DE3240905A1 - DIGITAL SIGNAL DISCONNECT - Google Patents

Description

-S-

RCA 77465 / Sch / Ro.
US Ser. No. 319.109
AT: November 6, 1981

RCA Corporation, New York , NY (V.St.A.)

'' Digital signal separation circuit

The invention relates to digital signal separation circuits and, more particularly, relates to digital filter circuits, which are both band-pass filtered and low-pass filtered Deliver digital signals.

In television receivers in which the baseband video signal is processed in digital form, it is often necessary to separate signals that are in adjacent frequency bands lie. If, for example, a video signal is filtered over the full video bandwidth in order to achieve the intermingling to separate interleaved luminance and color signal components then remains in the low frequency band at the comb filter output, usually referred to as color output, some luminance information is left over. This low-frequency luminance information, the so-called vertical detail information, must then be extracted from the

BfiD ORIGINAL

comb-filtered color signal, which arises at this output, be separated out and again with the comb-filtered luminance signal to a fully restored one Luminance signal are combined. With the NTSC television system, the color signal extends up to around 1.5 MHz below the 3.58 MHz color subcarrier frequency, and for a normal scene is essentially all of it Vertical detail information in the lower 1.0 MHz of the am Color output contained the resulting signal. The luminance and color signal components in the output of the Luminance channel can then switch between the upper frequency of the vertical detail signal, 1.0 MHz, and the lower frequency Frequency of the color signals can be separated by about 2.1 MHz.

An arrangement for separating the vertical detail information of the luminance information at the output of the comb filter for the color signal is found in US Pat. No. 4,096,516. The video signal is hereby a scanned data system with a CCD comb filter (charge-coupled construction element) comb-filtered. The comb-filtered color signal at the output of the color comb filter is separated by two filters, which are each constructed with discrete frequency-selective filter elements. A low pass filter with a bandwidth of 0 to 1.5 MHz separates the vertical detail information from the color information and couples the vertical detail information in the luminance signal at the output of the luminance comb filter. A band pass filter couples the band-filtered color signals from the output of the color comb filter to a color signal processing circuit.

In a digital television receiver, are wherein Leucht'-density and color signals by a digital comb filter separated from each other, it is also desirable that vertical detail information from separate out the color information in order comb filtered with the

Luminance signal to reunite when the. Video signal Comb-filtered over the entire video frequency band will. In order to minimize the number of circuit elements required to carry out this separation are needed, it is desirable to use a single filter that has a low pass filter output for the Has vertical detail information and a band pass filter output for the color signal.

According to the invention, a digital filter is now provided in which a plurality of delayed and weighted taps Signals are combined at a first output, for which a first amplitude-frequency response results. The filter contains a shift register with a predetermined delay. Unweighted signals that are delayed by at least a value of half the specified delay, are included in the first Output occurring signals combined to a signal at a second output, at which a second Amplitude frequency response results. In a first embodiment of the invention, the shift register is one FIR filters (finite impulse response filter) an output tapped Shift register in which the tap signals are weighted and combined in a so-called adding tree will. A band filter characteristic is obtained at the output of this addition tree. Signals from this output will be then with signals that have been taken from the center tap, subtractively combined to form signals at one second output for which a low-pass characteristic applies.

In a second embodiment of the invention weighted input signals are supplied to taps of an input tapped shift register, which has a has the first output for which a band filter characteristic applies.

Input signals that differ by the same length of time

BATH ORIGINAL

over the first output are delayed like signals at the Center tapping will be created with the first exit Signals are subtractively combined to form signals at a second output for which a low-pass characteristic applies. ■

Show in the accompanying drawings

Fig. 1 shows the digital video signal processing part of a Television receiver with a signal separation circuit according to the invention,

2 shows an output tapped FIR filter according to the invention,

FIG. 3 shows, as a block diagram, a more precise representation of part of the FIR filter according to FIG. 2,

4 and 5 characteristic curve of the FIR filter according to FIGS. 2 and 3 and
20th

6 shows an input-tapped FIR filter as a block diagram according to the invention.

According to FIG. 1, the part of the television receiver constructed according to the invention is for processing the baseband digital signal illustrated as a block diagram. Video signals from a video signal source 10, e.g. the video detector can be in a television receiver, an analog / digital converter (hereinafter referred to as A / D converter designated) 12 is supplied, which converts the analog video signals into digital signals, for example in Converts the form of consecutive 8-bit words. the Digital signals come to the input of a digital comb filter 14, which the signals in the usual way as Luminance signals Y and color signals C separate, though

-ιοί the color signal also contains low frequency (Voillkal ~ Deiall-iH <] nnlo). The digital comb filter 14 can be constructed so that au in the Ar! works, win μη In d »in ArI I Iu-I" I) ImII al To lev J a ion Γιπ «μ <· KnhancnnitMil. 'j by .John P. KohbI, 84 KMP ¹ I ¹ J' :, ^r j4Ί -51 (1074) bcHchrJeben J ti L. The separated Y signal is given via a delay element 16 to an input of an adder 30. The delay τ of the delay element 16 is chosen so that it essentially matches the delay which occurs in a The vertical detail signal occurs when it is processed by a separating circuit 20 which is located between the output C of the camera filter 14 and a second input of the adder 30. The adder 30 combines the vertical detail signal with the comb-shaped light bulb signal u ssui i * I mini wlfciderhtT'ioiil ο I Ltcui!, ouchtdichtoss I gnal. Da »wlndoiiiercjeHto'l ΊI ο Mhfnnl Y arrives at öiruun WJngang el now AcldJerorfl 12, where nine winto Ikalanhebunqu» i qrial component lu-irfjofit oil Il nn Hlqnal Y is added. Daa h l.tiraus rctui'l tlcronde attached signal Y we d fed to a Leuchtdlchteeignal processing switch 40, which is ao controllable that the brightness and contrast of the luminance signal can be changed. The processed lighting density signal is converted back into analog form with the aid of a digital / analog converter (D / A converter) 46. and the analog luminance ^{signal Y 1} is then applied to a tüngamf of a matrix 60.

The location in which the isolating circuit 20 must be opened is <in the Farbauagang C do η KainmJ'iltors 14 connected. Color signal C, which also contains vertical detail signal components, is fed to inputs via a color bandpass filter 22 and a delay element 24. The color bandpass filter 22 supplies a bandpass-filtered color signal which is fed to the input of a color signal processing circuit and to an input of a subtracting circuit 26

BATH

is- The delay element 24 delays the color signal by a delay duration τ, which is chosen so that it matches the delay of the bandpass filter 22. the Subtracting circuit 26 subtracts the comb filtered and band-pass filtered color signal from the comb filtered delayed color signal, and thereby arises through cancellation the color information is a vertical detail signal at the exit. The vertical detail signal is fed to adder 30 and the input of a non-linear detail signal processing circuit 34 supplied, which has a non-linear transfer function, as shown in Fig. 1 and on the Pages 12-15 of the article "A CCD Comb Filter for Color TV Receiver Picture Enhancement "is explained by D.H. Pritchard, in Volume 41 of the RCA Review on page 3 ff of March 1980 is described »The non-linear signal processing circuit 34 removes low amplitude signals, emphasizes medium amplitude signals and attenuates signals of high amplitudes. The circuit 34 can contain, for example, a RAM memory, and the The transfer function depends on the RAM memory under the control of a computer (not shown) saved data. During inactive video intervals, such as during the vertical blanking interval into RAM memory under the control of a computer via read / write control line 36 and RAM address lines 38 new data can be saved. The processed vertical detail signal is sent to the adder as Vertical increase signal supplied.

The color signal processing circuit 50 may contain a color signal enhancement circuit and a color signal composite demoduJaf-.nr as it is described in the UiS Patonfcanmoldun «! - OSSN 297,556 {title; ¹¹ DICITAr, COLOR TEIiIOVTSlON SIGNA !, DEMODULATOR ", registration date: August 31, 1981). The» Farbßicjnn 1 -vorarbn Π circumBBchai I unq '»0 orzt.'injl loflo Fill I) Mh (I ι.ί l'HMii I Μ · Ί in, wIp r »twn) · 'ηι IhI I fi> t enz-

BATH ORIGINAL

Signals (B-Y) and (R-Y) or I and Q. The mixed color signals are converted into analog form with the aid of D / A converters 52 and 54, and the now analog color signal mixtures are fed to the matrix 60. The matrix 60 combines the mixed color signals with the luminance signal for generating red, green and blue color signals for a picture tube (not shown).

In the arrangement according to FIG. 1, the separating circuit 20 serves to separate the vertical detail information which is shown in the low-frequency component of the signal supplied by the color comb filter is contained by the color information that is contained in the high-frequency part of the signal supplied by the color comb filter. Used as a color bandpass filter a digital FIR filter, then you need because of that linear phase response of the filter only a single bandpass filter for this separation. The bandpass filtered Signal can then be subtracted from the full bandwidth delayed signal to cancel the color information practically without residual components which cause point ripples on the edges of the restored color signal could, and thus only the vertical detail information remains for subsequent reinsertion into the comb-filtered luminance signal.

FIG. 2 shows an isolating circuit which is suitable for the arrangement according to FIG. 1 and which is constructed according to the invention is. It contains a digital FIR filter with a tapped shift register 100, weight factor circuits 102-118 and an addable circuit 120-130. At this and the following embodiments of the invention the broad arrows in the drawings represent parallel lines for digital information, over which Digital words made up of several bits are transmitted from one element to another.

BATH ORIGINAL

Comb-filtered color signals in the form of, for example 8-bit words are fed to the first stage of a shift register 100 as shown in FIG. Each stage of the shift register 100 can temporarily control and pass on a word of the color signal under the control of a clock signal. Thus, each stage of shift register 100 numbered 1-21 can have 8 bits at the same time to save. The delay of the shift register 100 is a product of the number of stages and the cycle time of the clock signal. The circuit of FIG. 2 therefore stops FIR filter 21st order (21-stage) with taps, which are at the outputs of levels 1, 5, 9, 11, 13, 17 and 21.

Weight factor circuits 102-118 are connected to the output taps of shift register 100 and multiply the tap signals by those in the Fractional factors shown in the figure. The FIR filter has an impulse characteristic that is relatively concentrated around the tapped middle stage 11. Signals from the stage 11 are in this example by the weight factor circuit 102 weighted by the weight factor 1/2, and the weighted tap signals are then fed to an input of an adder 130. Signals from levels 9 and 13, both of which are two levels from the Intermediate level 11 are away from the weight factor circuits 108 and 118 with the factor - (5/16) weighted and fed to inputs of an adder 120. Signals from levels 5 and 17, each six levels away from the intermediate level, will be from the weight factor circuits 106 and 116 to the Factor (5/64) weighted and fed to inputs of an adder 122. Signals from stages 1 and 21, both are ten levels away from the intermediate level, the weighting factor circuits 104 and 114 with the

Factor. - (1/64) weighted and then fed to the inputs of an adder 124. The symmetrical distribution of the tapped levels 1, 5 and 9 as well as 13, 17 and 21 around the Intermediate level 11 and the symmetrical distribution of the

Γ) weight irifaktorwortc give the FIR filter a linear Phase character intJk, as it eats necessary for the subsequent Elimination of the color signals.

IHe outputs of adders 122 and 124 are with. Entrances an adder 126 coupled, which the supplied Combined signals and the output of which is at the input of an adder 128. This combines the signals from the The adder 126 has signals from the adder 120 and its output is at the input tfinefi of the adder 130. The latter combines the weighted tap signals, which have been summed in adder 128, with the weighted ones Center tapping signals and has on his Output a band filter characteristic. In this way band-filtered color signals are produced at the output of adder 130.

The delay is the FIR filter described above substantially equal to the signal delay from the input of the filter to the intermediate stage 11 around which the Impulse characteristic of the filter is distributed. The delay element 24 from FIG. 1 is used to compensate for the Delay of the signal path for the full bandwidth in the circuit 20. In the arrangement according to FIG this compensation delay by connecting the output of stage 11 to an input of the subtraction circuit 26. In this way, the comb-filtered color input signals are offset by a compensating delay of eleven steps delayed at the output tap of step 11 , and full bandwidth unweighted signals from stage 11 are input to the subtracting circuit

BATH ORIGINAL

32 4 GS 0-5

fed. This subtracts the bandpass filtered color signals from adder 130 from the full bandwidth signals for the erasure of the color information in the full bandwidth signals, so that at the output of the subtraction circuit 26 only the vertical detail information remains. The arrangement is flexible in that the signals are full bandwidth as well from any point can be derived on which the signals of the full bandwidth with suitable phase position for the cancellation available.

The weight factor circuits and the adder tree from FIG. 1 are shown in greater detail in FIG. The output tap of stage 1 of the shift register 100 _{is denoted by T 1} , the output tap of stage 5 by T ₅ , etc. Since the weighting factors from FIG be weighted by which coefficient multiplier circuits can be omitted. For example, the tap output of stage 5 (Tc) is divided by 64 as shown in block 154 and divided by 16 according to block 158. These divisions are made by coupling appropriate higher order bits from the tap T ₅ to appropriate inputs for relatively lower order bits from subsequent adders 174 and 176, respectively. Block 154 illustrates the coupling of the three more significant bits of tap Tr to three lower-order bit inputs on one input side of adder 174, the tap outputs being divided by 64. Similarly, the five high order bits at tap T _{5 are} coupled to five low order bit inputs to one input side of adder 176, division by 16 (indicated in block 158).

The weighted tap signals are finally on

324U905

The output of adder 194 is summed to give a weighting of the tap output of the stage with the factor (1/64) + (1/16) = 5/64. In Fig. 3 show the Weight factor blocks 150-170 all have a similar coupling to the downstream adders, whereby the lowest-digit bit at each adder input has a value of 2 Has.

According to FIG. 3, signals from the tap T. are divided by 64 (indicated by block 150) when they are fed to an input of an adder 172. Signals from tap T ₂₁ are divided by 64 (block 152) when applied to a second input of adder 172. Signals from tap T ₁₇ are divided by 64 when applied to a second input of adder 174 (block 156) and are also divided by 16 when applied to a second input of adder 176 (block 160). Signals from the center tap T ₁ <. are divided by 2 when applied to an input of an adder 186 (block 170). Furthermore, unweighted signals from the tap T .. are fed to an input of the adder 198. Signals from the tap T _Q are divided by 4 and 16, respectively, as they are applied to inputs of adders 178 and 180 (blocks 162 and 166, respectively). Signals from the tap T ₁ <j, when fed to the second inputs of adders 178 and 180, are divided by 4 or 16 (blocks 164 and 168).

The output of adder 174 is coupled to an input of an adder 184, and the output of the adder is coupled via an inverter circuit 182 to a second input of the adder 184, the same time a logic "1" is supplied as a so-called carry-in bit. The inversion and the carry-in bit cause one

BAD ORIÖiNAL

2's complement formation of the output signal from the adder 172, whereby the coefficients of the signals from the Levels 1 and 21 become negative. The output of adder 184 is coupled to an input of an adder 192. 5

The outputs of adders 178 and 180 are with the inputs of adders 188 coupled, the output of which via an inverter circuit 190 at an input of a Adder 192, which also has a logical "1" is supplied as a carry-in bit. This inversion and the carry-in bit result in a 2's complement formation of the Output from adder 188 so that the coefficients the tapped levels 9 and 13 become negative "

The output of the adder 176 is connected to a second input of the adder 186, the output of which is connected to an input of an adder 194. The output of the adder 192 is connected to a second input of the adder 194. Since all of the weighted tap signals converge in the output signal of the adder 194, this adder has the desired bandpass filter characteristic for the color signals. The output of the adder 194 is connected to an input of the adder 198 via an inverter circuit 196 which, together with a logic "1" as the carry-in bit, generates a 2-digit compilation of the color signals for the subtraction of the full bandwidth signal from the center tap T ₁ ^ causes. At the output of the adder 198 there is thus a low-pass characteristic, as a result of which the low-frequency vertical detail information is made available.

The characteristics at the outputs of the circuits according to FIGS. 2 and 3 are from FIGS. 4 and 5 for a clock signal frequency from about 14.32 MHz can be seen. Fig. 4 shows a bandpass characteristic 200 as it is at the output

ί. -τ

of the adder 130 in Fig. 2 (which corresponds to the adder 194 in Fig. 3 corresponds). The characteristic curve in FIG. 4 shows a small maximum between approximately 0.7 and 1.1 MHz, however signals in this frequency range are attenuated by the FIR bandpass filter by more than 30 dB. The upper frequency limit of the passband is between 5 and 6 MHz, but remember that the baseband video signal because of the IF transmission characteristic of the television receiver normally drops above 4.2 MHz. Therefore, there are practically none in the pass band above 4.2 MHz Signals contain more.

At the output of adder 26 in FIG. 2 (which corresponds to adder 198 in FIG. 3) the in FIG. 5 shows the low-pass characteristic curve 210, within which the vertical detail information of the color comb filter from the luminance channel. The part 212 of the characteristic curve according to FIG. 5 lies above the cutoff frequency the television IF pass band of 4.2 MHz (baseband), and above this frequency there is practically none Signal more.

The isolating circuit according to the invention can also be implemented in an input-tapped form, as shown in FIG.

is shown. In this version, the FIR filter a 20-stage shift register 302 is used, and between 2-stage and 4-stage sections of the shift register adders 320-330 are inserted. Weighted input signals from the comb-filtered color signal reach the filter input 300 and via weight factor circuits 304-316 to the input of the first stage of the shift register 302 and to the adders therebetween. The shift register stages are clocked by a common clock signal.

BATH ORIGINAL

The comb-filtered color input signal arrives via weight factor circuits 304 or »316 with a weighting factor of - (1/64) to the input of the first stage or one input of an adder 330, the second input of which coupled to the output of the last shift register stage 20 is. The input signal is also passed through a weighting factor circuit 306 to the input of an intermediate the shift register stages 4 and 5 switched adder 320. Via a weighting factor circuit 314, the input signal also arrives at an input of an intermediate the shift register stages 16 and 17 switched adder 328. The weight factor circuits 306 and 314 weight their input signals with a factor of (5/64) “weight factor circuits 308 and 312 with the inputs of adders 322 and 326 are connected, weight the input signal rait a factor - (5/16). An adder 322 is located between the shift register stages 8 and 9, while a further adder 326 is between the shift register stages 12 and 13 is located. The weight function circuit 310 weights the input signal by a factor of 1/2 and feeds it in between the input of a adder 324 connected to stages 10 and 11.

The impulse characteristic of the FIR filter according to FIG. 6 is concentrated around adder 324, which is equidistant from the first and last stages. Weighted signal components are accumulated when passing through the shift register and the adders in the latter, and at the output of adder 330 there is finally a bandpass filter characteristic. To separate the vertical detail information of the color information, the bandpass-filtered color signal is sent to an input of a subtraction circuit 350 which subtracts the color information from the full bandwidth input signal. Since the Middle of the impulse characteristic of the FIR filter at

Adder 324 apart is from the output time by ten stages of shift registers 11 to 20, the signal is full bandwidth dam second !!!. Lncjang represents "to adder A 3 50 6 in balance shift register 340, which is also b contains ten Schl.oberegisteiratufen and by the same clock .nlgnal. how shift register 302 is clocked. Is * ^r> arise at the outputs of the adder 3 and 30 of the bandpass filter 350 and Subtrahierechaltung Tiefpaßfilterkennlinien as BSSW in Figs. 4 ßl.nd.

In the circuit according to FIG. 6, it is now assumed that the adders practically do not respond to the signals supplied Add term. However, if the adders used produce significant transit times, then the insertion an additional shift register stage in the signal path between the output of the adder 330 and the input of the subtraction circuit 350 be appropriate. The compensation shift register 340 would then have an additional compensation stage added.

When using the FIR filter according to the invention are operated by coded offset 2's complement data, in which the data are distributed around the zero point in the middle of the number system, then one can use complementary Achieve filter shapes by simply adding the sign of the Weighting factors and / or the signs of the last two signal combination stages are reversed. For example, if the signs of the weighting factors of the circuits 104, 106, 108, 114, 116 and 118 of FIG. 2 are all reversed, then the adder 130 gives the low-pass filter characteristic, and the subtracter 26 gives the bandpass filter characteristic. If one continues to use the adder 130 against a subtracting circuit and the subtracting circuit 26 is exchanged for an adder, then the adder 26 results in the low-pass filter characteristic,

BATH ORIGINAL

however, there is one for the subtracting circuit 130 Bandpass filter characteristic with reversal of the polarity of the filtered information regarding the information at the filter input. If you continue to get the signal full bandwidth subtracted from the stage 11 in the embodiment of FIG. 2 from the output signal of the adder 130 instead to do the opposite, then results for the subtraction circuit 26 with a low-pass filter characteristic Phase reversal. These complementary filter shapes may be desirable in applications in which the following Signal processing stages require phase-inverted signals. Analysis has shown that at least twelve such complementary filter shapes are possible in the circuit according to FIG. Complementary filter shapes can also be obtained for the circuit according to FIG. 6 by taking the signs of the weighting factors and / or the sense of combination of the adder 330 and the Subtracting circuit 350 reverses.

It should be noted that the isolating circuits according to the invention are also used in television receivers in which luminance and color signals are not separated from one another by comb filtering. Luminance and color signals can then be separated directly by the isolating circuits. At a such receivers, the values of the weighting factors or the clock frequency of the FIR filter can be selected so that that the crossover frequencies (transition bands) in the output characteristic moved to a higher frequency will. In the NTSC television system, this crossover frequency would be at about 3.2 MHz. At the low-pass filter output in such an arrangement signals occur in a pass band of up to about 3.2 MHz, and at the bandpass filter output signals with frequencies of 3.2 MHz up to the upper limit of the video frequency range would appear. if

3 2 4 O 9 Γ) Β

if you feed the digital video signal from the A / D converter to the input of the isolating circuit, then the. Luminance information signals would appear at the output with the low-pass characteristic, and the color information signals would appear at the output with the band-pass filter characteristic
ui'ifc-ho i.non.

BATH ORIGINAL

Claims (13)

PATENTANWÄLTE "^;""--DR. DIETETR-V, -6EZOl-D-- "- ■ DIPL. ING. PETER SCHÜTZ DIPL. ING. ¹ WOLFGANG HEUSLER MARIA-THERESIA-STRASSE 22 POST BOX Θ6Ο2 6Ο D-6OOO MUNICH 86 324CT905 APPROVED BY THE EUROPEAN PATENT OFFICE EUROPEAN PATENT ATTORNEYS MANPATA1RES EN BREVETS EUROPEENS TELEPHONE 089/4 70 60 06 TELEX 522 638 TELEGRAM SOMBEZ RCA 77465 / Sch / Ro.
US Ser. No. 319.109
AT: November 6, 1981 RCA Corporation, New York, NY (V.St.A.) Digital signal isolation circuit Claims η y) television receiver with a source of digital video signals, characterized by a comb filter (14) to which the digital video signals are fed to generate an output signal which contains frequency bands of luminance information signals and color information signals, by a delay device (100; 302, 340), the input of which the chimney-filtered output signal is supplied, for delaying the supplied signals, with a shift register (100; 302) which has a plurality of signal taps (1-21; 1-20) and causes a predetermined time delay between its input and its output and also provides an unweighted input signal that is by one 324Ü905 Period is delayed, which is equal to at least half of the predetermined delay time, by a Coupling circuit (102-118; 304-316) for weighted signals, which are connected to corresponding signal taps of the shift register is connected and the signals passing through it are weighted by one with the coupling circuit for the weighted Signals coupled device (120-130; 320-330), which interacts with the shift register and at a Output at which there is a first amplitude / frequency characteristic results, a first information signal, luminance or color information signal, provides, and by a signal combination circuit (26; 350) which has a first input at the output of the shift register cooperating device and with a second input to the delay device for feeding of the delayed unweighted input signal is connected and the supplied signals at an output at which a second amplitude / frequency characteristic results for a second information signal, luminance information or color information, combined. 2.) television receiver according to claim 1, characterized characterized in that the input of the shift register (100) is the output signal of the comb filter and that the plurality of signal taps (1-21) are output taps which are around a centrally located output tap (11) lying around that the coupling circuit (102-118) for the weighted Signals with an input connected to the output taps and the supplied signals as weighted Signals at an output supply that the circuit (120-130) interacting with the shift register is connected with one input to the output of the weight coupling circuit and that the signal combination circuit (26) is connected with a second input to the centrally located output tap and from -3-there receives the delayed unweighted signals. 3.) television receiver according to claim 2, characterized characterized in that the coupling circuit (102-118) for the weighted signals has a plurality of Contains weighting factor circuits which are symmetrical about the centrally located output tap (11) and weight have factors that are also symmetrically distributed around the centrally located output tap, and that the cooperating with the shift register circuit (120-130) contains a Äddierbaum circuit, which with the Weight factor circuits is coupled and at the output of which the first amplitude / frequency characteristic is present. 4.) television receiver according to claim 3, characterized characterized in that the adder tree circuit (120-130) contains a final adder (130), the a first input is supplied with the weighted tap signals from the plurality of output taps (1-21) and the weighted signals from the centrally located output tap (11) at a second input fed. 5.) television receiver according to claim 4, characterized characterized in that the signal combination circuit (26) has a subtracting circuit for subtractive Summary of the signals that appear at the central output tap (11) with those from the last one Adding stage (130) has delivered signals. 6.) television receiver according to claim 5, characterized characterized in that the subtracting circuit (26) includes an adding circuit (198) and means (196; CI = "1") to form the 2's complement of one of the dem O Δ. HU -4 adder contains signals fed to it. 7.) television receiver according to claim 1, characterized characterized in that the circuit (320-330) cooperating with the shift register has a plurality of adders and that the shift register (302) contains a plurality of stages (1-20) and the Adders are alternately coupled to certain of these stages, and that the shift register between one centrally located stage (10) and an output stage (20) has a predetermined delay that also the coupling circuit (304-316) for the weighted signals when the video signals are supplied to the weighted video signals alternately coupled shift register stages (1-20) and adders (320-328) supplies, and that with the shift register cooperating circuit (330) also with the output stage of the shift register and with the coupling circuit for the weighted signals is coupled and at its output a sum of weighted and delayed signals Signals. 8.) Television receiver according to claim 7, characterized in that the coupling circuit (304-316) comprises a plurality of weight factor circuits for the weighted signals. 9.) television receiver according to claim 8, characterized characterized in that the circuit (320-330) cooperating with the shift register Has final adder. 10.) television receiver according to claim 9, characterized characterized in that the delay device (302,340) has a second shift register (34Ö) with a delay which is substantially the same BATH ORIGINAL half the delay of the first-mentioned shift register (302) and provides an unweighted delayed input signal. 11.) television receiver according to claim 10, characterized characterized in that the signal combination circuit (350) a subtractive signal combination circuit for the subtractive combination of signals that of the final adder (330) of the circuit cooperating with the shift register and of the second Has shift register (340) delivered signals. 12.) television receiver according to claim 11, characterized characterized in that the subtractive signal combination circuit (350) has an adder (198) and means (196, CI = I) having a The input of the adder is coupled to the 2's complement formation of a supplied signal. 13.) television receiver according to claim 3 or 8, characterized in that the Gevrichtsffektschaltungen (102-118; 304-316) by displacement and adding circuits are formed.