chrominance helper signal.
48. 48. A method according to claim 46, in which chrominance frequency components of low vertical frequency and low horizontal frequency are derived from the main line group.
49. 49. A method according to claim 46, in which chrominance frequency components of low vertical frequency received in the main line group are subtracted from corresponding components received in the auxiliary line group.
50. 50. A method according to claim 46, in which the active picture corresponds to a picture of aspect ratio 4:3 and the main line group corresponds to a picture of aspect ratio 16:9.
51. 51. A method according to claim 46, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
52. 52. Apparatus for receiving a color video signal of the type formed by fields of scanned lines and which includes separate luminance and chrominance components, in which the active picture occupies only some of the scanned lines of each field forming a main group of lines, and an auxiliary group of lines is used to carry helper information, and the auxiliary line group carrying as its chrominance component a chrominance helper signal, the apparatus comprising means for processing the received chrominance helper signal by increasing the effective horizontal frequency of the chrominance components by a factor r, and means for adding the thus-processed chrominance helper signal to the chrominance component received in the main line group to provide a chrominance output signal.
53. 53. Apparatus according to claim 52, in which the adding means comprises a cross-fader for cross-fading between the chrominance frequency components received through the main line group and the chrominance frequency components derived from the chrominance helper signal.
54. 54. Apparatus according to claim 52, including means for deriving for the output signal chrominance components of low vertical frequency and low horizontal frequency from the main line group.
55. 55. Apparatus according to claim 52, further comprising means for subtracting chrominance frequency components of low vertical frequency received in the main line group from corresponding components received in the auxiliary line group.
56. 56. Apparatus according to claim 52, in which the active picture corresponds to a picture of aspect ratio 4:3 and the main line group corresponds to a picture of aspect ratio 16:9.
57. 57. Apparatus according to claim 52, in which the main group of lines occupies a central part of the active picture area, and the auxiliary group of lines occupies border regions above and below the main group of lines in the picture.
58. 58. A chrominance video signal encoder, comprising: an input for receiving a chrominance video input signal; an output for providing an encoded chrominance video output signal; vertical sub-sampling means coupled between the input and the output for vertically sub-sampling the chrominance signal to produce a smaller number of lines therefrom, and horizontal expansion means coupled between the input and the output for extending each input line onto two or more output lines.
59. 59. An encoder according to claim 58, further comprising horizontal high-pass filter means coupled between the input and the output.
60. 60. An encoder according to claim 58, further comprising vertical low-pass filter means coupled between the input and the output.
61. 61. A chrominance signal video decoder, comprising: an input for receiving an encoded chrominance video input signal; an output for providing a decoded chrominance video output signal; horizontal compression means coupled between the input and the output for compressing the chrominance signal such that each output line is formed from two or more input lines; and vertical super-sampling means coupled between the input and the output for vertically super-sampling the chrominance signal to produce a larger number of lines therefrom.
62. 62. A decoder according to claim 61, in which the said input is a first input for receiving an encoded chrominance video input signal from a first group of lines, and including a second input for receiving a second video input signal from a second group of lines, and adding means coupled to the output and the second input for adding the decoded signal to the second video input signal.
63. 63. A decoder according to claim 62, in which the adding means comprises a cross-fader.
64. 64. A decoder according to claim 62, further comprising a subtractor having its non-inverting input coupled to the output of the horizontal compression means, and vertical sub-sampling means coupled to the second input to receive the second input signal and to produce a smaller number of lines therefrom, the inverting input of the subtractor being coupled to the output of the vertical sub-sampling means.
65. 65. A decoder according to claim 64, further comprising horizontal high-pass filter means coupled in series with the vertical sub-sampling means.
66. 66. A decoder according to claim 64, in which the horizontal compression means is connected to the input and the vertical super-sampling means is coupled between the horizontal compression means and the output.
67. 67. A decoder according to claim 66, in which the subtractor is connected between the horizontal compression means and the vertical super-sampling means.