FI86785B - Wide picture television system with enhanced definition using a number of signal transmission channels - Google Patents

Abstract

A system is described for sending and receiving a television picture with high definition (HDTV). The signals are sent on two ordinary 6 MHz NTSC channels. A first signal, the main signal, is generated from the HDTV- source's signals by an encoding circuit 112 in order to be compatible with existing NTSC receivers 122 so that in these receivers it generates a picture which is not distorted to any appreciable degree. The main signal contains psycho-physically concealed video information which can be used in a receiver 128 for enhanced definition for generating a wide-screen picture with higher resolution than a normal television picture. A second signal, the auxiliary signal AUX, is generated in the transmitter by decoding the main signal 140 and subtracting the decoded main signal from the original HDTV source's signals. The auxiliary signal, which contains frequency components within the range 0 Hz - 20 MHz, is divided into three bands, A (0 Hz - 6 MHz), B (6 MHz - 12 MHz) and C (12 MHz - 18 MHz). The bands B and C are frequency converted for the use of the band 0 Hz - 6 MHz and are time-division multiplexed according to the line-by-line principle. The combined B and C bands are subsequently time-division multiplexed with the A band according to the image-field by image-field principle with regard to the stationary pictures. In the case of moving pictures, only the A band is transmitted. The receiver 136 decodes the main signal and by use of the movement signal sent with the main signal it decodes an auxiliary signal. The decoded main and auxiliary signals are combined to reproduce an HDTV picture. <IMAGE>

Description

1 86785

Extended definition widescreen television system with multiple signal transmission channels

BACKGROUND OF THE INVENTION The present invention relates to a television signal transmission system using more than one transmission channel for transmitting signals representing high definition widescreen.

A conventional television receiver, such as a receiver according to NTSC transmission standards, has a 4: 3 aspect ratio (the ratio of the width of the displayed image to the height). Recently, there has been interest in using higher aspect ratios in television receiver systems, such as 2: 1, 16: 9, or 5: 3, because such higher aspect ratios more accurately approximate the aspect ratio of the human eye than the 4: 3 aspect ratio. Videos with 5: 3 and 16: 9 aspect ratios have received special attention because these ratios approximate the aspect ratio of many movie movies. In addition, there has been interest in increasing the level of detail of the reproduced 20 images beyond the level of a conventional television image in order to more accurately approximate the level of detail in a motion picture. Television systems that transmit signals representing high-definition widescreen should be designed to transmit a video signal that is compatible with existing television receivers because widespread adoption of an incompatible system can be difficult.

A review of several proposed high-definition widescreen television systems can be found in an article by Robert Hop-30 kins entitled "Advanced Television Systems," IEEE Transactions on Consumer Electronics, February 1988, p. 1 to 15. Of the eight systems described herein, five are compatible with existing NTSC television receivers to the extent that the transmitted 35 signals can be received by a conventional NTSC receiver without pre-processing with a converter, to produce a 4: 3 picture with little weaker relative to a standard NTSC image. These systems were developed by: AT&T Bell Laboratories (Bell system), Dr. William Glenn of the 5 New York Institute of Technology (Glenn system), Del Ray Group (Del Ray system), North American Philips (NAP system) and NBC and David Sarnoff Research Center consortium (NBC system).

The Bell system uses two 6 MHz televi-10 channels to transmit signals representing a high-resolution widescreen image. One channel carries a constant NTSC signal representing the relatively low-frequency luminance and chrominance signal components of a high-resolution widescreen image. The second channel carries 15 high frequency luminance and chrominance signals that can be used by a special receiver to complement the image represented by the NTSC signal.

The Glenn system also uses two channels, one of which carries an NTSC-compatible signal. The second 20 channels in the Glenn system use only 3 MHz of the 6 MHz bandwidth, and carry image detail information that has been filtered to reduce its temporal resolution (i.e., to reduce its frame rate). Frame memory is used to combine an NTSC signal with a low ke-25 high rate high precision signal. The widescreen effect is achieved in the Glenn system by increasing the vertical blackout period while decreasing the horizontal blackout period.

The Del Ray system uses only one 6 MHz 30 channel that carries an NTSC-compatible signal. This signal defines a high-precision image for six field periods. The high-resolution image is reconstructed with a special receiver that combines six fields in the frame memory. The Del Ray system also increases the vertical 3,867,885 dimming period to achieve a widescreen effect.

The NAP system transmits a standard NTSC signal on one channel and a supplemental signal on the other channel.

5 The complement signal is combined with the NTSC-compatible signal on a special receiver to form a widescreen, progressively scanned image. The horizontal resolution of the image produced by the NAP system is essentially the same as that of the NTSC image.

10 A digital approach to transmitting enhanced quality video using light wave technology in a single digital channel on a "Broadband Integrated Services Digital Network" (BISDN) is described by J. A. Belli-et al. in an article entitled "Television coding for 15 Broadband ISDN" published at the IEEE Global Telecommunications Conference; Globecom '86, Houston, Texas 1-4 December 1986, Conference Record, Volume 2 of 3, IEEE (US), p. 894 to 900. In the system described therein, a high resolution high definition video signal with a constant aspect ratio of 1050 lines is filtered from the vertical, separated from it in a 2: 1 ratio, converted to an interlaced format of 525 lines, and encoded for digital transmission. The encoded signal is also decoded, converted back to progressive scan, interpolated from vertical to 1050 lines (luminance only), and subtracted from the high precision signal to form a difference signal that is quantized, entropy encoded, and then multiplexed into a common transmission channel with the encoded main video signal. 150 Mbps. Such a system is not compatible with conventional (analog) television transmission channels and requires a special digital decoder to be used in a conventional television receiver.

4,86785 The NBC system is the starting point for the invention shown below. The system is described in more detail in an article by M. Isnardi et al., Entitled "Docoding Issues in the ATV System"; IEEE Transactions on Consumer Electro-5 nics, February 1988, p. 111 to 120. Part of this system is also described in U.S. Patent 4,782,383, entitled "Apparatus for Processing High Frequency Information in a Widescreen Television System," issued November 1, 1988. Another document describing 10 of this system , has also been written by MA Isnardi et al. and entitled "Encoding for Compatibility and Recoverability in the ACTV System," IEEE Transactions on Broadcasting, Volume BC-33, no. 4, December 1987, IEEE (New York, US) ss. 116 - 123. These references are incorporated herein by reference. The system and patent application described in the referenced article use one 6 MHz channel to carry a complementary NTSC signal. When this signal is processed in a conventional NTSC receiver, a conventional television picture is produced. However, when this signal is processed in a conventional receiver that decodes and combines different component signals, an extended resolution widescreen image is produced. Component signals that complement the basic NTSC signal are located in inefficiently used areas of the NTSC spectrum. Although this high-resolution widescreen image has more horizontal and vertical resolution than a conventional television image and is very pleasing to the viewer, it may have less oblique resolution. Still, the overall level of 30 separation is still lower than in film.

SUMMARY OF THE INVENTION

Therefore, it is desirable to provide an additional signal to an NBC-type system that increases the resolution level of the reproduced image to more closely match the resolution of the film-35.

5,867,85 A system for producing a coded main and auxiliary signal representing a high definition picture includes a video source for producing a high definition video signal, a first signal encoding device coupled to the source for producing a coded main video signal compatible for reception and display in a conventional television receiver. a signal encoding apparatus associated with a source for producing an encoded additional video signal suitable for connection to the main video signal in a special receiver to produce a widescreen display. The present invention is characterized in that the video signal generated by the first encoding device contains widescreen information and represents an enhanced image with a higher resolution level 15 than in a conventional video image, but lower resolution than a high resolution image. A decoding apparatus is connected to the first signal encoding apparatus for decoding the encoded main video signal to form a main video signal. A signal separation device is connected to the decoding device 20 for generating a difference signal representing the difference between the video input signal produced by the source and the decoded main video signal produced by the decoding device, and a second signal encoding device connected to the signal separation device for generating an additional video signal.

The principles of the invention are also implemented in a television signal receiving system comprising first and second receiving devices for producing a respectively encoded main video signal received from a second analog transmission channel, said video signal being obtained from a high definition source, said encoded main video signal being compatible for reception and display in conventional television receivers. is a conventional aspect ratio, and a sig-35 signal processing device for decoding and combining, 86735

O

received signals to produce an enhanced resolution video signal. Such a system according to the invention is characterized in that the signal produced by the first named receiving device contains a widescreen-5 format and represents an extended resolution television image with a resolution higher than that of a conventional video image. The additional signal represents the difference between the high-precision widescreen signal from which the encoded main video signal is obtained and the additional signal obtained from the encoded main video signal. The signal processing apparatus includes a main signal processing apparatus coupled to the first receiver and controlled by the encoded main video signal to produce a baseband decoded main video output signal. The auxiliary signal processing device connected to the second receiving device and controlled by the encoded auxiliary video signal produces an output signal supplementing the baseband. The combining device, the first and second inputs of which are connected to the respective main and auxiliary signal processing apparatus, combines the baseband decoded main video signal and the baseband supplement signal by summing to form a video output signal representing a widescreen video image with a resolution greater than widescreen resolution.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a block diagram of a two channel television signal transmission system operating in accordance with the present invention.

Fig. 1a is a diagram illustrating the operation of a single 30 channel encoder suitable for use in the system shown in Fig. 1.

Figure 1b is a block diagram of a single channel encoder suitable for use in the system shown in Figure 1.

7 86785

Figures 1e - 2, 2a, and 3a to 3c are diagrams illustrating the operation of the various components of the single channel encoder shown in Figure 1b.

Figure 4 is a block diagram of a 1050 line / frame (L / F) 5 interleaved 525 L / F progressive scan converter suitable for use in the single channel encoder shown in Figure Ib.

Figures 5a and 5b are block diagrams of progressive interleaved scan converters suitable for use in the single channel encoder shown in Figure Ib.

Figures 6-8 are block diagrams of circuits suitable for use as a side-center signal separator and processor of the single channel encoder shown in Figure Ib.

Fig. 9 is a block diagram of a circuit suitable for use as an NTSC encoder of the single channel encoder shown in Fig. Ib.

Figures 10 and 10a to 10c are block diagrams and a map 20 illustrating the structure of the various vertical-temporal and horizontal-vertical-temporal filters used in the single-channel encoder shown in Figure Ib and the single-channel encoder shown in Figure 14.

Figures 11a and 11b are block diagrams of circuits suitable for use as a high in-frame center circuit of the single channel encoder shown in Figure Ib.

Figures 12 and 12a to 12d are block diagrams of silicon-30s suitable for use as any of the time expander or time compressor used in the single channel encoder and diagrams shown in Figure Ib to illustrate circuit operation.

86785 8

Fig. 13 is a block diagram of an amplitude compensation and quadratic modulation circuit suitable for use in the single channel encoder shown in Fig. Ib.

Figure 14 is a block diagram of a single channel decoder suitable for use in the present invention.

Figures 15 and 16 are a block diagram of circuits suitable for use as an in-frame averaging circuit-10 of the single channel decoder shown in Figure 14.

Fig. 17 is a block diagram of a square demodulator and amplitude expander suitable for use in the single channel decoder shown in Fig. 14.

Fig. 18 is a block diagram of a luminance-chrominance signal separation circuit suitable for use in the single channel decoder shown in Fig. 14.

Fig. 19 is a block diagram of a circuit suitable for use as a Y-I-Q format decoder in the single channel decoder shown in Fig. 14.

Fig. 19a is a block diagram of a side panel-center panel splitter suitable for use in the format decoder shown in Fig. 19.

Figures 20 and 21 are block diagrams of interleaved 25 progressive scan converters suitable for use in the circuit shown in Figure 14.

Fig. 22 is a block diagram of an auxiliary channel luminance signal encoder suitable for use as an auxiliary channel encoder of the system shown in Fig. 1.

Fig. 22a is a pixel diagram useful in explaining the operation of the circuit shown in Fig. 22.

Fig. 23 is a block diagram of a motion encoder and sub-pass filter suitable for use in the single channel encoder shown in Fig. Ib 35.

9 86785

Fig. 24 is a block diagram of an auxiliary channel chrominance signal encoder suitable for use in the auxiliary channel encoder shown in Fig. 1.

Fig. 24a is a video field diagram useful in explaining the operation of the circuit shown in Fig. 24.

Fig. 25 is a block diagram of an exemplary circuit for modulating and demodulating an additional signal generated by the additional channel encoder shown in Fig. 1.

Figures 26 and 27 are block diagrams of corresponding chrominance and luminance complement signal decoders suitable for use in the advanced compatible television dual channel decoder shown in Figure 1.

Fig. 28 is a block diagram of circuits suitable for use as a motion signal detector of the auxiliary channel luminance signal encoder shown in Fig. 22 and the luminance auxiliary signal decoder shown in Fig. 27.

Fig. 29 is a block diagram of an advanced compatible television dual channel decoder shown in Fig. 1.

Detailed description

In the drawings, arrows 25 marked with a single line represent buses carrying multi-bit parallel digital signals or signal paths carrying analog signals or single-bit digital signals. The type of signal carried by the bus or signal path is determined in the context in which it is processed. For clarity, compensatory delays may have been omitted from some signal paths. One skilled in the art of designing digital signal processing circuits knows where such delays are needed in a given system.

10 86785

Figure 1 is a block diagram of a dual channel television signal transmission system incorporating an embodiment of the present invention. In Figure 1, wideband high definition wide screen (HDTV, high definition 5 television) signals Y, I and Q from source 110 (e.g., a camcorder) are applied to transmitter 112. Signal Y includes luminance image information using the band between 0 Hz and 20 MHz, and signals I and Q contain chrominance key information using a frequency band between 10 Hz and 10 MHz.

The signals Y, I, and Q are encoded by a single channel encoder 114 of the type according to the document and patent application referred to above by Isnardi et al. The encoded signal is then transmitted by antenna 115. The television signals transmitted by antenna 115 are received by antenna 120 associated with a standard NTSC receiver 122 and antenna 124 associated with a widescreen extended definition television (EDTV) display 128 in a widescreen EDTV receiver. The widescreen EDTV receiver includes 20 single channel decoders of the type according to the above-mentioned Isnardi et al document and patent application, in which the received signal is decoded into its extended precision luminance and chrominance signal components Y ', 1' and Q '. Signals Y ', 1' and 25 Q 'are displayed on a wide screen EDTV display 128.

The system shown in Figure 1 further includes an additional channel encoder 142 that encodes for transmission a signal representing the difference between widescreen EDTV signals, as shown in widescreen EDTV receivers 30 where 125 and 128, and widescreen EDTV signals produced by the original source 110. This difference signal is generated by subtracting the output signals Y ', 1' and Q 'produced by the single channel decoder 140 from the respective signals Y, I 35 and Q produced by the source 110. The single channel decoder 140 may be similar to the decoder 125. The auxiliary television signal, which encoder 142 is producing is transmitted by antenna 130.

The auxiliary channel signal transmitted by the antenna 130 and the main channel signal transmitted by the antenna 115 are received by the antenna 132 associated with the widescreen HDTV reception system. This system includes a two-channel decoder 134 for decoding the main channel signal and the auxiliary channel signal, and further combining the decoded material luminance-10 and chrominance signals to produce corresponding widescreen HDTV component signals Y ", I" and Q "suitable for display on a widescreen HDTV. in the display unit 136.

Fig. 1a is a diagram illustrating signal-to-15 encoding techniques used in a single channel encoder 114. The encoder first reduces the bandwidth of HDTV signals to less than 6.0 MHz, converts HDTV signals from 1050 line / frame (L / F) interleaved scan format to 525 L / F in progressive scan format, and encodes the signals obtained by sit-20 into an NTSC-compatible video signal having four component signals. One component that produces constant aspect ratio and standard resolution color images in a conventional, or standard, NTSC receiver 122, and three components that represent standard image enhancements when played back on a widescreen EDTV receiver 128. The three enhancement components are combined with a component corresponding to describes the standard aspect ratio, the standard resolution, so that they use the same frequency band as the standard component, but are physically or perceptibly hidden in the standard picture formed by the television receiver 122.

The first component signal (i.e., the standard component) is a 525 L / F, 2: 1 interlaced signal with a constant 4: 3 aspect ratio. This component includes the middle band of the band-35 projected HDTV signal, which is temporally extended to use the i2 86735 for almost the entire active line time. The first component signal also includes lateral horizontal low frequency information temporally compressed into the left and right horizontal image sweep areas of the display produced by the constant NTSC receiver 122 5. The low frequency sideband formation is physically hidden from the viewer because it is not displayed on a standard NTSC receiver. From this first component signal, an in-frame average of 10 frequencies above 1.5 MHz is taken, for the reasons described below, before being combined with the second and third component signals. Figure 2 illustrates the time expansion and time compression of the first component signal.

The second component signal is a complementary 2: 1 yh-15 braided signal in which the high frequency information of the left and right sidebands is each extended to use half of the active line time. Thus, the extended sideband information uses substantially the entire active line time of the second component signal. Figure 3c 20 illustrates the generation of the second component signal.

The third component signal is a complementary 2: 1 lo length scan signal obtained from band-limited HDTV signals. It contains high frequency horizontal luminance separation information between approximately 5.0 and 6.0 MHz. This information is transmitted down frequency to use the frequency band between 0 and 1.0 MHz.

The second and third component signals are both averaged within the frame, compressed in amplitude, and then used to modulate the corresponding 90 ° phase-shifted suppressed subcarrier subcarrier signals ASC and ASC, whose frequency is an odd multiple of the multiple of the horizontal line scan. within the chrominance band of the signal spectrum 35. In addition, these alternating subcarrier signals are subjected to a 180 ° phase change from one field to another, where the field is defined as a period of 262 horizontal lines. Thus, any distortion in the constant NTSC image produced by the modulated second and third components is manifested as a change in complementary color at a frame rate (30 Hz). This type of distortion is not usually observed because the human eye is relatively insensitive to the rapid, mutually complementary color changes at the saturation levels typically found in television signals.

The fourth component signal is a 2: 1 interleaved "auxiliary signal" that contains vertical temporal (V-T) luminance separation information that would otherwise be lost when converting the HDTV signal from 525 LPF progressive scan to 525 LPF interlaced scan. This signal helps the EDTV receiver to reconstruct the missing information in moving objects and to reduce or eliminate unwanted flicker and motion disturbances in a progressively swept widescreen image formed on the EDTV display 128 and the HDTV display 136. Figure 2a illustrates how the auxiliary signal is generated. and 525 LPF of interlaced scan signals.

The fourth component signal is band-limited to 750 to 25 kHz, shifted over the first component signal by compressing the portions corresponding to the side bands and expanding the portions corresponding to the center band. The resulting signal is used to modulate a carrier signal that is 90 ° out of phase with the image carrier signal (also known as a video subcarrier signal) and is combined with the modulated first, second, and third components. The fourth component signal is hidden in a conventional NTSC display because any distortion it may cause is spatially correlated with the first component in the overlay process.

14 86735

The luminance signal portion of the first component signal is pre-filtered using a horizontal-vertical-temporal (H-V-T) comb filter to eliminate probable crosstalk between it and the chrominance signal portion of the first component-5 signal. In addition, the first, second, and third component signals are averaged within the frame using a VT comb filter to substantially reduce VT crosstalk between the first component and the second and third corners, allowing the second and third component signals to be easily separated in receiver decoders 125 and 134. Figs. le illustrate an in-frame averaging of the first, second, and third component signals, and combining them to form an encoded NTSC signal.

Although it is advantageous in the encoding and decoding processes, the comb filtering performed in the processing steps described above tends to reduce the transverse separation of the EDTV images displayed on the display 128. In Taa-20 frequency spectrum analysis, the luminance signal Y 'produced by a single channel decoder has a conventional NTSC video frequency spectrum up to approximately 1.5 MHz, and its frequency spectrum is characterized by alternating peaks and wells between 1.5 MHz and 6.0 MHz (i.e. .frequency spectrum of the comb - filtered signal 25). Frequencies above 6.0 MHz are missing from signal Y '. In addition, since the auxiliary signal used to regenerate the progressive scan luminance signal from the transmitted interlaced scan signal is band-limited to 750 kHz, luminance information at 750 kHz and 20 kHz may be missing during the alternating line periods of signal Y '.

Signals I and Q, the color information signal components of the HDTV signal, are both band-limited up to 600 kHz before being split between the first and second component signals. Since these signals, as parts of the first and second component signals, are V-T filtered when taking the in-frame average, the frequency spectrum of the repeated signals 1 'and Q' has alternating peaks and bumps, which is characteristic of the comb-5 filtered signal.

As discussed above, each of the difference signals ΔΥ, ΔΙ, and Δ0 is obtained by subtracting the 525 L / F progressive scan signals V, 1 ', and Q' produced by the single channel decoder 140 from the corresponding widescreen scan signals Y, I, and 10 of the 10TV L / F of the HDTV source 110. Q. Since there is a vertical shift between the displayed location of at least one of the lines of the interlaced scan signals Y, I and Q and the corresponding progressive scan signals Y ', 1' and Q ', each difference signal ΔΥ, ΔΙ and AQ can have a considerable amount of energy at relatively low frequencies.

Thus, the luminance difference signal ΔΥ may have energy in the frequency spectrum in the range between 0 Hz and 20 MHz, and the chrominance difference signals ΔΙ and Δ0 may have 20 energies in the frequency spectrum in the range between 0 Hz and 10 MHz. As described below, the auxiliary channel encoder 142 processes the signals ΔΥ, ΔΙ, and AQ to produce an auxiliary video signal having a bandwidth of 6 MHz.

Fig. 1b is a block diagram showing details of 25 widescreen HDTV sources 110 and a single channel encoder 114. The source used in the practice of this invention includes a widescreen 1050 LPF interlaced scan camera 10 synchronized to encoder 114 with a composite synchronization signal CCPS 30 and a frame rate 30. with a timing signal FT generated by the studio timing signal generator 11. The camera 10 used in this embodiment of the invention produces 525 lines of video information per field, or 1050 lines per frame. In each frame, the lines of one field are interleaved with the lines of the next field. Although only 1050 L / F camera sources are shown, it is contemplated that a 1050 L / F video recorder (VTR) or 1125 L / F interlaced scan camera or VTR may also be used as a source of video signals Y, I and Q. When using an 1125 L / F source, it may be desirable to reduce the number of lines per frame to 1050 by cropping the 1125 L / F image or interpolating.

The camera 10 used in this embodiment of the invention is a 525 LPF progressive scan camera that is controlled to move the image it produces vertically on the side of the line from field to field. This image transfer is performed by applying a frame rate (30 Hz) square wave signal FT produced by the studio timing signal generator 11 to the focus control of the camera 10. Controlled by a signal FT having a single value (e.g., a logic one), the camera used in this embodiment of the invention shifts the image down half of one line period. As a result of the second value of the signal FT, the camera 10 returns the image to its original non-shifted position. Moving the image does not affect the timing of the sig-20 signal produced by the camera.

The red, green, and blue (R, G, and B) output signals produced by the camera 10 are applied to a matrix 12, which converts the primary color signals to the luminance signal YA and the color difference signals IA and QA. The signals YA, IA and QA are digitized by analog / digital converters (ADCs) 14, which are controlled by a clock signal 8xfsc produced by the studio timing signal generator 11. The ADCs 14 produce widescreen HDTV signals Y, I and Q with a sampling frequency of sixteen times the frequency fse of the constant NTSC color subcarrier signal 30.

The following is an overview of the coding circuit shown in Fig. Ib. The signals Y, I and Q are applied from 1050 interleaved 525 to progressive scan converter 16. The output signals YF, IF and QF produced by the scan converter 16 are output from progressive interleaved converter 86735 17 to 17a, 17b and 17c, which produce signals YF IF 'and QF', each of which are 525 LPF interlaced scan signals. The scan converter 17c also produces a field difference signal YT, which is used to generate an auxiliary signal.

As discussed above, the auxiliary signal is used by single channel decoders 128 and 140 to generate 525 L / F progressive scan signals Y ', 1' and Q 'from the received 525 L / F interlaced scan signals. The signals YF ', IF' and QF 'are low pass filtered and applied 10 to a side-center signal separator and processor 18 which separates the signals into respective side and center band portions, separates the sideband portions into high and low frequency components, compresses the low bandwidth components of the sideband signals and middle band signals. The signals representing the expanded center band, YE, IE and QE, and the signals representing the compressed low frequency components of the sidebands, YO, 10 and Q0, are combined by a side-center combiner 28 to produce signals YN, IN and QN. These signals are applied to an NTSC encoder 31, the output of which is the first component signal C / SL of the single channel EDTV signal described above.

The high frequency components of the sidebands, YH, IH, and QH, produced by the separator and processor 18, are applied to an NTSC encoder 60, the output signal circuit 62 of which expands 25 with respect to time. The output signal of circuit 62 is the second component signal ESH of the one channel EDTV signal described above.

The signal YF 'produced by the progressive interleaved converter 17c is bandpass filtered between 5.0 MHz and 6.0 MHz by a filter 70, shifted from 0 MHz to 1.0 MHz by a modulator and a low-pass filter 72, and compressed with time to chance together with the center band portion of the first component signal produced by the format encoder 74. The output signal of encoder 74 is the third component of the EDTV signal.

18 86785

The second and third component signals are averaged within the frame by respective circuits 64 and 76 to produce signals X and Z. The signals X and Z are amplified compressed and used by circuit 80 to modulate the two 90 ° phase shift subcarrier signals ASC and ASC '. The output signal of circuit 80 is signal M. From the first component signal C / SL, an in-frame average is taken by circuit 38, resulting in signal N. Signals M and N are combined by sum-10 with ground 40 to produce signal NTSCF.

The field difference signal YT produced by the converter 17c from progressive to longitudinal sweep is temporally positioned by the format encoder 78 to be in a format corresponding to the first component signal. The resulting signal 15 is applied to a motion encoder and a low-pass filter 79 to generate a fourth component or auxiliary signal YTN.

The signals NTSCF and YTN are converted to analog form by respective digital / analog converters (DAC) 54 20 and 53. The analog switch 58 adds the composite synchronization signal OCPS generated by the studio timing signal generator 11 to the signal NTSCF. The signal generated by the switch 58 and the auxiliary signal are applied to a radio frequency (RF) transverse modulator 57, which forms the first television signal component. This signal is transmitted via antenna 115 by transmitter 55.

The following is a more detailed description of the circuit shown in Figure Ib. The studio timing signal generator 11 generates composite synchronization signals CCPS and OCPS 1050 for the LPF 30 interlaced scan camera 10 and 525 LPF for the interleaved coded output signal MAIN. In addition, the generator 11 produces clock signals 4xfsc, 8xfsc and 16xfsc with respective frequencies of four, eight and sixteen times the frequency of the NTSC color subcarrier signal, da-35 sampled oscillating signals ASC, ASC 'and fc, 19 86735 line speed signals F The switching signal SW is produced to control the input of the signal OCPS to the coded output signal MAIN. Circuit 11 includes a conventional resonant crystal controlled oscillator 5 which generates a clock signal 16xfsc. Circuit 11 generates from the signal 16xfsc a 90 ° phase shift alternating subcarrier signals ASC and ASC having a frequency of 3.1 MHz substantially equal to 395 times half the horizontal line scan frequency, and a signal fc having a frequency substantially equal to 5 MHz. The signals ASC, ASC 'and fc can be generated, for example, by incrementing the counter (not shown) with the signal löxfsc and entering the counter value into a read only memory (ROM) (not shown) programmed to produce sample values representing the three signals. In addition, the ROM may produce an output signal indicating a predetermined pixel sampling time (H) or a group sampling times (Fs) in each horizontal line period of the signal NTSCF. One of these signals may be applied to an additional counter (not shown) programmed to produce signals, such as the signal FT described below with reference to Figures 11a and 11b, which occur at frame or field rate. An exemplary device for generating various clock and timing signals is described in more detail in U.S. Patent Application 241,277, entitled "Video Signal Synchronization System as for an Extended Definition Television System", which is incorporated herein by reference.

The widescreen HDTV signals Y, I, and Q produced by source 110 are applied from 1050 interleaved 525 to progressive scan to converter 16, which produces signals YF, IF, and QF. The scan converter 16 is part of a single channel decoder 114. Fig. 4 is a block diagram showing a circuit suitable for use in a scan converter 16. The circuit shown in Fig. 4 represents one of the three channels of the scan converter 16 35 (for signals Y, I and Q, respectively).

20 86 7S5

The other two channels (not shown) may be identical to those shown and will not be described in detail. This circuit interpolates the sample values between the 1050 LPF interlaced input signals and connects these 5 interpolated samples to the interpolator tower samples 525 to generate LPF progressive scan signals at the output ports of the two dividing circuits 482. The interpolated samples generated by the scan converter 16 are obtained by motion adaptive processing. In the regions of the image that represent motion, the interpolated samples are formed from samples separated by a single horizontal line period (i.e., the vertical average). In stationary areas of the image, the interpolated samples are formed from samples separated by a single frame period (i.e., a temporally taken average).

In Figure 4, the 1050 LPF interlaced scan signal is applied to a horizontal low pass filter 438. The filter 438 can average 16xfsc of successive samples of signals Y, I, and Q to produce samples that can be subsampled with the signal 8xfsc without spurious image distortion. The signal produced by the filter 438 is applied 524 to the delay element 440 of the horizontal line section (524 H), which is the first of four delay elements connected in series. The other three delay elements 442, 444 and 446 produce respective signal delays of 1H, 1H and 524 H. The output signal of the delay element 442 is a reference signal. The output signals of the delay elements 440 and 444 are promoted and delayed by an amount of 1 H relative to the reference signal. The input signal of the scan converter and the output signal of the delay element 446 are advanced 30 and delayed by one field period of the 1050 LPF interlaced scan signal relative to the reference signal, respectively.

The output signals of the delay elements 440 and 444 are applied to the respective input ports of the multiplexer 450. The multi-35 plexiglass 450 is controlled by the signal FT produced by the studio driving signal generator 11 to alternately produce a signal from the delay element 440 and a signal from the delay element 444 during successive field periods of the signal produced by the camera 10. The signal 5 produced by the multiplexer 450 is combined with the reference signal produced by the delay element 442 by an adder 458 and a difference circuit 456. The output signal of the adder 458 is a vertical average signal VAVG and the output signal of the difference circuit 456 is a vertical difference signal VDIFF. Signals VAVG and VDIFF 10 represent the respective added and subtracted combinations of a reference signal delayed by 525 H relative to the input signal and a signal alternately delayed by 524 H and 526 H relative to the input signal. The switching 15 of the signal to be combined with the reference signal is controlled by the signal FT to monitor the image shift of the camera 10.

The input signal of the scan converter and the output signal of the delay element 446 are combined in an adder 460 and a difference circuit 462 to form a temporal average signal TAVG and a temporal difference signal TDIFF, respectively. The temporal mean signal TAVG and the temporal difference signal TDIFF always represent combinations of signals separated by 1050 H in one frame period.

The signals VDIFF and TDIFF are applied to a conventional motion detector 464, the output of which is applied to a conventional motion spreader 466. The detector 464 may include, for example, a comparator (not shown) comparing the relative values of the respective temporal difference signal and vertical difference signal TDIFF and VDIFF. In addition, the sensor 464 may include a horizontal low-pass filter (not shown). The motion spreader 466 may include a circuit (not shown) that sets a threshold for the motion signal produced by the sensor 464, and averages the image of the threshold signal 35 in horizontal, vertical, and temporal motion 22 86785 to spread the boundaries of the regions to the non-motion regions. The motion signal is propagated to mitigate edge phenomena in any motion adaptive processing that uses the motion signal. In general, treating a stationary area as a mobile area causes fewer problems than treating a mobile area as a stationary area.

Motion spreader 466 produces motion indicating signals M and (1 to M), which are applied to respective multipliers 468 and 470. Multipliers 468 and 470 are configured to scale the respective signals VAVG and TAVG with these motion indicating signals. The output signals of multipliers 468 and 470 are summed by adder 472 to produce an interpolated signal.

15 The signals M and (1 - M) are mutually complementary sub-values. The magnitude of M is directly proportional to the level of intra-field motion among the video images represented by the reference signal and the signal above and out of the field. Thus, when there is a lot of intra-field movement in a part of the displayed image 20, the interpolated signal of that part has a relatively large component obtained from the vertical average signal VAVG and only a small component obtained from the temporal average signal TAVG. The relatively 25 stationary areas of the image, i. where the level of intra-field motion is low, these relationships are reversed.

Switching between the vertical average and the temporal mean in the presence or absence of intra-field motion produces interpolated signals with full temporal separation in the moving parts of the image and full vertical separation in the stationary parts of the image. The inventors have found that this type of motion-adaptive processing produces better results than pure temporal or pure vertical interpolation.

23 86735

The output signal of adder 472 is scaled by a factor of half in two divider circuits 474 to produce a digital signal using the same range of amplitude values as the signal applied to sweep converter 5 16. The output signal produced by two divider circuits 474 port is connected by two to the input ports of the dividing circuit 482. The adder 480 and the two dividing circuit 482 form the average of the sample values produced by the camera 10 and the interpolated sample values representing the lines between the 1050 LPF interlaced scan images. Thus, the output signal produced by circuit 482 can be considered to be the average of successive line pairs 1050 of the LPF progressive sweep signal. The inventors have found that this is a good approximation of the 525 LPF progressive scan signal.

Referring to Figure Ib, the 525 LPF progressive scan signals IF, 20 QF and YF produced by the scan converter 16 are applied to different progressive scan interleaved scan converters 17a, 17b and 17c, respectively.

A circuit suitable for converter 17c is shown in Figure 5a, and a circuit suitable for transducers 17a and 17b is shown in Figure 5b.

The scan converter shown in Fig. 5a includes a signal averaging circuit formed by one field (525 H) delay elements 510 and 512, an adder 514, and two dividing circuits 516. The signal produced by the signal averaging circuit is the average of the input signal YF 30 or B, the delay element 512 A. These signals are separated by two field periods. The average signal is subtracted in the difference circuit 518 from the signal X produced by the delay element 510 to produce a field difference signal YT. Signal X represents the field of the video signal between signals A35 and B. Signals X and YT are applied to the various 24,86735 terminals of switch 520. The contact of switch 520 is controlled by a square wave signal Fs having an effective period of fifty percent and a frequency substantially equal to the horizontal line scan frequency defined by the NTSC standard. The signal Fs controls the switch 520 to alternately produce line samples from the signal X, and line samples from the signal YT. These samples are written to the dual port memory 522 under the control of the signal 8xfsc and read from the memory as signals YF 'and YT in parallel under the control of the signal 4xfsc 10, the frequency of which is substantially equal to four times the frequency of the color subcarrier. The signal YF 'is a 525 LPF interlaced scan luminance signal. The signal YT is used to generate the auxiliary signal described above with reference to Figure 2a.

The scan converter shown in Figure 5b includes a single field compensating delay 530 and a dual port memory 532. Delayed samples of IF or QF signals are stored and read from memory 532 under the control of the respective signals 8xfsc and 4xfsc. The output signal of the memory 532 is a 525 20 LPF interlaced scan signal IF 'or QF', which is suitably time dependent on the signal YF '.

Referring to Figure 1b, the signals IF ', QF' and YF 'produced by the respective sweep converters 17a, 17b and 17c 25 are applied to the respective low pass filters 19a, 19b and 19c. Filters 19a and 19b reduce the horizontal bandwidth of the signals IF 'and QF' to 500 kHz. Filter 19c lowers the horizontal bandwidth of the signal YF 'to 5 MHz. The output signals IF ", QF" and YF "produced by the respective low pass filters 19a, 19b and 19c are applied to a side-center signal separator and processor 18. Details of the processor 18 are shown in Figures 6, 7 and 8.

Fig. 6 is a block diagram of a portion of a processor 18 that separates the luminance signal YF "into a signal YE representing temporally expanded midband pixels, a signal YO representing time-compressed sideband pixels, and a signal YH representing a sideband pixel; high-frequency components of pixels.

5 Signals YE and YO are used to form the first component of the encoded EDTV signal, and signal YH is used to form the second component signal.

In Fig. 6, a low-pass filter 610 filters the signal YF ”to produce a signal YL using the frequency-10 new band between 0 Hz and 700 kHz. The signal YL is subtracted from the delayed signal YF "by a separation circuit 612 to form a signal YHO using the frequency band between 700 kHz and 5 MHz. The delayed signal YF", the signal YHO and the signal YL are applied to a demultiplexing circuit 1516. Circuit 616, described below with reference to Fig. 8 , passes a portion of the signal YF "corresponding to the center band of the image as a signal YC and portions of the signals YHO and YL corresponding to portions of the sideband of the image as signals YH and YL '. The signal YC is expanded with time exponent 622 to produce a signal YE, as shown in Fig. 2, The signal YL 'is time-compressed, also as illustrated in Figure 2, by circuit 628 to produce a signal YO with a bandwidth of 4.2 MHz. The time-expanding circuit 622 and the time-compressing circuit 628 may be implemented with the circuit described below with reference to Figures 12 and 12a-12d.

The circuit shown in Fig. 7 is similar to the circuit shown in Fig. 6, except that the passband of the low-end filter 710 is 0 Hz to 83 kHz instead of the band 6 Hz to 700 kHz of the filter 610 of Fig. 6, which it corresponds to.

Fig. 8 is a block diagram of a demultiplexer circuit that may be used as the demultiplexer 716 of Fig. 7, or, as illustrated, the demultiplexer 35616 of Fig. 6. The circuit shown in Fig. 8 includes three multiplexers 861035 26 series 810, 812 and 814 each connected to a respective quantity comparison circuit 8. 818 and 820. Each of these quantity comparison circuits is connected to receive a count value from the counter 822. The counter 822 is clocked with a signal 5 4xfsc. The counter 822 is reset by a signal H produced by the studio timing signal generator 11, which occurs once in a horizontal line period, indicating the location of the first image sample in each line.

Controlled by count values 1 to 84 and 671 to 754, the quantity comparison circuits 817 and 820 control the respective multiplexers 810 and 814 to pass the signals YHO and YL, respectively. Multiplexers 810 and 814 pass the blackout signal BLK as a result of all other number values. Similarly, the quantization comparator circuit 818 controls the multi-15 plexiglass 812 to pass the signal YF "at numbers between 75 and 680 and to pass the blackout signal at other times. -20 to facilitate the reconstruction of the signal in the decoder (described below).

As discussed above, time expander 622 and 722 and time compressors 628 and 728 may be implemented with circuits such as those shown in Figure 12. The circuit shown in Figure 12 uses four dual port memories 1216a-1216d to expand the sequence of pixel values over time (by repeating samples) or to compress time. (by removing samples). The circuit shown in Figure 12 further includes a pair of spike filters 1220 and 1222, 30 for amplifying the high frequency components of the signals produced by the two-port memories, and a two-point linear interpolator 1230 for combining the spiked signals to form time-expanded or time-compressed video signals. Spike filters 3520 and 1222 compensate for the low-pass filtering required in two-point linear interpolation. The pixel counter 1210 and the two programmable read-only memories (PROMs) 1212 and 1225 control the circuit shown in Figure 12. By appropriately programming the PROMs 1212 and 1225, the circuit of Figure 12 can be used to implement a number of expansion and compression factors.

During operation, the video input signal S, which may be, for example, the signal YC produced by the demultiplexer 616 of Fig. 6, is applied to a two-port memory 1216a and three series-connected delay elements 1214a, 1214b and 1014c. The output signals of the delay elements 1214a, 1214b and 1214c are applied to the input ports of the respective dual port memories 1216b, 1216c and 1216d. The write address signal of the memories 1216a to 1216d is produced by the pixel counter 1210. The signal M is also applied to the PROM 1212, which forms the read address signal 15 for the memories 1216a to 1216d, and the interpolation factor DX, which is applied to the two-point linear interpolator 1230. 1225. If the circuit shown in Figure 12 is used to expand the signal over time, the PROM 1212 is programmed to produce a read address signal N that increases more slowly in value than the signal M. This causes memories 1216a-1216d to repeat sample values. If, on the other hand, the circuit is used to compress the signal with respect to time, the read address signal N increases in value faster than the signal M. This causes the memories 1216a -25 1216d to jump over the sample values. The PROM 1225 can be programmed, as shown in Figure 12d, to change the spike factor PX applied to the spike filters 1220 and 1222 under the control of various interpolation factors.

Fig. 12b shows details of spike filter-30 from 1220 and 1222, and a two-point linear interpolator 1230. Fig. 12c shows details of high-pass filter 1240 used in spike filter 1220, an identical filter is used in spike filter 1222.

Referring to Figure Ib, the separated and time-expanded 35 midband signals IE, QE, and YE separated 28,86785 and the time-compressed sideband signals 10, QO, and YO produced by processor 18 are applied to a side-center combiner 28. The combiner 29 may include a counter (not shown), which is reset by signal H and clocked by signal 4xfsc. The number signal produced by this counter is applied to a multiplexer (not shown) which combines the signals YE and YO, as shown in Fig. 3a, to form an output signal YN. The combiner 28 includes a circuit (not shown) of the same type for combining signals IE 10 and 10 to form a signal IN, and for combining signals QE and QO to form a signal QN.

The signals YN, IN, and QN produced by combiner 28 are applied to NTSC encoder 31. Encoder 31 includes a horizontal-15 ta-vertical temporal (HVT) filter 34 that processes the signal YN by a comb filter transfer function that removes spectral components of the luminance signal that may be modulated. and mixed by the modulated alternating subcarrier signal. The output signal of the HTV filter 34 is the signal YP. The color difference signals IN and QN are applied to a transverse modulator 30 which, under the control of a clock signal 4xfsc, generates an NTSC chrominance signal CN. The signal CN is applied to a vertical temporal (VT) filter 32, which comb-filters the signal CN 25 to remove spectral components corresponding to the modulated alternate subcarrier and the high frequency luminance signal. The output signal of the filter 32 is the signal CP. The signals CP and YP are summed by a signal combiner 36 to form a signal C / SL, which is the first component signal of the encoded EDTV signal.

Fig. 9 is a block diagram showing further details of the NTSC encoder 31. In Fig. 9, the latch circuit pair 910 and 912 are controlled by clock signals 2xfsc and 4xfsc to time division multiplex the signals IN and QN into a signal 35 which alternates I and Q samples. The circle at the input of the holding silicon 86785 29 indicates that it is controlled by the 2xfsc inversion of the clock signal. The signal produced by latch circuits 910 and 912 is applied to another pair of latch circuits 914 and 916, which change the polarity of the alternating sample pairs, as shown in Figure 9 to produce a signal CN. The signal CN is applied to the VT bandpass filter 32. The signals 4xfsc, 2xfsc and f it are produced by the studio timing signal generator 1a.

Fig. 10 is a block diagram of an FIR filter suitable for use as filter 32. The filter shown in Fig. 10 uses samples from four consecutive fields to generate its output signal. Each of the nine taps of the filter is multiplied by the corresponding filter coefficient ai-a9. Fig. 10a is a diagram of coefficient values that can be used to configure the filter shown in Fig. 10 as a VT bandpass filter 32 or a VT bandpass filter as used in the HTV bandpass filter 34.

Figure 10b is a block diagram of a filter suitable for use as an IDA bandpass filter 34. In reality, filter 34 has a flat frequency response curve between 0 Hz and 1.5 MHz, and a comb-type frequency response curve between 1.5 MHz and 4.2 MHz. The zeros on the comb curve are positioned to substantially prevent components of the signal YN that may appear as crosstalk in the modulated chrominance signal. In Fig. 10b, the signal YN is applied to a horizontal low-pass filter 1020 having a passband of 0 Hz to 1.5 MHz. The output signal of the filter 1020 is subtracted from the signal YH produced by the compensating delay element 1022 to produce a high-pass filtered luminance signal. This high-pass filtered luminance signal is applied to a VT bandpass filter 1021, as described above with reference to Figures 10 and 10a. The output signal of the VT bandpass filter 1024 is combined with a low-pass filtered luminans-35 inner signal produced by the compensating delay element 1024 to produce an output signal YP. As shown in Fig. 9, the signal YP and the signal CP produced by the VT bandpass filter 32 are summed by a combiner 36 to produce a first component signal 5 C / SL.

Referring to Figure Ib, the signals IH, QH, and YH produced by the page-center signal eot and processor 18, which represent the high-frequency components of the wideband sidebands, are input to the NTSC encoder 60. The encoder 60 may be the same as the NTSC encoder described above. 31. The output signal NTSCH of the encoder 60 is applied to a time-expanding circuit 62 which expands the high-frequency signals of the sideband, as shown in Fig. 3c, to generate a signal ESH using a portion of the horizontal line time corresponding to the center of the signal C / SL. -kaistaosaa. The time expanding circuit 62 may be implemented with a circuit as described above with reference to Figures 12 and 12a-12d. The ESH signal is the second component signal of the widescreen EDTV signal.

The luminance signal YF 'produced by the progressive interleaved scan converter 17c is applied to a bandpass filter 70 which passes frequencies in the range of 5 MHz to 6.0 MHz to an amplitude modulator and a low pass filter 72. A modulator 72, which may be conventional in design, mixes the band end 70 the signal produced by the signal to a sampled blue signal fc having a frequency of substantially 5 MHz. The signal fc is generated by a studio timing signal generator 11. Circuit 72 includes a low pass filter that substantially removes the baseband 30 signal and modulation signal components from above 1.0 MHz. The operation performed by circuit 72 is substantially a frequency shift of high frequency luminance information from the 5 to 6.0 MHz band to the 0 to 1.0 MHz band. The signal produced by the modulator and low-pass filter 72 is applied to a format-35 encoder 74, which compresses the signal over time to use a portion of the horizontal line period corresponding to the center band portion of the signal C / SL.

The signal C / SL produced by the NTSC encoder 31, the signal ESH produced by the time-expanding circuit 62, and the signal produced by the format encoder 74 are averaged in-frame by the respective in-frame averaging circuits 38, 64, and 76, respectively.

The in-frame averaging process makes the signals in each of the two fields 10 of any frame 10 identical. This process is important so that the second and third component signals can be combined with the first component signal in a manner that allows a wide screen EDTV signal to be easily separated by an EDTV decoder.

As discussed above, the C / SL signal is averaged within the frame only for frequency components above 1.5 MHz and only in the midband range. Components and sideband portions below 1.5 MHz are not averaged within the frame to maintain vertical and temporal separation of the reconstructed image. Fig. 11b is a block diagram of a circuit suitable for use as a high-frame averaging circuit 38. In Fig. 11b, an input signal, in this case a C / SL signal, is applied to a pair of single field (262 H) delay elements 2520 and 1122 connected to the output. Y1 + C1 is applied to the input port of the averaging circuit 1128, and its second port is connected to receive either the input signal Y2 + C2 or the output signal of the delay element 1122 through the multiplexer 1125. The multiplexer 1125 oh-30 is divided by a frame signal Fs applied to its control input port to produce a signal Y2 + C2 during one field of the frame, and to produce an output signal of the delay element 1122 during the second field of the frame. The multiplexer 1125 always outputs a signal that is in the same frame as the signal Y1 + Cl produced by the vii-35 ve element 1120.

32 86735

The average circuit 1128 scales the signal Y1 + Cl by a factor of -1/2, scales the signal given by the multiplexer 1125 by a factor of 1/2, and sums the scaled signals. The signal produced by the averaging circuit 1128 is high pass filtered by a filter 5 1130 to remove components below 1.5 MHz. The output signal of filter 1130 is applied to gate 1132. Gate 1132 is controlled by a signal which may be generated, for example, by a circuit including a pixel counter (not shown) and a decoder (not shown) to pass the signal from filter 1130 only during the center band portion of the signal. The output signal of port 1132 is applied to the input port of adder 1134. The second input port of the adder 1134 is connected to receive the signal Y1 + Cl produced by the delay element 1120. The output signal of the adder 1134 15 is the signal N shown in Fig. 1b, the output signal of the high intra-frame average pair 38.

A circuit suitable for both in-frame average circuits 64 or 76 is shown in Figure 11a. In Fig. 11A, the signal IN, in this case the signal ESH, is applied to a pair of delay elements 1110 and 1112 connected in series. The output signal Y1 + C1 of the delay element 1110 is applied to the input port of the average circuit 1118 and the input signal IN, or the output signal of the delay element 1112 is applied to the circuit 1118. input port through multiplexer 1115. The multi-plexer 1115 is controlled by a field signal Fs at a field rate (30 Hz) to alternately pass the signal IN and the output signal of the delay element 112 during the alternating field periods of the input signal. The signal produced by the multiplexer 1115 is always in the same frame period as the signal Y1 + Cl produced by the delay element 1110. The average circuit 1118 scales each of its input signals by a factor of half and sums the resulting signals to form an intra-frame average output signal.

Referring now to Figure Ib, the output signal of the high-frame average circuit 38 of the high frame 35 is applied to one input port of the summer 33. The output signals X and Z from the respective in-frame averages 64 and 76 are compressed in amplitude and used by modulator 80 to modulate the alternating subcarrier signals ASC and ASC at 90 ° phase shift to produce a signal M. The signal M is applied to the second input port of the adder 40. The output signal NTSCF of the adder 40 is a combined first, second and third component signal of the widescreen EDTV signal. The NTSCF signal is applied to a digital / ana-10 logic converter (DAC) 54.

Figure 13 is a block diagram of a circuit suitable for modulator 80. In Fig. 13, the signals X and Z corresponding to the respective intra-frame averages 64 and 76 are output to the respective PROMs 1310 and 1312. Both PROMs 1310 and 1312 are programmed with an amplitude compression function, exemplified by the graphical input-output function PROM. next to 1312. The output signals of the respective PROMs 1310 and 1312 are applied to the respective first input ports of the multipliers 1314 and 1316, respectively. The second input port of the multiplier 1314 is connected to receive the alternate subcarrier signal ASC ', and the second input port of the multiplier 1316 is connected to receive the alternate subcarrier signal ASC. The signals ASC and ASC 'are produced by the studio timing signal generator 11. The output signals of the multipliers 1314 and 1316 are summed by an adder 1320 to produce a cross-modulated output signal M.

Referring to Fig. 1b, the frame difference signal YT produced by the progressive-to-interleaved conversion circuit 17c is input to a format encoder 78. The format encoder 78 may include a circuit of the same type as described above with reference to Figs. 12 and 12a to 12d. Encoder 78 expands the centerband portion of the signal YT and compresses the sideband portions, as illustrated in Figure 1f. The signal produced by the format encoder 78 is applied to a motion encoder 86785 34 and a low pass filter 79. An example circuit for use as a motion encoder and a low pass filter is shown in Fig. 23.

The motion encoder of circuit 79 reverses the variations in signal-5 yT, which represents the motion of the image, to +10 IRE or -10 IRE depending on whether the signal is positive or negative, respectively. This translation improves the YT behavior of the decoded signal at the receiver when there is a significant amount of interference on the transmission channel between the transmitter and the receiver. The use of the motion coded "auxiliary signal" is described below with reference to Figures 22 and 27.

In the circuit shown in Fig. 23, the signal produced by the format encoder 78 is applied to the input port of the adder 2368 and the comparator 2364. The comparator 2364 generates an output signal indicating whether the signal produced by the format encoder is positive, negative or zero. The signal generated by comparator 2364 directs multiplexer 2366 to pass a digital value representing 10 IRE units, -10 IRE units, or 0 IRE units-20 if the auxiliary signal is positive, negative, or zero, respectively. The signal produced by the multiplexer 2366 is applied to the second input port of the adder 2368. The output signal of adder 2368 is applied to a low pass filter 2378 having a passband between 0 Hz and 750 kHz. The output signal of the filter 25 2378 is the output signal YTN of the circuit 79. This signal is applied to the circuit DAC 53 shown in Figure Ib.

DACs 53 and 54 form analog versions of the respective signals YTN and NTSCF. The signal generated by the DAC 54 is applied to an analog switch 58, which is controlled by a signal SW generated by the studio timing signal generator 11 to place a composite synchronization signal OCPS, also generated by the generator 11, in the horizontal and vertical blackout periods of the analog NTSCF signal. Although the signal OCPS is shown as an analog sig-35 signal, it has been thought that the digital signal OCPS

35 86785 can be used. In this case, the analog switch 58 would be replaced by a conventional dual input multiplexer located between the adder 40 and the DAC 54.

The output signal of switch 58 is applied to one input of a radio frequency 5 (RF) cross modulator 57. The second input of the modulator 57 is connected to receive an auxiliary signal, an analog version of the signal YTN produced by the DAC 53. The modulator 57 produces a secondary band television signal MAIN, where the signal NTSCF is 10 as an in-phase component and the auxiliary signal as a 90 ° phase shift component. The signal produced by the modulator 57 is applied to a transmitter 55 which transmits the signal via an antenna 115.

As shown in Figure 1, the signal MAIN produced by the single channel encoder 114 is applied to the single channel decoder 140. A circuit suitable for the decoder 140 is shown in Figure 14. In review of its operation, the decoder shown in Figure 14 demodulates the signal MAIN in its in-phase and 90 ° to recover the NTSCF and YTN of the respective 20 components in phase shift. The NTSCF signal is processed to recover the first, second, and third component signals. These signals are further decoded and compressed or expanded, as appropriate, and combined to produce a luminance signal YF 'and color signals IF' and QF ', all of which are 525 LPF progressive scan signals. The auxiliary signal YTN is also decoded and used to convert the signal YF '525 to an LPF progressive scan signal. The color difference signals IF 'and QF' are converted to a progressive wipe format without the aid of an auxiliary signal. Finally, the progressive scan signals YF, IF and QF are converted to analog signals Y ', 1' and Q '.

In Figure 14, the signal MAIN is applied to an input unit 1422. The input unit 1422 includes a radio frequency (RF) tuner and amplifier circuit, a synchronous video demodulator, 36 86785 that separates the in-phase and 90 ° phase shift modulation component from the received video signal, and an analog / digital converter (ADC). The NTSCF signal produced by the ADC of the input unit 1422 represents the in-phase modulation component of the signal 5 MAIN; the signal YTN represents the modulation components in 90 ° phase shift.

The signal NTSCF is applied to a synchronizing signal separation and clock signal generating circuit 1432. The circuit 1432 10 includes a conventional circuit that separates the horizontal and vertical synchronization signals HS and VS, respectively, from the signal NTSCH and combines the signals HS and VS to form a composite signal sync. Circuit 1432 also includes a conventional phase locked loop 15 (PLL) that generates two clock signals CK4 and CK8 with respective frequencies of 4xfsc and 8xfsc, four times and eight times the frequency fse of the color synchronization burst signal component of the NSCH signal. Circuit 1432 generates from the signal CK4 90 ° phase shift alternating subcarrier signals ASC and ASC having a frequency of substantially 3.1 MHz, 395 times half the horizontal line scanning frequency, and a signal fc having a frequency of substantially 5 MHz. The signals ASC, ASC 'and fc can be generated, for example, by incrementing the counter (not shown) by the signal CK4 and applying the counter value to a read-only memory (ROM) programmed to produce sample values corresponding to the three signals. In addition, the ROM may produce an output signal H indicating a predetermined pixel sampling moment in each horizontal line period 30 of the NTSCF signal. This signal can be applied to another counter (not shown) programmed to produce signals, such as the signal FS described below with reference to Figures 15 and 16, which occur at frame or field rate. An example circuit for generating different clock and timing signals at a receiver is described in more detail in the aforementioned U.S. Patent Application 241,277.

The signal NTSCF produced by the input unit 1422 is also applied in-frame to the averaging difference circuit 1424. The circuit 1424 generates average pixel values and pixel difference values for the respective pixels in the two fields that form the frame. The average pixel values are the output signal N corresponding to the first component of the EDTV signal, and the pixel difference values 10 are the output signal M corresponding to the modulated second and third components of the wide-screen EDTV signal. Figs. 15 and 16 are block diagrams showing a circuit suitable for use as an intra-frame average difference receiving circuit 1424. Fig. 15 ha-15 illustrates the operation of the circuit 1424 during the first field of the frame and Fig. 16 illustrates the operation of the circuit during the second field of the frame.

In Figures 15 and 16, the NTSCF signal is applied as an input signal to two series-connected delay elements 2020 and 1522. The delay elements 1520 and 1522 each produce a time delay substantially equal to one field period (262H). The signal IN and the output signal of the delay element 1522 are applied to different input ports of the multiplexer 1525, respectively. The multiplexer 1525 is controlled by the frame no-25 speed signal FS produced by the timing signal generator 1432 to pass its two input signals in alternating field periods of the signal MAIN. The signal produced by the multiplexer 1525 is always in the same frame as the signal produced by the delay element 1520. The signal produced by the delay element 1520 30 and the signal produced by the multiplexer 1525 are applied to the average circuit 1528.

Circuit 1528 scales the signals produced by delay element 1520 and multiplexer 1525 by factors -1/2 and 1/2, respectively, and sums the scaled signals. This operation 35 removes all NTSCH components of the signal that are common to the two fields forming the frame, i.e. widescreen The first component signals of an EDTV signal with frequencies greater than 1.5 MHz. The output signals of the averaging circuit 1528 are applied to a horizontal over-5 pass filter 1530, which substantially removes the first component signals with frequencies below 1.7 MHz. The signals produced by filter 1530 are passed through port 1532, which is controlled by signal CS, which is generated by timing signal generator 1432. The signal CS directs port 1532 to pass only the center band portion of the signal produced by filter 1530. This part is an inverted version of the signal M, a wide screen of the cross-modulated second and third component signals of the EDTV signal. The signal produced by port 1532 is summed by adder 1534 with the signal produced by vii-15 ve element 1520 to produce a first component signal N, and is complemented by circuit 1535 to produce signal M.

Referring to Figure 14, the signal M produced by circuit 1424 is applied to a transverse modulator and amplitude 20 expansion circuit 1426. Circuit 1426 cross-modulates the signal M and expands the amplitude of the resulting in-phase and 90 ° phase shift signal to recover the second and third components of the EDTV signal, respectively. Fig. 17 is a block diagram of a circuit adapted for use as a transverse modulator and amplitude expanding circuit 1426.

In Fig. 17, the signal M is multiplied by the signals ASC and ASC in the respective multipliers 1710 and 1712. The output signals of the multipliers 1710 and 1712 are applied to the respective low-pass filters 1713 and 1715, which substantially remove the signal M and all high frequency modulation components. The output signals of filters 1713 and 1715 are applied to the respective programmable read-only memories (PROMs) 1714 and 1716. PROMs 1714 and 1716 are programmed with an amplitude expansion function, which is the inverse of the amplitude compression function used in the encoder to hide the second psychophy. component signal in a compatible composite signal. The output signal X of the PROMr 1714 is a decoded second component signal, the expanded high frequency components of the 5 sideband signals. The output signal Z of the PROMr 1716 is a decoded third component signal, the frequency-shifted high-frequency luminance signal components of a wideband EDTV signal.

Referring to Fig. 14, signal X is applied to sideband compression circuit 1428, which effectively reverses sideband data expansion by the encoder circuit. This operation produces a signal NTSCH representing the high frequency components of the sideband signals, restored to their correct time dependence on the time-compressed signal of the YIQ-15 format encoder 1444 with the centerband signal described below with reference to Fig. 19. The compression circuit 1428 may be implemented with a circuit of the type shown above. 12 and 12a - 12d.

The NTSCH signal is generated from the expanded sideband data of the second component signal X of the EDTV signal using a compression factor of 0.22. The NTSCH signal is applied to a luminance-chrominance separation circuit 1440 which separates the luminance (YH) and chrominance components 25 of the NTSCH signal and demodulates the chrominance signal component to obtain two color signal components (IH and QH). The signals YH, IH, and QH are applied to a Y-I-Q format decoder 1444 together with signals YH, IN, and QN formed from the first component signal N by a luminance-chrominance-30 separation circuit 1442. Circuits 1440 and 1442 may be identical; an exemplary circuit is shown in Figure 18.

In Fig. 18, a signal N or a signal NTSCH is applied to a bandpass filter 1810 and a delay element 1812, 35 which compensates for the processing delay through the filter 1810.

The filter 1810 used in this embodiment of the invention is a horizontal-vertical-temporal (H-V-T) bandpass filter. An example circuit for use as an HVT filter 1810 is shown in Figure 10c. This filter includes a horizontal bandpass filter 1030 having a passband of 3 MHz to 4.2 MHz and a VT bandpass filter 1031 defined by the FIR filter shown in Fig. 10 and the coefficient values shown in Fig. 10a. The output signal of filter 1810 is a separated chrominance signal. This signal is applied to the input port of the subtractor of the isolating circuit 1814, and its input port to be subtracted is connected to receive the signal produced by the compensating delay element 1812. The output signal of the separation circuit 1814 is a luminance component signal YN or YH.

The chrominance signal generated by the filter 1810 can be represented as a sequence of sample values I, Q, -I, -Q, I, Q

etc., where I and Q indicate samples of the I and Q color difference signals, and the sample marks indicate the sampling phase, not necessarily the polarity of the sample. This Kro-20 minimum signal is applied to the first and second latch circuits 1815 and 1816. The latch circuit 1815 is drifted by the phase I clock signal ICK produced by the clock generation circuit 1432 of Fig. 14 to maintain sample values of the chrominance signal representing the chrominance signal I-color difference signal. The latch circuit 1816 is controlled by an inverted version of the signal ICK produced by the inverter 1822 to hold sample values representative of the Q color difference signal component of the chrominance signal. The signals produced by latch circuits 1815 and 1816 are applied to respective complement-30 circuits 1818 and 1820. Circuits 1818 and 1820 are controlled by a signal produced by frequency divider 1824 to complement the alternating values of the data-sampled I and Q color difference signals. The signals produced by circuits 1818 and 1820 are demodulated signals IN or IH 35 and QN or QH, respectively.

41 86785

As discussed above, the signals YH, YN, IH, IN, QH, and QN are applied to a Y-I-Q format decoder 1444, where they are combined to form widescreen signals YF'0, IF ', and QF'. An example circuit that can be used as a format-5 decoder 1444 is shown in Fig. 19. In Fig. 19, the first luminance and color difference signals YN, IN, and QN are applied to a sideband-midband discrimination circuit 1940. The circuit 1940, which may include a demultiplexer (not shown) and a picture an element counter (not shown) separates from each line pixel values representing the low frequency components of the sideband signals from samples representing the midband signal. In this embodiment of the invention, samples 1 to 14 and 741 to 754 represent a sideband signal, while samples 15 to 740 represent 15 centerband signals.

Circuit 1940 produces data-sampled signals YO, 10, and QO representing sidebands applied to a time-expanding circuit 1942 that expands the signal over time by a factor of 6 to produce signals YL, IL, and QL representing low-band components of sideband signals returned to their correct time dependence. These signals are summed by the connecting circuit 1946 to form the sideband signals YS, IS and QS returned to the signals YH, IH and QH.

Circuit 1940 also produces data-sampled signals YE, IE, and QE that represent the time-expanded center band of the first component of the EDTV signal. These signals are applied to a time-compressing circuit 1944, which compresses the data-sampled signals over time 30 by a factor of 0.81 to produce center band signals YC, IC and QC.

The recovered side and center band signals are recombined with a 1960 wiper to form widescreen luminance (YF ') and color difference signals (IF' and QF '). 35 A circuit suitable for 1960s is shown in Figure 19a.

42 86735

In Fig. 19a, the threader is shown including a circuit 1910 for threading the luminance center band and sideband signals YC and YL, respectively, to form a widescreen luminance signal YF'0. In addition, Fig. 19a shows I-5 signal floats 1920 and Q signal floats 1930 having the same structure and operation as the described Y-signal floats.

In the coding operation, the midband and sideband signals are intentionally superimposed, for example by ten samples, to compensate for the distortion of the sample values that may occur in the expansion and compression processes at the boundaries of the side and midband regions. If there was no transition area in the lanes, the damaged samples could connect to each other and the junction was visible in the reproduced image. A transition reading of ten samples has always been found to be sufficient to compensate for the five sample values.

In Fig. 19a, the multiplier 1911 multiplies the sideband signal YS by the weighting function W in the transition areas, as shown by the waveform adjacent 20, before the signal Y is applied to the adder 1915. Similarly, the multiplier 1912 multiplies the midband signal YC by the complement weighting function (1 to W) waveform is described before signal YC 25 is applied to adder 1915. These weighting functions have a linear ramp-type characteristic in the transition ranges and have values between 0 and 1. They can be implemented, for example, with a combination of sample counter (not shown) indicating ROM (not shown), containing sample values of 30 to represent weighting functions. The output signal of adder 1915 is a threaded widescreen luminance signal YF'0.

Referring to Fig. 14, the signal produced by the transverse modulator and the amplitude expander 1426 is applied 35 to the time expanding circuit 1430. The time expanding circuit 1430, which may be implemented with a circuit as described above with reference to Figs. 12 and 12a to 12d, expands the wide display of the signal Z. The third component signal of the EDTV signal, to use the entire active video portion of the horizontal-5 line period. The signal produced by the time-expanding circuit 1430 is applied to an amplitude modulator 1432.

The modulator 1432 multiplies the signal produced by the time-expanding circuit 1430 by the signal fc produced by the clock generation circuit 1432 to return the high frequency luminance signal to its original frequency band. The high frequency luminance signal produced by modulator 1434 is applied to a high pass filter 1436 which removes frequencies below 5 MHz. This filter removes the baseband component and all low frequency modulation components from the signal produced by the modu-15 generator 1432. The output signal of the high-pass filter 1436 is applied to one input of the adder 1436 and its other input is connected to receive the signal YF'0. The adder 1436 combines the high frequency components (between 5.0 MHz and 6.0 MHz) of the snow-20 nance signal with the widescreen luminance signal YF'0 to form a wideband widescreen luminance signal YF '.

The auxiliary signal YT is recovered from the signal YTN produced by the input unit 1422 by applying the signal YTN to the format decoder 1460 and the core circuit 1458. The format decoder 1460 expands the center band portions and compresses the center band portions of the signal YTN to re-generate the tuned auxiliary signal 1458. Circuit 30 1458 switches the converted auxiliary signal with values between -10 IRE units and +10 IRE units to 0 IRE units. In addition, circuit 1458 subtracts 10 IRE units from the magnitude of all sample values of the signal YTN that are greater than 10 IRE units. This reverses the level shift performed by the encoder 35 and eliminates any low-amplitude scientific interference (less than 10 IRE) from the converted auxiliary signal. The output signal of the core circuit 1458 is applied from interleaved to progressive scan to a converter 1450, which is described below with reference to Fig. 21.

The signals YF ', IF' and QF 'are applied to the long progressive scan converters 1450, 1452 and 1454, respectively. Figures 20 and 21 are block diagrams of exemplary scan converters 1452 (or 1454) and 1450, respectively. The sweep converter shown in Fig. 20 includes an adder 2014 and a dividing circuit 2016 that take the average of the input signal (i.e., IF 'or QF') with the signal A provided by the delay element 2012 to produce a signal X. Signal A is delayed by one frame period relative to signal B. The signal X and signal C produced by the delay element 2010 15 are applied to the respective dual port memories 2018 and 2020. Sample values are written to memories 2018 and 2020 under signal CK4 and read from memories 2018 and 2020 under signal CK8. The output signals of memories 2018 and 2020 are applied to the respective input ports of multiplex-20 ser 2022. The lines of signal X are alternated by multiplexer 2022 with lines of signal C 525 LPF to generate a progressive scan signal IF or QF.

The scan converter shown in Fig. 21 includes all the circuits shown in Fig. 20 and further includes an adder 2118 which sums the reconstructed auxiliary signal YT to the frame average signal X to form the signal X '. As discussed above, the signal YT represents the difference value between the approximations of the frame average of the sample value lines and the actual sample values. Thus, the signal X 'represents intervening sample lines whose errors have been corrected in the averaging process used to generate the samples. The output signal of the scan converter shown in Fig. 21 is 525 LPF progressive scan sig-35 signal YF.

45 86785

Referring to Figure 14, signals YF, IF, and QF are applied to DAC circuit 1462, which generates the corresponding analog signals Y ', 1', and Q '. As shown in Fig. 1, the signals Y ', 1' and Q 'are subtracted from the wideband 1050 LPF interlaced scan signals Y, I and Q to form the wideband difference signals ΔΥ, ΔΙ and Δ (2). These wideband signals are applied to the auxiliary channel encoder 142.

As shown above with reference to Fig. 1, the wide-band separation signal ΔΥ has a frequency spectrum ranging from 0 Hz to 20 MHz, and the signals ΔΙ and Δ0 have a frequency spectrum ranging from 0 Hz to 10 MHz. The auxiliary channel encoder 142 encodes these signals into an AUX signal with a 6 MHz bandwidth. In addition, the signal AUX may include a digital audio signal. Encoder 142 uses a separate circuit to encode the luminance separation signal ΔΥ and the two color difference signals ΔΙ and AQ. The circuit for encoding ΔΥ: η is shown in Fig. 22, and the circuit for encoding ΔΙ: η and Δ (2: η is shown in Fig. 24).

20 In Figure 22, the signal ΔΥ is divided into components "A", "B" and "C" by a 0-6 MHz low pass filter 2244, a 6-12 MHz bandpass filter 2246 and a 12-18 MHz bandpass filter 2248. Frequencies between 18 MHz and 20 MHz contain little energy and are not transmitted. Thus, component A has a frequency spectrum of 0 to 6 MHz, component B has a frequency spectrum of 6 to 12 MHz, and component C has a frequency spectrum of 12 to 18 MHz. Component A is connected to one input of switch SI.

Components B and C are downconverted by respective se-30 mixer and low pass filter circuits 2250 and 2252 to use a frequency band between 0 Hz and 6 MHz. The line pairs of the downconverted components B and C are averaged by the respective line average circuits 2254 and 2256. The average line lines are connected to the various input terminals of switch S2, which operates at 35 15 kHz (half of the 32 kHz horizontal line speed), respectively. The output signal of the switch S2 includes the average lines of the B-component signal alternately with the average lines of the C-component signal. The averaging and shifting of the B and C component lines reduces the vertical bandwidth of the horizontal high frequency components represented by the B and C component signals. Since the alternating lines of the output signal of the switch S2 represent luminance information in the 6-12 MHz and 12-18 MHz frequency bands, this function 10 efficiently compresses the information from the 6-18 MHz frequency band to the 0-6 MHz frequency band. The output terminal of switch S2 is connected to the second input terminal of switch S1.

The switch SI is controlled by the switching signal produced by the OR gate 2258. The input signals of the OR gate 2258 are a 15 Hz 15 (half of 30 Hz frame rate) square wave signal and a motion detection signal produced by the motion detector 2260. In the absence of motion in the image produced by the widescreen HDTV source 110, i. when the image remains in place, the switch S1 switches on each frame, producing the field sequence-20 shown in Fig. 22a. However, when there is movement in the figure, the switch SI is controlled to give only component A.

Although many motion detectors are known, the converted auxiliary signal (the four-to-25 component component of the widescreen EDTV signal described above) is preferably used to generate a motion indicating signal, as described below. The converted auxiliary signal may be input to the auxiliary channel encoder 142 by a single channel decoder 140, as shown in Figure 10. This signal is produced in the manner described in Figs.

Figure 28 illustrates the interface between a single channel decoder 140 and an additional channel encoder that produces an auxiliary signal. In Fig. 28, the converted auxiliary signal provided by the format decoder 1460 is applied to the core circuit 1458, as shown in Fig. 14, and to the motion detector 2260. At values of the converted auxiliary signal between 10 IRE units and -10 IRE units, the signal produced by the detector shows no movement. at values outside the range, the signal generated by the sensor indicates that there is intra-frame motion in the image being processed.

Fig. 24 is a block diagram of a circuit that can be used to encode the difference signals ΔΙ and AQ. The circuit shown in Fig. 24 reduces the horizontal and vertical bandwidth of both signals ΔΙ and AQ, multi-plexes the time-limited bandwidth ΔΙ and Δθ signals, and then compresses the obtained signal with respect to time. The time-compressed signal is divided into high and low frequency components, which are produced alternately at field rate to form the signal CA. The signal CA and the signal YA produced by the auxiliary channel luminance signal encoder are time division multiplexed to produce a video signal having a multiplexed component format, as illustrated in Fig. 24a.

More specifically, in Figure 24, the signals ΔΙ and Δ0 20 are applied to respective 2.4 MHz low pass filters 2410 and 2416, which reduce the separation of the ΔΙ and ΔΟ signals. The band-limited signals produced by these filters are applied to the respective line average circuits 2412 and 2418. The circuits 2412 and 2418 take an average corresponding to the sample values from the successive lines ΔΙ and AQ of the respective signals. The signals produced by the averaging circuits 2412 and 2418 are applied to the respective input ports of the line rate multiplexer 2414. The line rate multiplexer 2414 produces line averages of the ΔΙ and Δθ signals during alternating line periods 30 to compress both ΔΙ and AQ signals into a single 2.4 MHz signal. The signal produced by the multiplexer 2414 to the compressor in terms of time by a factor of 5 in the time compressing circuit 2422. The output signal of the circuit 2422 is a 12 MHz signal (2.4 x 5 = 12) for each horizontal drink-35 cycle (910 sample cycles during 150 sample cycles. This 48 86735 signal is divided into two to a component having the respective frequency spectra of 0-6 MHz and 6-12 MHz by a low-pass filter 2424 and a bandpass filter 2426, respectively.The signal produced by the bandpass filter 2426 is downconverted to 0-6 MHz by a mixer 2428, the output signal of which is applied to a field delay element 2430. The signals produced by the low pass filter 2424 and the field delay element 2430 are applied to the respective input ports of the field rate multiplexer 2432. The multiplexer 2432 alternately produces 10 signals input to its input ports during successive field cycles as a signal CA. Although not shown, it is contemplated that 2430 vol can be replaced by a frame delay element, and that the multiplexer 2432 can be controlled by a motion signal 15a, such as the signal produced by the OR gate 2258 of FIG. In this case, the high and low frequency components of the signals ΔΙ and AQ would be transmitted on a frame basis from the stationary areas of the image, and only the low frequency components of the signals ΔΙ and AQ would be transmitted from the moving areas of Fig. 20. This change would adjust the temporal separation of the signals ΔΙ and AQ to match the separation of the signal ΔΥ.

The signal CA and the signal YA produced by the circuit illustrated in Fig. 22 are applied to the different input ports of the luminance / chromium-25 nancial multiplexer 2434, respectively. The multiplexer 2434 passes the signal YA in the first 755 sample periods in each active line period, the signal CA in the next 150 sample periods, and the blackout signal in the last 5 sample periods. The digital audio signal 30 shown in Fig. 1 can be input by a device (not shown) during the vertical blackout period of the additional video signal, and included in the signal YA / CA.

Fig. 25 is a block diagram of RF modulation and demodulation circuits suitable for use in the system shown in Fig. 155. In this circuit, a 6 MHz signal 49 86785 YA / CA is applied to a line demultiplexer 2510 which produces from the even numbered line periods of the signal YA / CA to the time expansion circuit 2512 and the odd numbered line periods to the time expansion circuit 2516. Circuits 2512 and 2516 expose their so that each line of the input signal corresponds to two lines of the output signal and uses a frequency band between 0 Hz and 3 MHz. The signals produced by the time expansion circuits 2512 and 2516 are applied to a modulator 10 2514, which can either double-sideband modulate (DSM) a pair of 90 ° phase-shifted carrier signals with two input signals, or double the single-sided single-band modulated (DSSM) single-carrier signal. and a lower sideband for the second signal. The signal produced by modulator 2514 produces little interaction with the parallel channels in existing terrestrial television signal transmission signals because the carrier is located in the center of the transmission spectrum, and any interaction present is not as coherent as that produced by the interfering conventional television signal. The output signal of the modulator 2514 is an AUX signal which is transmitted via antenna 130 to a dual channel decoder 134. Decoder 134 receives the AUX signal via antenna 132.

The RF demodulator 2522 in the two-channel decoder 134 demodulates said two time-expanded signals and outputs the results to the time-compressing circuits 2524 and 2526. The circuits 2524 and 2526 compress their respective input signals with a time factor of half-30 and provide the obtained signals to the line multiplex. 2528. The multiplexer 2528 alternately produces signals input to its two input ports during alternating horizontal line periods to recover the signal YA / CA.

Fig. 29 is a block diagram of an exemplary two- to 35-channel decoder 134. In Fig. 29, antenna 32 is connected to two tuners 2902 and 2522, which tune the respective RF channels carrying the transmitted main and auxiliary signals. The single channel decoder 2904 de-encodes the signal produced by the tuner 2902 to produce component signals Y ', 1' and Q '. Tuner 2902 and decoder 2904 may be similar to the single channel decoder described above with reference to Figures 14-21. The auxiliary signal decoder 2906 expands the 6 MHz auxiliary signal produced by the tuner 2522 to produce an 18 MHz complementary luminance signal-10 to aY and 2.4 MHz to supplement the color difference signals ΔΙ and AQ. The additional signal decoder is described below with reference to Figs. 26 and 27. The signals ΔΥ, ΔΙ and AQ are combined with the 525 LPF progressive scan signals Y ', 1' and Q 'produced by the single channel encoder 2904, corresponding to 15 wool summers 2908, 2910 and 2912. 1050 L / F interlaced scan signals Y ", I" and Q "would be generated and applied to a widescreen HDTV display 136. The display 136 may be a conventional 525 L / F progressive scan display controlled by a field rate signal to move 525 20 displayed lines at half the line spacing from field to field 1050 LPF to provide an interlaced sweep display.

The circuit shown in Fig. 26 includes a luminance / chrominance demultiplexer 2610 that separates 150 chrominance samples CA, 755 luminance samples Y from each horizontal line period of the signal YA / CA produced by the multiplexer 2528 of Fig. 25. This circuit inversely performs the multiplexing operation performed by the multiplexer 2434 of Figure 24 described above. Samples YA process-30 are played with the circuit described below with reference to Fig. 27. The samples CA are applied to a field delay element 2612. The field delayed signals produced by the element 2612 are upconverted by a mixer and a bandpass filter 2614 to use a frequency band between 6 MHz and 12 MHz. The signal produced by mixer 35 and bandpass filter 2614 is summed 86735 51 by adder 2616 to the low frequency chrominance signals produced by demultiplexer 2610. The output signal of adder 2616 is applied to the second terminal of switch 2620 and its second terminal is connected to receive a field delayed version of the signal produced by adder 2616 via field delay element 2618.

The 15 Hz signal directs switch 2620 to pass the output signal produced by adder 2616 during frame periods in which the signal CA represents low frequency chromium 10 signals, and to pass the signal produced by delay element 2618 during frame periods in which signal CA represents high frequency chrominance signals. Alternatively, when motion adaptive signal CA multiplexing is used in the coding circuit, the circuits 15 shown in Fig. 26 may include motion adaptive demultiplexing circuits of the type described below with reference to Fig. 27. The signal produced by switch 2620 is expanded with time by a factor of 5 in circuit 2622, resulting in 750 signals of CA per horizontal line period. This operation 20 reduces the frequency spectrum of the signal CA from 0 to 2.4 MHz. The time-expanded signal CA is applied to a line-rate demultiplexer 2624, which, during alternating horizontal line periods, produces lines of samples from the reconstructed signal A1 'to one output port and 25 lines of samples from the reconstructed signal AQ' to the other output port. The signals ΔΙ 'and AQ * are applied to respective horizontal line interpolators 2626 and 2628, which form supporting sample lines from existing sample lines to produce signals ΔΙ and AQ, each having 525 sample lines per field period and having a horizontal frequency spectrum of 0 MHz.

An example circuit for use as a luminance replenishment signal decoder is shown in Fig. 27. In Fig. 27, the signal YA is applied to one input port of the multiplexer 2710, and its output port is connected to the input port of the frame delay element 52 8 6 7 8 5 2714. The output port of the frame delay element 2714 is connected to the second input port of the multiplexer 2710. The control signal of the multiplexer 2710 is a signal H / L which is logical from the OR motion signal MOTION generated by the motion sensor 2711 and the 15 Hz square wave signal. This signal changes state at the frame rate. The motion detector 2711 may be similar to the detector 2260 described above with reference to Figures 22 and 28. The multiplexer 2710 is controlled by the control signal H / L to pass the signal YA during alternating frames when no motion occurs or during each frame in moving parts of the image. In intermediate frames of stationary images, the control signal H / L controls the multiplexer 2710 to circulate the signals produced by the frame delay element 2714. The output signal of the frame delay element 2714 15 is a component signal of the low frequency (0 Hz to 6 MHz) signal ΔΥ. This output signal is applied to one input port of adder 2716. The second input port of adder 2716 is connected to receive the high frequency components (6 MHz to 18 MHz) of the signal ΔΥ with stationary 20 images and zero signals with moving images. In the stationary parts of the image, the horizontal frequency spectrum of the signal produced by the adder 2716 is 0 Hz to 18 MHz, but the temporal update period is only 1/15 s, and in the moving parts of the image, the horizontal frequency spectrum is 25 Hz to 6 MHz and the temporal update period is 1/30 s.

To generate the high frequency components of the signal ΔΥ, the signal YA produced by the luminance / chrominance demultiplexer 2610 of Fig. 26 is applied to the demultiplexer 30 2718. The demultiplexer 2718 is controlled by a square wave signal having a frequency of 15 kHz (half 1050 LPF sample lines for mixer and bandpass filter circuits 35 2720 and 2722.

53 86785

In turn, circuits 2720 and 2722 perform the downconversion of the components of the signal ΔΥ 6 to 12 MHz and 12 to 18 MHz performed by the additional channel encoder 142. The mixer and bandpass filter 2720 modulates the 6 MHz carrier signal 5 with signals input to its input port and passes the obtained signals through a bandpass filter (not shown) having a passband of 6 MHz to 12 MHz. The filter removes the baseband signal and any interfering modulation components. Similarly, mixer and filter 2722 modulates a 12 MHz 10 carrier signal with signals input to its input port and bandpass filters the resulting modulated signals to use the frequency band between 12 MHz and 18 MHz.

The signals produced by the mixer and the bandpass filter 2720 are applied to one input port 15 and the 1H delay element 2724 of the multiplexer 2728. The delay element 2724 is connected to provide its output signal to the other input port of the multiplexer 2728. The multiplexer 2728 is controlled by the 15 kHz line rate switching signal alternately the output signal of the circuit 2720 and the output signal of the delay element 2724 during the respective alternating horizontal line periods. The 1H delay element 2762 and the multiplexer 2730 are configured to similarly alternately output the output signal of the circuit 2722 and the output signal of the delay element 2726 during the respective alternating horizontal line-25 periods. The line rate signals applied to the multiplexers 2728 and 2730 are synchronized to the signals produced by the encoder circuits 142 such that in stationary images, the multiplexer 2728 produces line average signals ΔΥ using the 6 MHz to 12 MHz frequency band, and the mul-30 diplexer 2730 produces line average signals . In moving parts of the image and in alternating frame periods of stationary images, the signals produced by the multiplexers 2728 and 2730 are not valid.

86735 54

The output signals of the multiplexers 2728 and 2730 are summed by an adder 2732 to produce, during periodic periods of the image, a signal representing the components of the ΔΥ signal in the 6 MHz to 18 MHz frequency band.

5 This signal is applied to one input port of multiplexer 2734. The second input port of the multiplexer 2734 is connected to receive a zero signal produced by a zero signal source 2736. The multiplexer 2734 is controlled by a signal H / L produced by the OR gate 10 2712 to pass the signal produced by the adder 2732 in alternating frames in stationary portions of the image. , and pass the zero signal produced by source 2736, otherwise.

The output signal of multiplexer 2734 is applied to a delay element 2740 of one frame (1050 H) at one input port of multiplexer 2738. The second input port of the multiplexer 2738 is connected to receive the frame delayed signal produced by the delay element 2740. The multiplexer 2738 is controlled by the control signal 20 produced by the AND gate 2742 to pass the stationary image produced by the multiplexer 2734 in alternating frames (i.e., when the signal produced by the multiplexer 2734 includes high frequency components of line averages) and during moving parts of the image (i.e., by the multiplexer 2734). zero). Only during frame periods between stationary images, when the signal YA contains only low frequency components, the multiplexer 2734 is controlled to pass the signal li produced by the delay element 2740. In this configuration, the multiplexer 2738 passes a signal of zero zero with moving parts of the image, and a high frequency ΔΥ signal with fixed parts of the image. The control signal of the multiplexer 2738 is a logic AND 15 Hz of the frame rate switching signal and the complement of the signal MOTION produced by the inverter 2744.

ss 86735

The output signal of the multiplexer 2738 is applied to the second input port of the adder 2716 to produce a reconstructed signal ΔΥ. As shown above, the frequency spectrum of the signal ΔΥ is 0 Hz to 18 MHz and the temporal refresh period is 5/15 seconds for stationary parts of the image, and the frequency spectrum is 0 to 6 MHz and the temporal refresh period is 1/30 second for the moving parts of the image.

In the described embodiment of the invention, the high and low frequency components of the signal ΔΥ are multiplexed 10 motion-adaptively on a frame basis. It is contemplated that these signals may be multiplexed on a field basis or on the basis of a number of field periods larger than the frame period.

Although not shown in the drawings, it is contemplated that the dual channel decoder 134 may include timing signal generating circuits of the same type as the timing circuits 1432 described above with reference to Figure 14, the 16xfsc clock signal used by the decoder 134, the 6 MHz and 12 MHz data sampled carrier signals, and frame-20 rate to generate line rate control signals. To ensure proper reconstruction of the picture, it may be desirable to synchronize encoder 142 to decoder 136 by including a timing reference signal, such as a semi-random sequence, in either or both of the MAIN and AUX signals. An example circuit for generating and decoding such a timing reference signal is described in the aforementioned U.S. Patent Application 241,277 and U.S. Patent 4,309,712, entitled "Error Coding for Video Disc System", which are incorporated herein by reference. Alternatively, the signal AUX may include a modified synchronization signal, such as that used in a television system proposed by NHK, a Japanese broadcaster described in R. Hopkins' article entitled "Advanced Television Systems", IEEE Transactions on Consumer Electro-35 nics, February 1988, ss. 4 to 5. Exact timing information 56 86785 can be obtained from the zero bypass points of this synchronization signal.

As shown with reference to Fig. 29, the signal ΔΥ produced by the luminance expansion signal decoder, and the signals ΔΙ and Δ0 produced by the chromium-5 energy complement signal decoder are combined with the respective signals Y ', 1' and Q 'to produce widescreen HDTV signals Y ", I" and Q ". for display on a widescreen HDTV display 136. An image produced under the control of signals Y ", I" and Q "includes substantially full HDTV resolution in stationary areas and a higher resolution than a standard NTSC image, but smaller than an HDTV image, in moving areas .

Claims (11)

57 86785 A system for generating a coded main and auxiliary signal representing a high resolution image, comprising: a video signal source for producing a high resolution video signal; a first signal encoding device coupled to said source for generating said encoded main video signal compatible for reception and display in a conventional television receiver; a second signal encoding device associated with said source for producing an encoded auxiliary video signal suitable for combining with said main video signal in a special receiver to produce a widescreen display; and a transmission device including two transmission channels, one channel carrying said encoded main video signal and the other channel carrying said additional coded video signal; characterized in that: said main video signal generated by said first encoding device (114) includes widescreen display information and represents an enhanced picture with a resolution greater than a conventional video picture but smaller than said high resolution picture, a decoding device (140) coupled to said first signal encoding device to decode the encoded main video signal to obtain a decoded main video signal; a signal separation device coupled to said decoding device to generate a difference signal representing a difference between said incoming signal produced by the source 35 (110) and said decoded main video signal produced by said decoding device; and said second signal encoding device (142) is coupled to said signal separation device to generate said encoded additional video signal representing said difference signal. The system of claim 1, characterized in that: said separation signal generated by said signal separation device and encoded by said second signal encoding device (142) includes both luminance and chrominance components; and said encoded main video signal produced by said first signal encoding device (114) is form-15 interleaved and its line and field rates correspond to a given television transmission standard selected from the group consisting of NTSC, PAL and SECAM. The system of claim 1, further characterized in that: said incoming video signal uses a frequency band from 0 Hz to L MHz; said first signal encoding device includes a low pass filter device (19a) for processing said incoming video signal to produce a filtered video signal 25 using a frequency band 0 Hz to K MHz, where K is less than L; said signal separation device produces said separation signal using the frequency band 0 Hz to L MHz; and said second signal encoding device comprises: means (2244, 2246, 2248) for separating said difference signal into a first component signal (A) using the frequency band 0 Hz to J MHz and a second component signal using the frequency band J 59 8 6 7 3 5 MHz - H MHz, where J is less than H and H is not greater than L; means (2250, 2252) for encoding said second component signal to form a coded second component signal using the frequency band 0 Hz to J MHz; and means (SI) for time division multiplexing said first component signal and said encoded second component signal to generate said additional coded video signal using the frequency band 0 Hz to J MHz. The system of claim 1, further characterized in that: said incoming video signal is an N line / frame 15 interlaced scan signal, wherein N is an integer; said first signal encoding apparatus includes means (16) for converting said incoming video signal M line / frame to a progressive scan intermediate video signal, and means (114) for encoding said intermediate video signal to form said main video signal which is M line / frame a frame interlaced scan signal, wherein M is an integer not greater than N; said decoding apparatus includes means (1450) for processing said encoded main video signal to form said decoded main video signal, which is an M line / frame progressive scan signal; and said separating device includes means for subtracting said decoded main video signal from said incoming video signal to form said separating signal having 2M lines per frame. A television signal processing system comprising: first and second receiving devices for producing, respectively, an encoded main video signal obtained from a first analog transmission channel and an encoded auxiliary video signal obtained from a second analog transmission channel, said both video signals being obtained from a high-precision source; the encoded main video signal being compatible for reception and display in conventional television receivers having a constant aspect ratio; a signal processing device 10 for decoding and combining said received video signals to produce an enhanced resolution video output signal; characterized in that: said encoded main video signal produced by said first designated receiving device 15 (2902) contains widescreen information and represents an extended resolution television image having a higher level of resolution than a conventional television image; said additional signal represents the difference between a high precision widescreen signal from which said encoded main video signal is obtained and a second signal decoded from said encoded main video signal; and wherein said signal processing device includes: a main signal processing device (2904) coupled to said first receiving device and controlled by said encoded main video signal to produce a baseband decoded main video output signal; an additional signal processing device (2906) coupled to said second receiving device (2522) and controlled by said encoded additional video signal to produce a baseband supplemental output signal; and combining means (2908, 2910, 2912), the first and second inputs of which are connected to respective main (2904) and auxiliary video signal processing devices, respectively, for combining said baseband decoded main video signal and said baseband supplement signal by summing, to generate a video output signal representing a widescreen video image with a resolution greater than the resolution of said extended resolution television image. The television signal processing system of claim 5, further characterized in that: said additional signal includes a first and a second component signal, said first component signal representing a first complement signal for said main video signal using the first frequency band, said second component signal representing said second complement signal for a main video signal, and using a second frequency band, said second component being frequency shifted for transmission on said second frequency channel using the same frequency band as said first component, said components being time division multiplexed on said second channel, and wherein said additional video signal processing device includes to separate said first and second component signals therefrom; means (2720-2744) for decoding said second component signal; and means (2716) for combining said first component signal and said second component signal to form said baseband supplement signal. The television signal processing system of claim 6, further characterized in that: the first component signal (A) of said additional video signal includes a signal using the frequency band 0 Hz to N MHz, which represents a complement to said main video signal in the frequency band 0 Hz - N MHz; Said second component signal (B or C) of said additional video signal includes signals using the frequency band 0 Hz to N MHz, which represents a complement to said main video signal in the frequency band N MHz to P MHz; wherein P is greater than N; Said first and second component signals of said additional video signal are time division multiplexed at a given rate G fields per second, where G is an integer, for said fixed resolution television picture stationary sequences, and said additional video signal contains only said first component signal for said expanded video; and wherein said additional signal processing device includes: a device (2711) coupled to said main video signal processing device for generating a motion signal indicative of moving sequences of said extended resolution television image; demultiplexing devices (2710, 2718) for separating said first and second component signals from said additional signal; a frequency conversion device (2720 or 2722) coupled to said demultiplexer for processing said second component signal to form a frequency shifted second component signal using said frequency band N MHz to P MHz; a device (2714) coupled to said demultiplexer (2710) and including a signal storage device (2714) controlled by the presence of said motion signal to selectively reproduce said first field signal of the first component signal G during alternating periods 63 63 6 6 7 8 5, the first component signal to form; a device (2738) including a signal storage device (2740) coupled to said demultiplexer 5 and said frequency conversion device and controlled by the presence of said motion signal to selectively reproduce said frequency shifted second component signal during said alternating field periods of said G field period 10; a continuous frequency shifted second component signal; and means (2716) for combining said continuous first component signal and said continuous frequency shifted second component signal to form said baseband supplement signal. The television signal processing system according to claim 7, further characterized in that said G field periods correspond to one frame period. 9. A television signal according to claim 8 20 Iin processing system, further characterized in that: said second component signal of said additional signal includes first and second subcomponent signals (B, C) each using a frequency band of 25 0 Hz to N MHz, said first subcomponent signal (B) representing complementation of said main video signal in the frequency band N MHz to Q MHz, and said second subcomponent signal (C) representing supplementation of the main component signal in the frequency band Q MHz to P MHz, wherein said first and second subcomponent signals are time division multiplexed on a line basis; and said frequency conversion device includes: a device (2720) for modulating an N MHz carrier signal. . with said first subcomponent signal to form a frequency shifted first 64,867,85 subcomponent signal using the frequency band N MHz to Q MHz; means (2722) for modulating the Q MHz carrier signal with said second subcomponent signal to generate a frequency shifted second subcomponent signal using the frequency band Q MHz to M MHz; apparatus (2724, 2728) including a signal storage device for alternately reproducing a line of samples of said frequency shifted first subcomponent signal 10 to form a continuous frequency shifted first subcomponent signal; apparatus (2726, 2730) including a signal storage device for alternately reproducing a line of samples of said frequency shifted second subcomponent signal to form a continuous frequency shifted second subcomponent signal; and means (2732-2740) for combining said continuous frequency shifted first and second subcomponent signals to form said frequency-shifted second component signal. 65 8 6 7 8 5