JPH07500711A - Improvements in television systems - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] Retirement from television viewing system The back of a scream of surprise The present invention relates to apparatus and methods for transmitting and receiving television signals, and more particularly to improved television quality compared to the signals provided by current transmission systems; It is related to the transmission of signal.

This invention may be applicable to the NTSC system or other systems, but for convenience Regarding the L method, more specifically regarding the system using the 16:9 image aspect ratio. explained. The signal can be displayed in a special 16.9 aspect ratio receiver. , or an existing 43 with what is known as a "letterbox" display device. It can be displayed in an aspect ratio receiver.

In the letterbox display device illustrated in Figure 24, a wide aspect ratio image is The image is displayed in the vertical center of the screen at full horizontal width, and the top and bottom of the image are black. Be erased. This subjective effect is already evident in the fact that motion picture film is transmitted by television broadcasting. Occasionally seen when believed.

432 active lines, which are then letterboxed in a 4.3 aspect ratio receiver. It is already 625150/2: ill semi-P A,! − Displaying 6 images known in the transmission system on a 16.9 aspect ratio receiver , the image is expanded vertically by increasing the vertical scan of the raster. be able to. However, this makes the lines appear further apart on the screen. This results in the desired effect, i.e. the subjective image quality. This is the exact opposite of improving quality. Decreased definition in the horizontal direction be.

Another suggestion is to include additional information to give more vertical detail. Contains black bands at the top and bottom of the image. In this connection, a UK patent has been issued. Application GB-A-221,3342 is letterboxed with a reduced number of lines. Transmit an image and add low-amplitude invisible additional information to the top or bottom of the image. It is described that the data is transmitted in the erasure area that corresponds to 1.

Other legacy documents in this analysis include:

British patent number 1! GB-A-2186165 reduces scan amplitude of scan raster A CRT (Cathode Ray Tube) table that displays wide aspect ratio signals on a conventional display device. It describes the display device. However, this is due to the CRT circuitry of existing receivers. Requires changes.

According to the UK patent @GB-A-2203011 is a decoder similar to that at the receiver side. It describes a transmitter that also includes. Signal that should be sent as a regular video signal is decoded and the decoded signal is compared with the video input signal. Encoding/decoding operation Any difference representing the impairment introduced in the vertical blanking period, e.g. is sent along with the video signal and sent to the receiver at the receiver to counteract the effects of impairment. used.

British patent RGB-A-2238202 is based on HDT as main signal and complementary signal. It describes the transmission of V (high-definition television) signals. The division is a line to the main signal. in the image, accounting for three quarters of the total number of images, e.g. 432 out of 576. This is achieved by preserving the lower three-quarters of the spectrum in vertical detail. . The complementary signal is the spectral remainder of the vertical detail, which has a quarter of the number of lines. It holds the top quarter of the ridge.

9 fakes for light The invention in its various aspects is defined later in the independent claims and is now hereby incorporated by reference. Reference should be made to: Advantageous features are defined in the dependent claims.

In selected embodiments of this invention, which will be explained in more detail later on, PAL wide Although the aspect ratio luminance signal is transmitted in letterbox format, its main signal component is occupies the central part of the screen and the auxiliary (helper) signal occupies the upper and lower edges (voter). is occupying. The input signal is sampled by 4/3X2f, and divided into two parts. , so the low horizontal frequency component is 2f1 in the principal component, and the high horizontal frequency component is 2f1 in the principal component. is transmitted at the edge, eg above f, e. Edge signal up to 2f, c Multiplexed and preferably frequency inverted. The main signal is 3/4 of the number of lines and the edge (voter) signal comprises the remaining one-quarter. results and 2f, which occurs. The signal is sent to a phase-separated Westlone PAI-encoder for transmission. Can be added to the luminance input.

This system uses continuous spatial and temporal passbands for luminance and chrominance. give a range. This is achieved by a phase-to-phase separation, seven-tone clean PAL encoding system. However, this encoding system (to accommodate J and V carriers) Avoiding the need to cut a hole in the spatial and temporal brightness spectrum. Restriction of the spectral folding technique to auxiliary information aids in good compatibility.

Luminance passband is 6.5 KHz horizontally (relative to 1,5625 KHz scan) and perpendicular to the camera and film modes, respectively. The effective image height can reach 160 or 216 cycles. explained Usable aliases achieved for PAL sides B and C by the above example resolution is 6.3MHz and 150 or 190c/ph for brightness and 0.5MHz and 60c/ph for each of the chrominance components It is determined that The shape of the luminance passband is calculated by calculating the distance from the origin to the band edge J. The perceived spatial resolution is maximized by making it more isotropic for all spatial frequencies. Hexagonal to optimize.

Gaiho on the same plate The invention will now be described in more detail by way of example and with reference to the accompanying drawings, which In the drawings of FIG. 1 is a block diagram of an encoder according to the invention as used on the transmitter side. and FIG. 2 is a diagram showing the vertical-horizontal response of the luminance channel in the encoder of FIG. , Figure 3 shows the clean PAL assembler on the encoder (transmitter) side and the decoder (receiver) side. It is a configuration diagram of a sub-lifter, FIG. 4 is a block diagram of a decoder in a receiver embodying this invention, and FIG. It is a time interval indicating writing to and future reading from the storage device in the device, Figure 6 shows a PAL splitter for combinations Y and U+V as in the receiver. This is a detailed configuration diagram. Figure 7 shows the subband filter in the receiver of Figure 4, and Figure 8 shows the subband filter in the receiver of Figure 4. 4/3x2f, , to 4/3x4Lc post-filter is shown. Figure 9 illustrates the f & position filter open low sample relationship at the receiver at two phase positions. and Figure 10 shows the structure of a 4,/3X4f, -720 sample/line down converter at the receiver. It is a complete drawing, Figure 11 shows the receiver for converting from 432/2・1 to 432/], :1. It is a block diagram of an interlaced one-sequential converter, Figure 12 is a block diagram of the line rate up converter in the receiver, and Figure 13 is a diagram of the line rate up converter in the receiver. 14 is a block diagram of the [J soil V signboard rate converter in the receiver, and FIG. 15 is a block diagram of the vertical temporal post-filter at the receiver. 16 is a block diagram of a U/V display/to-up converter, and FIG. Vertical temporal for cascaded sequential-interlaced sequential conversion in transmitters and receivers FIG. 17 is a plot illustrating the intermittent frequency response. Figure 18 is a plot showing the frequency response of the stopper. This is a plot showing the response, and Figure 19 is a plot showing the luminance? & equipped filter vertical-horizontal response Figure 20 shows the luminance skipping one-sequential converter response in the receiver. The plot shown in Figure 21 is the luminance line rate up converter at the receiver. A plot showing the response, Figure 22 is a plot showing the chrominance prefilter response, and Figure 23 is a plot showing the luminance prefilter response. 211 is a plot showing the line rate up converter response, and Figure 211 is a plot showing the line rate up converter response. 3 illustrates a letterbox display on a 3 aspect ratio screen.

Surikajin Kakua@Kintoguri explanation FIG. 1 shows a block diagram of an encoder 10 embodying the invention. This encoder is have inputs 12.1.4.16.18 each receiving a signal of the image standard. Ru.

(a) 1250150/2:1 (b) 625150/2:1 (c) 625/25/1:1 (d) 625150/1:1 Format (a) has 1152 active lines per image, and formats (b) and (c) and (d) has 576 lines per image. Format (c) is usually 25 frames. obtained from a film that moves at a rate of 100 mm per second. Only one input is valid at any time . Inputs 12.14 and 16 have inputs of the form (a>, (b) and (c), respectively) Connected to converters 20, 22 and 24 for converting to a common standard (d). this Standard (d) is progressive scanning, ie non-interlaced. This format is vertical Good vertical resolution can be maintained without temporal aliasing . 1250150/2: Conversion from 1 has 1152 entities in line sequential grid. In the converter 20 using a 1511iX5 field filter for the effective line achieved. This filter takes into account the camera, display device, and visual characteristics to It is designed to produce progressively scanned images of the highest possible quality. 625150/2: Conversion from 1 uses an 8-line x 5-field filter for 576 sequential grids. 625/25/1: Conversion from 1 is done by simple field insertion. and is used only in film mode.

A selected one of the outputs from the transducers 20, 22 and 24 is connected to the vertical low voltage along with the input 18. It is processed by a bandpass filter 26. This is vertical downfiltering and format (d) gives the line rate conversion from 432150/1:1, where 43 2 represents the number of effective lines in the image. This is an upsampling by a factor of 3 ring (increasing the sampling rate), then a 41-tap vertical filter, and finally a 4-tap vertical filter. This is done by downsampling (reducing the sampling rate) by a factor of . centre The output of filter 26 is effectively 432150/1:1 as shown above. . The luminance output of the filter 26 is sequentially reconverted into an interlaced form by an interlaced converter 28. is converted to produce a 432150/2:1 signal.

Converter 28 also uses vertical and temporal filters to optimally pack the interlaced transmission spectrum. Apply the router function. In this way static vertical resolution comes at the expense of dynamic vertical resolution. be enhanced. This is a well-known technique that causes images to appear in letterboxed format. When it helps to achieve acceptable vertical resolution.

All of these operations are performed at the input horizontal sampling density (i.e., sampling rate) be done. The input horizontal sampling density is of the form (1440 samples per image width for a> and 720 samples per image width for formats (b), (c) and (d) It is. The signal may be analog or digital at this point, with a corresponding Theoretical vertical and temporal prefixes and postsummers for encoders when combined with encoders The filter product is shown in FIG.

Returning to FIG. 1, the output of converter 28 is then applied to horizontal frequency converter 30, which The transducer horizontally increases the sampling density, f,. If is the color subcarrier frequency If so, make it 4/3X4f,c, or 16/3f,e. specimen The oscillation frequency increases only in the horizontal direction. The output of the horizontal frequency converter is then - added to vertical prefilter 32; The prefilter 32 is a diagonal filter. The band edge is shown in Figure 2 labeled "compressed 2Le" sampling frequency. It is perpendicular to the line connecting the origin and r4/3X2f, , J, as shown in FIG. these eyes For the purpose of such a filter, every other tap has a weighting of zero, so Can be configured with minimal hardware cost. The output of filter 32 is is added to the 2=1 down sampler 34, which preserves the manual and half image offsets. The filter-passed signal is subsampled at a sampling rate of 4/'3×2f,e while Ru. This introduces aliasing at the upper band edge. Aliasing is Spectral folding causes high-frequency signal components to be reflected and creates interference within the signal bandwidth. Occurs when it causes a minute. This aliasing is now applied to the edge lines. , and by appropriate post-filtering in the decoder the improved form 16: 9 display device, so it does not impair compatibility with existing receivers. .

The 4/3X2f,e data stream is then split into f,c or The first stream of 2f,e retains the lower signal frequency at , and the higher signal frequency 2/3Xf, which holds the number. It is divided into a second stream and a second stream. The division is purely horizontal. i.e. the subband filter is purely horizontal and vertical as shown in Figure 2. This will result in a purely horizontal spectral split as shown by the straight line. mark The first data stream containing 3/4 of the book is the middle 4 of the image of the signal as transmitted. The third line occupies 3/4, or the middle 432 lines, and also contains 1/4 of the sample. The two data streams are located at the upper and lower edges of the image as transmitted, i.e. It occupies the remaining 144 lines divided into the upper and lower parts. The second data stream is multiplayer 2f1. gives the sampling rate of This multiplexing This can be accomplished in a sample-by-sample manner, or simply by sequentially passing the samples one out of the three lines. transmission line. obtained in this way The 2f,e signals are frequency inverted around fme to reduce visibility. clean PAL assembler 42 by selector 40 at an appropriate point in the scan. supplied to

The amplitude of this edge signal is scanned to ensure that it is not visible on normal display devices. reduced by Goering to give an average level corresponding to black.

The clean PAL assembler 42 is a so-called phase-separated Uniston clean PAL code. It is a coding system. The basic Uniston system is described in British Patent Specification GB-A-2 044577 and GB-A-2113037. The operation is [both Possible Improved PAL System J EBV Review, February 1986 (“’A Compatible Improved PAL System'', EBVR 1986). Further This development is based on international patent application number WO-A-92/100 published on June 11, 1992. It is described in 68.

The chrominance output of filter 26 is applied to vertical and temporal prefilter 44. However, this filter uses U and V carriers to expand the static vertical chrominance resolution. The waves have different temporal frequencies (6, 25 and 3/4) of the image frequency. 18.75Hz). That's about 6 Enables horizontal chrominance resolution of 0c/apw (cycles/effective image width) Although it is valuable for PAL systems■, it allows approximately 30 c/a pw It is not recognized as legitimate for PAL systems B and C, which only provide do not have.

The filtered chrominance signals U and V are then converted to a sampling rate of 4f,e. In addition to the sampling rate converters 46 and 48, respectively, to simplify the rest of the processing. and then filters 50.52 provide horizontal low-pass filters at f, c/2. correct chrominance when decoded by normal PAL, decoder The phase is corrected in a phase corrector 54,56 to ensure the correct position. U and The signals are then combined according to the switching signals received at input 60. forming U+■ and U-V in alternating lines combined in strainer 58 .

The phase-separated Uniston clean PAL encoding system uses a luminance sampled at 2f, e. degree component and f,. For the chrominance component sampled at channel. This Uniston system also applies to the luminance component and the link For the chrominance component up to the horizontal bandwidth (sample density) determined by the bandwidth It is transparent. In principle, the shape of these channels in frequency space is 6 conditions that may be changed at will, provided that the large sampling rate is not exceeded. Traditionally, the requirement to give good compatible image quality is to has prohibited spectral folding as required to give arbitrary .

This constraint is shown in this system by the vertical trajectory at l712 This was overcome by splitting the spectrum vertically and horizontally.

Figure 2 shows a plot in vertical/water frequency space for the luminance signal. and the Y-axis shows the cycle 7・' effective image height (c/a p h or simply c/p h) and the X-axis is the cycle/effective image width (C/a p The horizontal frequency at w) is shown.

The operation is therefore such that the signal spectrum is divided into subband codes in the ratio 3/4:1/4. The color subcarrier frequency f1. be divided along the horizontal frequency axis around It is. The lower three quarters are f3. from the origin. It extends to 2f, e sampling is transmitted in the letterbox area of the signal at a rate. This signal is spectral It does not include aliasing, so the artifact level is simply the subband filter. Determined by data design. The upper quarter extends from f, e to 4//3LC ing. This area of the spectrum has a frequency of 337 c/ap, equivalent to 6.5 MHz. Stretch the limit on its spectrum for a purely horizontal frequency (zero vertical frequency) of w. has spectral folding applied to This part of the signal spectrum Specimens describing spectral folding have little effect on compatibility. If not, they are interleaved and frequency inverted before transmission on the edge line.

A typical 423 tank aspect ratio receiver uses the method shown in Figure 24 to detect the letterbox portion of the signal. and ignore edge areas containing low-level signals that are translated as black. . Compatible as the signal can therefore be displayed in existing receivers It is. An improved 16:9 aspect ratio receiver fills the screen with the displayed image; The image takes the high detail information provided at the edges and applies this to the letterbox portion of the signal. generated by adding to the included low (and medium) detail information. Such reception The signal will be described with respect to FIG. 4, which shows a decoder for it.

First, the clean PAL assembler 42 and the corresponding splitter 8 in the receiver. 0 is shown in FIG. The clean PAL assembler is as illustrated in Figure 1. describes the lower three-quarters of the support spectrum in Figure 2 during the l/tarbox period. The luminance samples of 2f, e are also spectrally folded during the upper and lower edges (data). and multiplexed upper quarter. The edge is the frame where the specimen will be sent. Contains those specimens that belong only to the field. The chrominance signal is Chromina is present only in 216 lines per card (432 lines per image). This unused capacity channel is not used during the edge (data) line. The device may be used for additional auxiliary information or to convey auxiliary data. will be possible.

The luminance signal Y is received at terminal 82 and has already been sampled 2fIc''c. Direct and one-line delay to transversal filter 86 with filter function F1 84 and the output of filter 86 is the luminance component of the encoded PAL signal. form a minute. The chrominance signal is received at terminal 88 and is already connected to modulator 62 (FIG. ) is modulated with fme by is supplied directly to the Lanceversal filter 92 and via the one-line delay device 92, and The output of filter 92 then forms a chrominance component. of these two filters The outputs are summed in combinational circuit 94 and transmitted by transmit link ]00. form a PAL signal.

The encoder processing is assumed to be done with 4f, , in which case the modulation is simply This results in zero insertion and multiplication by 77' or ±1. However, this is never the essence The encoder and decoder can be used with any sufficiently high line-locked sampling could be carried out at a rate. The choice of subcarrier-related sampling is simply Of equipment! l noise and color burst-locked targets at the receiver. The derivation of the standardized clock is, if anything, similar to the line-locked clock. It is stable and stable.

As shown in FIG. 3, the splitter 80 and the receiver similarly each have a two transbars powered directly from the network 100 and through a one-line delay 106; It incorporates monkey filters 102 and 104. Filters 102 and 104 are They each have filter functions F3 and F4, and the filter 102 or and a filter 1゜4 consisting of U+■ and U-■ in alternating lines. gives the chrominance signal from

Numerical numbers Fl, F2. F3 and F4 describe Uniston PAL system The decision can be made based on the above-mentioned documents. Uniston PAL system is a clean PAL, and the assembler 42 creates a subband consisting of two synthesis filters. The splitter 80 operates as a sub-band decoder consisting of two analysis filters. Since it operates as a band encoder, it is closely related to the principles of subband analysis and synthesis. There is.

The improved receiver uses a decoder 120 of the type shown in FIG. 6. Referring to FIG. 4, the time diagram of the incoming PAL signal is shown as 4f , c and digitized (or at least sampled) to PAL input 122. is received, and a transponder is used to equalize the frequency response of the transmit channel 100. It is passed through a universal filter 124. This transversal filter has 27 It is desirable to have a symmetrical and usually folded cumulative ladder-shaped structure are using. Its coefficients may be, for example, as specified in Table 1 below. , and its characteristics may be as shown in FIGS. 17 and 18. Figure 17 is a cycle Charts plotted against horizontal frequency at effective image width (c/apw>) Figure 18 illustrates the frequency spectrum of the The overall response of the channel is illustrated.

Furthermore, 4f,,PAL is in the specified phase, i.e. 45 degrees with respect to the U subcarrier. sampled in

The output of channel equalizer 124 configures PAL splitter 80 as shown in FIG. is added to the two transversal filters that form the filter. One filter is a luminance standard. One produces book Y and the other produces chrominance samples U±■.

These filters typically contain 77 terms of two lines in a 4f, e grid, respectively. and is symmetrical, which divides the number of required coefficients into three equal parts. Shining The degree output is <2f, c4f where every other sample in the quincunx grid is zero, e It can be regarded as a signal. Since zero samples are not actually generated, the brightness scale The output rate from the splitter is actually 2f,c, which is the rate required at any one time. gives a further third division of the number of coefficients. Similarly, the chrominance output is It can be considered as a sample of 4f,e, three of which are zero. Hence The chrominance signal requires a sampling rate of f, which is required at any one point in time. gives a further 2-fold reduction in the number of coefficients taken.

A more detailed configuration diagram of the clean PAL splitter is shown in Figure 6. The following figures 7, 8 and 9 to 15 show sample delays, line delays, switches, multiplication Primarily, using the usual specifications of circuit components in the form of adders and adders. These drawings show the structure of an implemented circuit based on universal filters. The purpose of this is to provide the reader with a textual explanation of what is presented in a precise and concise manner. A detailed textual description of these drawings, which will be completely redundant to the skilled reader of the specification, is It is therefore unnecessary and therefore omitted.

The edge samples are frequency inverted by multiplying by fme in multiplier 126. Any non-linear additions in the encoder are gender is canceled. This non-linearity is as specified in Table 7 below. I ! is then 3-1 linear multiplexing performed in the encoder assuming sample-by-sample multiplexing. Recirculate the signal by selector switch 134 between three wires to restore It is added through storage 130 to be stored in circular line storage 132 .

The letterbox sample is sent to another selector through a compensating retiming store 136. switch 138, which letterboxes at appropriate portions of the scan. select samples (bottoms and edges) and apply them to the horizontal subband filter 140. In addition, this filter performs subband decoding and the samples are 3 letters per box. Requires to be multiplexed at a ratio of specimens to one edge. 2f, c letter box The specimen is removed from the edge line by simple switching by switch 138. Multiplexed with f and c samples.

Storage device 130 is a 72 line x 473 sample dual boat field storage device, Provides vertical retiming for edge lines. Storage device 136 has 144 lines x 473 lines Specimen storage and retying to the letterbox line as shown in Figure 5. Give ming. The resulting signal contains 216 lines per field. However, this is spread over the entire vertical scan period. This is the subsequent line rate Avoids the need for large amounts of storage in upconverter 148, but This complicates the description of the sampling frequency used. used in the decoder description. The term f- when For example, 283.7516 samples, ・′ line) should be considered as a spatial description of It is.

Letterbox combiner 140 is followed in series by a postfilter 142 and a sampling rate. converter 144, interlace-sequential converter 146, and line rate up converter 14. 8 is placed.

The chrominance signal U±V from PAL splitter 80 is the period of the letterbox line. 32 microsecond lines over the entire vertical scan period. Must be displayed in cycles. 144 lines x 235 samples double boat field Storage 150 stores 64 microseconds in a manner similar to that used for luminance. Three output lines spaced apart by Used to spread chrominance lines across. The only difference is the line period shortening. position determination, which in chrominance is acceptable in the post-decoder. chrominance retiming storage without exceeding the maximum possible sampling rate. It can be implemented as follows. When interpreting the meaning of subcarrier frequency, consider the luminance signal. The same precautions as just noted in -F are required for this issue.

The output of storage device 150 is then passed to sampling rate converter 152, U/V demultiplexer 152, and U/V demultiplexer 152. 154, vertical temporal <vT> postfilter 156, and linear rate up is applied through converter 158 to provide the U and ■ output signals.

The individual component circuits will now be described in more detail.

Multiplexed letterboxes and edges describing the upper and lower subbands of Figure 2 ( The samples are recombined in a horizontal subband filter 140, but A block diagram of this is shown in FIG. Its structure is self-evident from this figure. Fi The sampling density of the router input and output is 4/3X2f,e, which is a manual offset (i.e., a quincunx sampling structure). This filter is letterboxed. The function of a low-pass subband synthesis filter that operates on edge samples and the function of a low-pass subband synthesis filter that operates on edge samples This simultaneously realizes the functions of a high-pass synthesis filter that operates in the same manner. the resulting signal are added together. The encoder subband filter and decoder subband filter are Reciprocal and transparent when cascaded.

The output data from the filter 140 has a sampling density of 4/3 x 2f, e. . This is just adequate to support the signal spectrum but the upper band Due to aliasing due to repeated spectra at the edge, the It has a quincunx structure that cannot be used directly. The post-filter 142 is This structure is orthogonal, increasing the sampling density to 4/3 x 4 f−c. Therefore, the artifacts due to out-of-band spectral repetition are over-sampled. Causes rias. It has a spatial frequency response as shown in Figure 19, Its coefficient versus sample relationship is shown in Figure 9, and its specifications are given in Table 2 below. It is.

The postfilter output is horizontally oversampled and related to fme Therefore, it is simply a pseudo line entrance. A truly line-locked sampling structure is displayed on the display. Required by the consumer electronics industry for use in Therefore, horizontal Sampling rate conversion is 4 taps 256th to line locked 720 samples/active line This is done by phase sampling rate converter 144. The configuration diagram of the sampling rate converter is It is shown in FIG.

The signal at this point has a near minimal sampling density both horizontally and vertically. Ru. However, vertical post-filtering depends on the display spot profile. is performed normally. This operation is not perfect, and for good quality display, , to higher line rates to enable electronic vertical t11 filtering. Requires conversion. In this way the response of the electronic post-filter is dominant and Signal frequencies that cause static or dynamic vertical aliasing are attenuated. Luminance display rate up conversion is optional, but if included, two parts It is done in minutes. The first one is 432/221 to 4 in converter 146. 32/1: It is an interlaced-sequential conversion to 1, and is applied to the encoder's sequential-interlaced converter. 108 cycles/vertical frequency above the image height placed on a 25 Hz carrier Use cross-field filtering to return The second part is far from complete. Static remaining after post-filtering with invisible spot profile 432 to 576 effective in transducer 148 to reduce vertical aliasing It is a line rate up conversion to line. Interlaced sequential converter and line rate up The block diagrams of the converters are shown in Figures 11 and 12, respectively, and their frequencies and The coefficient patterns are given in Figure 20 and Table 3 and Figure 21 and Table 4, respectively. . Figure 16 shows cascade sequential-interlaced sequential in a combined transmitter and receiver. The vertical and temporal frequency responses for the transformation are illustrated.

The 216 line/field period at the input to the interlaced sequential converter 146 is 288 appears at times occupied by regular television lines. The line period up to this point is 21. maintained at 64 microseconds with a 1/3 microsecond gap or 851 /3 microseconds. This is an optional choice and the converter From the first operation performed by Because the line period is shortened to microseconds, the hardware complexity, delay capacity There is no difference between This shortening is achieved by a pair of double boat line storage devices. one of these is written while the other is read. All interpolators is valid between three lines where data movement is stopped and one inactive line following it. Pay attention to one thing.

The interlaced sequential transform has vertical and temporal artifacts, e.g. This is done by a symmetrical vertical and temporal filter to minimize the "tuning".

The interpolator has two signal paths: a direct path to the input line and a coincident output line; Contains interpolation paths for the remaining output lines. The interpolation filter is predominantly lower vertical Four fixed taps in the current field, giving the frequency, and the sequence of the encoder - giving a predominantly high vertical frequency placed on a 25 Hz carrier by an interlaced transducer; There are five taps in the field before and after the game. Direct line and interpolation The lines are fed into a display line rate up converter in parallel at 720 samples/line each. . Line rate up converter 148 performs upsampling by a factor of 3. , apply a 15-tap vertical filter and downsample by a factor of 4. Work by doing. Therefore, the filter has four phases; three of them have four coefficients and the fourth one has three coefficients. Ru. The input to the line rate up converter has a correct line period of 32 microseconds. A continuous sequence that is at the correct frequency but is missing every fourth line. It is a line. The lines required to contribute to the filter output are recirculating double boats the latest input line held in the line store, as it appears, is sent to the first delay device; The contents of this are then passed to the next delay device (and so on). The switch order is Then accept new data for three lines and then recirculate for one line. It is something to do.

The chrominance signal at the output of chrominance retiming storage 150 is f , U+V at the sampling rate of e. This is equivalent to the U component centered at the origin. and centered at 108c/aph and 12.5Hz (for 16:9 aspect ratio). ■Can be considered as an ingredient. The U component is therefore a suitable vertical and temporal Can be separated by post-filtering. Similarly, the ■axis switch When used, the V component is at the origin and the U component is 1.08 c / a p h and 12 ．． It is placed at 5 Hz, allowing separation of the components.

The separation of the U and V components is therefore a purely vertical-temporal operation. horizontal chrominan The signal width is approximately 0.6MHz for PAL systems B and G, and approximately 0.6MHz for system ■. The horizontal sampling density is approximately 1.1 MHz, and the horizontal sampling density is 180 samples/effective line. That is, it can be fully supported by the CCIRREC601 sampling density.

Therefore, the sampling rate conversion to 180 samples/' line is the output of the retiming store. and direct and switched chrominance signals is U±V, ■±U, U±V, ■-! :U... 360 samples after multiplexing form a combined signal with a sampling density of /line. This is a vertical/temporal filtering color separation and display line rate up conversion in the same hardware module Therefore, it is possible to do this. The U±■ sampling rate converter is shown in Figure 13. ing.

The vertical temporal postfilters for U and V separation are the same; These filters can be realized by the same hardware components and This is advantageous because it is asymmetrical. The lower vertical frequency components are from the previous field. , that is, the high frequency components extracted after field delay are taken from. The filter aperture is slightly larger than for luminance, and the latter It has 7 terms in the field and 6 terms in the previous field. The storage control logic One of the four potentialities of the timed input signal must be considered , a filter configuration diagram is shown in FIG. The combined pre- and post-filter response is The post-filter specifications are shown in FIG. 22 and in Table 5.

The output of the chrominance postfilter is alternating U and ■ samples, which Each is at 180 samples/effective line, spread over the output effective scan period. It is held on 216 lines, with one line out of every four empty.

Chrominance display rate amplifier converter 158 is 108 cycles/image height - is not required to support a vertical resolution of F and is therefore a purely vertical interpolation process. A block diagram of the display rate up converter 158 is shown in FIG. The responses and specifications are given in FIG. 23 and Table 6. It is used for brightness ), except that its input is u, v, u, v... It is heavily mapped onto an interlaced sampling grid. The interpolator is a knob sample with a ratio of 4. ring, use a 39 tap filter, and then down by a factor of 3 Do sampling. Therefore eight phases each with five coefficients There is.

The line storage device is not used except when a new 32 microsecond input line is available. It's recirculating.

The phase of the interpolator changes once per output line period, but for both U and ■ samples (for two It's the same. U and ■ samples give separate data streams at 180 samples/effective line. can be separated by a simple switch to ensure The output is therefore 2 It is a 4.1:l format with a brightness of 7MHz.

As you may have noticed, the various filters shown separately for each country are the same. may actually be combined or otherwise arranged to give an overall power/output function. can be Some or all of the quantities may be software It can be realized by

Therefore, to summarize, this system has an overall input of 4/3X4f, e of 43 Start by reducing to 2 wires. The signal is then preformed using the overall function in Figure 2. 4/3X2f, filtered and downsampled by 2. sampling of Give a rate.

According to this invention, the signal is divided into two parts by two filters, and the As a result, the main low horizontal frequency components are transmitted in the letterbox or main signal and High horizontal frequency components are transmitted in the edge (data) region. This is the so-called sa This is a band division. That is, as long as a filter is complementary, it is incomplete. It doesn't matter what happens. The edge signals are then reordered to one-third the number of lines.

It is highly desirable to frequency invert the signal to take advantage of the companding frequency.

The letterbox and edge lines are then recombined. Each field in turn edges 361! , followed by 216 of the main or letterbox signal, followed by another Contains 36Il at the edge of . This is 288 at a sampling rate of 2f,e Gives the line/field total. This sampling rate is determined by the phase separation or Uniston P It is suitable to be applied to the luminance input of an AL encoder.

Although this explanation has been given in relation to the highly preferred 3:1 split, other ratios are also considered. It will be. Similarly, it is possible to deviate slightly from 2f and e as the average sampling rate. performance, but some loss of quality may occur.

Table −↓ Channel isolator coefficient list in units of 11512 Filter coefficient at z y x 0 0 13 4゜ table space Luminance post filter coefficient list Clothes-” Luminance line rate up converter coefficient list 11512 unit Filter coefficient at z y x once the table Chrominance vertical/temporal post filter coefficient list z y x quality grade Once dressed Chrominance line rate up converter coefficient list z y x Akira 2 Tadada Nira person-e Letterbox decompression function Decoder compander search table The characteristics are symmetrical with respect to the black level, with each half consisting of four linear sections.

The maximum value of the black bar is 220 and 41 as a scale when the gray level is 700 mV. It's a gray level.

Therefore, the maximum edge amplitude is (41/220) x 700 = 130 mV. -500711 (9) vtEeiL16: 9L/1rt, pb'L'L'f )’? ”r r″S)D ,　■ 〇-1 ,9 Q ss marrow @1liil III II II II Field Kuchi 1 Kuchi Ward) Ward b - mouth toro X〒〒〒　1 Field Q Yu (1, Id/○) Takaso 1st stage silkworm 0 (Ltd/○)'#l Inquiry Natsutare 7:/(3-:L-0 Temporal frequency (Hz) To F/C;r　22 Temporal frequency (Hz) 'l/'V/'1 ,,,,Continuation of PCT/G8 92101988 front page (81) Designated countries EP (AT, BE, CH, DE.

DK, ES, FR, GB, GR, IE, IT, LU, MC, NL, SE), 0A (BF, BJ, CF, CG, CI, CM, GA, GN, ML, MR, SN, TD, TG), AT, AU, BB, BG, BR, CA, CH, C3゜DE, DK, ES, FI, GB, HU, JP, KP, KR, LK, LU, MG, MN, MW, NL, No, PL, RO, RU, SD , SE, US (72) Inventor Eastern Pruck, James Nidwar de England, S.M.T.2T (72) Inventor: Thomas, Graham Alexander, East Sussex, England, TN 21.8 Earls E, Heathfield, Broad Oak, Street End Lane Groundless), Half Moon Cottage

Claims (25)

[Claims] 1. the luminance signal, the first part of which consists of a signal component having a low horizontal frequency, and said first and second portions, said second portion representing a signal component having a high horizontal frequency; dividing the first portion into a first selected line of the transmitted signal and dividing the first portion into transmitting a luminance signal, the method comprising: transmitting two portions by a second selection line of the signal; How to do it. 2. the first portion is comprised of signal components below a predetermined frequency; 2. The portion according to claim 1, wherein the portion consists of frequency components above this predetermined frequency. Method. 3. 3. The method of claim 2, wherein the predetermined frequency is fsc. 4. The first portion includes approximately three times as many frequency components as the second portion. , the method of claim 1. 5. Claim 1, wherein the transmitted signal is applied to one input of a phase-spaced encoder. Method described. 6. The transmitted signal is a PAL type signal, and the phase-gap encoder is a Weston PAL type signal. 6. The method of claim 5, wherein the method is an A-code encoder. 7. A chrominance signal is applied to the other input of the Weston PA encoder. 7. The method according to claim 6. 8. before the second portion is combined with the first portion to form the transmitted signal; 2. The method of claim 1, wherein the method is temporally compressed. 9. before the second portion is combined with the first portion to form the transmitted signal; 2. The method of claim 1, wherein the frequency is inverted. 10. The first selection line is a vertical scanning center line, and the second selection line is a vertical scanning center line. 2. The method of claim 1, wherein the upper and lower lines of . 11. Generating a sampled signal with a sampling frequency of 4/3×2fsc, this step of the signal, the first and second portions containing different frequency components of the signal. the first part has a sampling frequency of 2 fsc and the second part has a sampling frequency of 2/3 fsc; dividing into the first and second parts having a main frequency of 2 fsc; compressing the second portion into a number, and alternating the first and second portions in time; A method of transmitting a luminance signal, the method comprising transmitting a luminance signal to a luminance signal. 12. The sampled signal having a sampling frequency of 4/3×2fsc is 4/3×4fsc. Generating a sc signal and downsampling to 4/3 x 2fsc 12. The method of claim 11, wherein the method is generated by: 13. the first portion includes a signal component below fsc, and the second portion includes a signal component below fsc; 12. The method of claim 11, comprising frequency components above fsc. 14. Claim where the transmitted 2fsc signal is applied to one input of the phase-spaced encoder. The method according to item 11. 15. the signal is a PAL type signal and the phase-separated encoder is a Weston PA 15. The method of claim 14, wherein the method is an encoder. 16. A chrominance signal with a sampling rate fsc is input to the Weston PAL encoder. 16. The method of claim 15, wherein the input is applied to the other input. 17. The second portion is combined with the first portion to form the transmitted signal. 12. The method of claim 11, wherein the method is temporally compressed beforehand. 18. The second portion is combined with the first portion to form the transmitted signal. 12. The method of claim 11, wherein the method is previously frequency inverted. 19. the first portion is transmitted at the center line of a vertical scan and the second portion is transmitted vertically. 12. The method of claim 11, transmitted in direct scan top and bottom lines. 20. A first part of the luminance signal is composed of a signal component having a low horizontal frequency, and said first and second parts, said second part consisting of signal components having a high horizontal frequency; a dividing means for dividing into minutes, and a dividing means for dividing the first part into a first selection line of the transmission signal; and means for transmitting said second portion as a second selection line of a signal. , a device for transmitting luminance signals. 21. Means for generating a sampling signal having a sampling frequency of 4/3×2fsc, standard The first and second parts contain different frequency components of this signal. and the first part has a sampling frequency of 2 fsc and the second part has a sampling frequency of 2/3 fs. means for dividing into said first and second parts having a sampling frequency of c; 2; means for compressing said second portion to a frequency of fsc; for transmitting a luminance signal, including means for transmitting portions alternating in time; equipment. 22. The lines of the transmitted luminance signal are arranged such that the first line forms the first part and the second line forms the second part. wherein the first portion comprises a signal component having a low horizontal frequency and the first portion comprises a signal component having a low horizontal frequency; separating such that the second part represents signal components with high horizontal frequencies, the former retiming the first and second portions relative to each other; and recombining the parts so that a low horizontal frequency component is formed from said first part; and forming an output signal in which a high horizontal frequency component is formed from the second portion. and a method for receiving a transmitted luminance signal. 23. The arriving samples are temporally divided so that the group of arriving samples forms the first part and the arrival target is such that the group of books forms a second part, said second part having a circumference of 2/3 fsc; stretching to a wave number and recombining the first and second portions to The first and second parts constitute different frequency components of the luminance signal, and the output signal is 4/4. forming an output signal having a sampling rate of 3×2 fsc. , 2 fsc. 24. The lines of the transmitted brightness signal are arranged such that the first line forms the first part and the second line forms the second part. the first portion comprises a signal component having a low horizontal frequency, and to separate such that the second part represents signal components with high horizontal frequencies; means for relatively retiming said first and second portions; recombining the first and second parts so that the low horizontal frequency components are from the first part. an output signal having a high horizontal frequency component formed from said second portion; Apparatus for receiving a transmitted luminance signal, the apparatus comprising means for forming a luminance signal. 25. The arriving samples are temporally divided so that the group of arriving samples forms the first part and the arrival target is means for causing the group of books to form a second part, said second part at 2/3 fs; c. and means for recombining said first and second parts. In addition, the first and second parts constitute different frequency components of the luminance signal and the output To form an output signal where the force signal has a sampling rate of 4/3 x 2 fsc receiving a transmitted luminance signal having a sampling rate of 2 fsc, equipment for.