GB2152782A - Television display - Google Patents

Abstract

An input video signal is displayed at an increased (eg doubled) field rate by repeating each field. In order to reduce spatial and movement distortions, a conventional interlaced signal is displayed as a group-interlaced picture in which each field of a displayed frame contains equally spaced groups of a plurality of spatially consecutive lines of the picture. Alternatively (Fig. 8), with a group-interlace inputsignal, a conventional interlaced display may be used. The group in each case is preferably a pair of lines. In each method a storeholding only one frame is required if read and write can be simultaneous. Further reduction of flicker etc. may be achieved if the lines of a group are made unequal in brightness. <IMAGE>

Description

SPECIFICATION Television display This invention concerns the display oftelevision signals. It is well knownthatdisplayon a cathode ray tube or device having similar persistence characteristics results in certain impairments, particularly large area flicker at the field frequency and inter-line twitter at the picture frequency. With conventional broadcast standards employing field rates of 50Hz (as in system I) or 60Hz (system M), such impairments can be objectionable, particularly with the greater brightness of modern displays.

It has been suggested (UK patent application no.

20501 09A) thatflicker can be reduced by using picture storage to permit scanning of the display vertically at twice the inputfield frequency. In one proposal, each field of an input signal is displayed twice in succession. Thus if the two fields of the nth picture (assuming 2:1 interlace) are designated An and Bn and input sequence A0B0A1 . . . . B1 is displayed as a sequence A0A0,B0B0,A1A1 where the Ao etc simply indicates the second display offield A,. A normal display is as shown in Figure 1,which illustrates four consecutive fields ofthe conventional display; at (a) the sequence ofthefields in time, and at (b) the appearance of the resultant display.The square blocks illustrate the appearance of a vertical line moving from leftto right, whilstthe arrows indicate the apparent horizontal movementfrom one input field to the next ie (for a 50 fields/s display) each 20ms. Figure 2 is a similar diagram forthe field repeat display at 100 fields/s, the divided block indicating that it appears twice. This requires only a single field store, and the large-area flicker is doubled, but the inter-line twitter is still atthe original picture frequency. Perturbed line (or field) scanning waveforms are required to achieve the even, even, odd, odd field sequence. The arrows and dots under (b) offig 2 etseq. now indicate the apparent movement per output field ie each 1 Om sec for 100 field-sec displays.In a second proposal, each frame is displayed twice in succession, viz A0B0Ao'Bo'A1 B1A1,B' etc. as can be seen from figure 3. Although this doubles both the frequency of the large-area flicker and inter-line twitter, it causes (as indicated by the arrows) an apparent reversal of motion once everyfourfields and therefore produces undesirable jerky effects in moving pictures.

In television standards currently in use, equal spacing of the lines in each field is employed. A pair-interlace system has, however, been proposed for high definition television (see Data Compression for High Definition TV: an NTSC Compatible Approach - G.G.Apple, H.E.Tsou. National Telesystems Conference. NTC '82. Systemsforthe Eighties, Galveston, TX, USA, 7-10 Nov. 1982 (NewYork: IEEE 1982) p.E1.3/1-4.). In this system, even fields carry lines 1,2, 5,6,9,10 etc whilst odd interlaced fields carry lines 3, 4,7,8,11,12... (wherethe numbers refertothe position of the lines in the picture, not the conventional sequence). This is illustrated in figure 4.The system is suggested in orderto facilitate down-conversion from (say) 1050 to 525 lines by simple line averaging.

The invention is defined in the appended claims.

Some embodiments of the present invention will now be described with reference to the accompanying drawings, in which: Figure 5 is a blockdiagramofaconversion apparauts; Figure6 is a diagram similartofigures 1 to3 illustrating the operation of one embodiment ofthe present invention; Figure 7 shows the appearance of consecutive fields ofthe resulting display; and Figures 8 and 9 are diagrams analogous to figures 6 and 7 illustrating a second embodimentofthe invention.

In one embodiment of the present invention, an input video signal in conventional 2:1 interlace format (at, say, 25 frame/second) is converted, for display, into a pair-interlacesignal at twice the input frame rate (50 frames/second). Figure 5 shows, in outline, a block diagram of a suitable apparatus. Video at an input 1 is entered into a frame store 2 under control of a store input address control unit 3, and read out attwice the frame rate with the aid of a store output address control unit4. The size of store required will depend upon whether reading or writing can take place simultaneously. If this is so, a minimum of one television frame of storage is sufficient.One possible format is illustrated in figure 6 using the same conventions as employed forfigures 1 to 4, AnBo indicating inputfields from which the displayed lines originate. The line pairing results in the simultaneous display lines from two fields ofthe original and results in a double image on moving pictures (as can be seen from figure 7 which shows the appearance of six displayed fields) but the mean position ofthis double image is now the sameforalternate pairs of display fields, so thatthe displayed motion is the same as that offigure 1, as indicated by the arrows which, in figure 6 (and figure 8) refer to the mean image positions.It should be noted, of course, thatthe illustrations shown are idealisations and in practice each component of these images will, due to the movement, be somewhat blurred.

Again, the large area flicker is doubled. Inter-line twitter is now replaced by inter-pair twitter, but at double the input picture frequency (50Hz), unlike that ofthe odd/odd/even/even sequence of figure 2. It is believedthattheincreased pitch ofthistwitter(double thatoftheconventional display)ismorethancompensated for subjectively by the higherfrequency.

As an alternative to the "staggered"frame scheme of figure 6 (A1Bo) (A1B1) (A2B1) etc, the sequence (A1B1) (air11 (A2B2) (A2B2) etc. could, of course. be used, which might be of advantage forthe display ofTV signals originated from cinefilm.

The cases considered so far all involve the use of each input field twice at full luminance value. Since both odd and even inputfields (A and B) are used for each outputfield, further improvement in movement rendition can be obtained by varying the amounts of A and B between a pair of output fields. For example, in figure 6, with A and Weighted by 1 and 1/3 respectively in one output field and vice versa in the next, the mean position ofthe double image in each such field changes by equal amounts, each of halfthat previously occurring. Thusthe same average rate of movement is displayed but with greater smoothness.

The values derive from the use of two inputfields only and the need to displace the positions from the centre of (A+B) by 1 1/4 of the distance between A and B (note that (1/2 + 1/4) = 3 (1/21/4)).

Thefurther processing described introduces degradations in spatial resolution and flicker. For this reason it may be considered necessary that such algorithms be made movement adaptive, reverting for example to the unmodified values in these parts of the picture containing no motion.

This is indicated in figure 5 buy a movement detector 5 and interpolator 6.

Suppose nowthatthe inputTVsignal employs line-pair interlaceperhaps an HDTVsignalas proposed in the above-mentioned paper-although the line standard used is of course irrelevantto the present discussion. Assume that the frame rate is 25Hz and we wish to drive a display at 50 frames/ second and interlaced 2:1.

In this case, the higher-rate display employs conventional 2:1 interlace, with a conversion from the pair-interlace input. The input signal is illustrated in figure 4, and the conversion in figure 8. The appear anceofthe resultantdisplay is illustrated by figure 9.

This is essentially the reverse processtothatillus- trated in figures 6 and7, and again a double image results on moving objects giving the same movement rendition as in figure 6.

Here, the large area flicker and inter-line twitter are doubled, with the inter-line twitter having, of course, a pitch of one line ratherthan two. It is recognised that we are dealing with a sampled system and signal reconstruction after processing. A variety of both spatial and temporal interpolations may be of use at the display and these can be treated as post-filtering operations and designed mathematically. Of course thefiltering action oftheeye/brain is partofthe whole.

However,the simplercasestend to be of greatest interest since strong filtering means a wider interpolating aperturewhich in the temporal case being considered here means using more than 2 input fields, hence: (a) more fields storage and hence cost and space; (b) greater smearing.

ltwould seem to be advantageous if the scheme selected for 100Hz display can be one which does not introduce movementerrors in the first place (as does the system offigure 3) since such electrical filtering as may be used can then be less strong.

The overai I resultwill also depend on the degree of pre-filtering used at the source. It must not be forgotten thatthe integrating action essential to all cameras in order to maximise sensitivity means that each field image A or B hasfundamental movement blurextending'rom one image to the next. In present practice other blurring (eg due to lag) occursaswell. In due coursethe latter will be avoided (CCD images) and also extra filtering will be introduced to reduce vertical aliasing (due to the raster structure) and possibly temporal aliasing also (wheel spoke effect).

It is considered that at the receiving end ofthe chain the differences between the simpler 100Hzdisplay schemes discussed earlier may be much more significantwhen such pre-filterings have been applied.

The phenomena discussed are also affected bythe non-linear relationship between the display luminance values and the transmitted signal, the latter being gamma corrected.

Claims (7)

1. Atelevision display apparatus comprising input means for receiving a video signal with a defined field rate; store means connected to the input means for storing the input video signal; display means for displaying a television picture; scan control meansforcausing said display means to scan with field rate increased from the corresponding rate of said input signal; and means for selecting in accordance with the scanned location from the stored signal portionsthe appropriate signal portion for application to the display means, characterised in thatthe store means is arranged for receipt of a signal in which the lines of each field correspond to equally spaced lines of the picture and the scan control means and selection means are arranged to produce displayed fields each containing equally spaced groups of a plurality of consecutive lines ofthe picture.

2. Atelevision display apparatus comprising input means for receiving a video signal with a defined field rate; store means connected to the input meansfor storing the inputvideosignal; display means for displaying a television picture; scan control meansforcausing said display means to scan with field rate increased from the corresponding rate of said input signal; and means for selecting in accordance with the scanned location from the stored signal portions the appropriate signal portion for application to the display means characterised in that the store means is arranged for receipt of a signal in which the lines of each field correspond to equally spaced groups of a plurality of consecutive lines of the picture and the scan control means and selection means are arranged to produce displayed fields each containing equally spaced lines of the picture.

3. A display apparatus according to claim 1,in which the input fields are 2:1 interlaced, and the said groups each contain a pair of consecutive lines of the picture.

4. A display apparatus according to claim 2 in which the said groups each contain a pair of consecutive lines of the picture, and the displayed picture is 2:1 interlaced.

5. A display apparatus according to claim 1,2,3 or 4 in which the scan control means is arranged to doublethefield rate.

6. A display apparatus according to claim 5, in which the selecting means is controlledto generate in respect of each frame ofthe input, two frames for display, the first containing lines from that frame of the input signal, and the second containing linesfrom the second field ofthatframe of the input and the first field ofthefollowing frame of the input.

7. A display apparatus according to any one of the preceding claims, including interpolation means.

8 A display apparatus according to claim 7, including a movement detector arranged to permit operation ofthe interpolation means only in those parts ofthe picture which contain a representation of movement.