GB2191360A - Circuit for processing video signals - Google Patents

Abstract

A circuit for processing transmission primaries (R,G,B) from a video signal source comprises an electronic matrix (68) for optimizing colour reproduction. The matrix has selectable coefficients by means of which the transmission primaries are either matrixed as intensity-linear signals or masked as logarithmated signals. <IMAGE>

Description

SPECIFICATION Circuit for processing video signals This invention relates to a circuit for processing transmission primaries from a video signal source.

As is known, thethreetransmission primaries R, G, B obtained by photoelectric conversion in a video signal sources, e.g. atelevision camera or film scanner, are linearly combined in an electronic matrix in order to provide the negative contents of the trichromatic responses necessaryforreproduction, and colour reproduction can be influenced and optimized through the choice of suitable matrix coefficients. When reproducing colour films by means of a television film scanner, such linear matrixing is used in orderto obtain a projection-reiated colour reproduction of positive films. Projection-related colour reproduction requires that a colour picture on the screen of a television monitor has the same colour reproduction asthedirect projection ofthe picture on to a projection screen.

In connection with television film scanners a different principle is also known (Professional Video, September 1984, pp.26-27), which seeks an objectrelated colour reproduction. In this method, which is preferably used forthe reproduction of negative films, it is necessary to eliminate the so-called colour density crosstalk present in the dye layers ofthe film material.

As the colour densities are logarithmically related to the transmissions and, in the case of a linear image converter characteristic, with the electric R, G and B signals, it is necessary to provide an electronic correction matrixforthe logarithmated R, G and B signals, i.e. so-called logarithmic masking.

Since with linear matrixing precise object-related colour image reproduction is not possible and with logarithmic masking precise projection-related colour image reproduction is not possible, the need has increased overthe pastfewyears for colour reproduction which can be switched from projection-related to object-related and vice versa. For this purpose a television film scanner mustswitchablycontain both linear matrixing and logarithmic masking, which greatly increases signal processing costs.

Afurther problem arises with the digitizing ofthe video signals to be processed. ForAID conversion of linear R, G, B signalsforsubsequent matrixing, it is necessary to have an 11 to 12 bit resolution in order to keepquantization errors belowthevisibilitythreshold.

For signals with a video band width, such A/D converters are very complicated and at present cannot be manufactured monolithically. For logarithmic masking, the R, G and B signals could undergo logarithmation priorto A/2conversion and then roughly8 bits would be sufficientforthe quantization visibilitythreshold. However, r, the characteristic in the reproduction of colòurfilms needed forthe presently required contrast raRge of 1 to 275 could not be achieved forvieo frequencies.

ThereforetJzetorobiem ofthe invention isto provide a circuit making it possible to achieve switchabie digital linear matrixing/loga rith mic masking with reduced effort and expenditure.

Accordingly, the invention provides a circuit for processing transmission primaries from a video signal source ofthe kind in which the primaries are supplied to an electronic matrix for optimizing the colour reproduction, wherein the matrix has select able coefficients whereby thetransmission primaries may either be matrixed as intensity-linear signals or masked as logarithmated signals.

The invention has the advantage that matrixing and masking can be performed with a single assembly, i.e.

the electronic matrix.

In the embodiment ofthe invention, it is particularly advantageous for the transmission primaries to be quantization pre-equalized during digital processing for reducing the visibility ofthe quantization.

An embodiment ofthe invention will now be described with reference to the accompanying drawings, wherein Figure 1 is a block circuit diagram for performing known linearmatrixing; Figure 2 is a block circuit diagram for performing known logarithmic masking; and Figure 3 is a block circuit diagram according to the embodimentofthe invention.

Figure 1 represents a known circuit arrangementfor processing video signals for a projection-related colour reproduction of positive films. In this,transmission primaries R, G and B are produced by scanning a colourfilm, notshown in the drawings, by means of optoelectric converters 1,2 and 3 and each ofthe transmission primaries is supplied via an amplifier stage 4,5, 6to an input a linear matrix 7, where the transmission primaries R, G, B are conventionally linearly combined by a resistance network. Thus, transmission primaries R', G' and B' influenced and optimized by corresponding matrix coefficients are produced at the outputs of matrix 7.These matrixed transmission primaries are then passed across gradation pre-emphasis stages 8,9,10 and converted into 8 bits digital signals by means of A/D converters 11,12 and 13. In afollowing digital signal processing circuit 14, the digital video signals are further processed in known manner, such as e.g. conversion of the sequential input signal into a line jump signal, aperture and further corrections. The digital transmission primaries produced atthe outputs ofthe signal processing circuit 14arethen converted in DIA converters 15,16,17 into analog, gradation preequalized transmission primaries Rill, GilT and Bill, which are available at the outputterminals 18, 19 and 20.

Figure 2 shows a known circuit for processing video signals for object-related colour reproduction, preferably in the case of negative films. Once again transmission primaries R, G and B are produced by scanning a colour negative film, now shown in the drawing, by means of optoelectric converters 21, 22 and 23 and the transmission primaries are supplied across respective amplifier stages 24, 25, 26 to the inputs of an AID converter 27, 28, 29. Due to the following gradation pre-emphasis gamma corrector each A/D converter 27, 28, 29 must be able to process 12 bit digital signals, which is possible e.g. through the use of two 8 bit AID converters. These digitized transmission primaries, which are present in linear form, are converted in following memory units 31,32, 33, e.g.PROMs using a so-called look-uptable, into logarithmated transmission primaries. The correction forthe object-related colour reproduction then takes place in a logarithmic masking stage 34. Thethus corrected transmission primaries are then passed across gradation pre-emphasis stages 35,36,37 and are delogarithmated in stages 38,39,40. Stages 38, to 40 can also each be in the form of a PROM with a look-up table.As in the circuit according to Figure 1, in a following digital signal processing circuit 41, the transmission primaries supplied undergo further processing operations and arethen converted with the aid of DIA converters 42,43,44 into analog transmission primaries RilT, GilT and Bilt, which are available at the output terminals 45,46,47.

Figure 3 shows a circuit arrangement according to the embodiment ofthe invention,whereonce again transmission primaries R, G and B are produced by scanning a colourfilm, not shown in the drawing, by means ofoptoelectricconverters 51,52,53 and are amplified in following amplifier stages 54,55,56.

These amplified R, G and B signals are then supplied to quantization pre-equalization stages 57,58,59 with a non lineartransmission characteristic according to ClE-L*-formula Y = 1.16 - 0.16. Thistransmission characteristic used as the quantization characteristic on the basis of this formula makes it possible for the transmission primaries to only require8 bitdigitization in thefollowing A/D converters 61,62, 63. Thus, compared with conventional circuits, this makes it possible to economize by needing only one AID converter in each ofthethree colourchannels.The three transmission primaries R, G and B are then supplied as 8 bit digital signals to a digital signal processing circuit 64, where the transmission primar- ies undergo known processing operations, such as contour correction, colour correction, conversion of the sequential signals into line jump signals, etc. The outputs of this circuit 64 are connected to respective coding stages 65,66,67, which can comprise memory or digital arithmetic circuits with PROMs or RAMs and in which the quantization pre-equalized transmission primaries undergo code conversion into intensitylinear or logarithmated transmission primaries.These intensity-linear or logarithmated transmission primaries are supplied to a matrix 68 with variable coefficients, and in which, in accordance with the input signals, there can eithertake place a matrixing of the intensity-linear signals or a masking of the logarithmated signals. With the aid of a control signal at terminal 69, it is possible to modify or adjust the co-efficients required in each case. It is then possible to take from the outputs of matrix 68 correspondingly corrected linear or logarithmated transmission primaries. The output signals of matrix 68 undergo code conversion into logarithmatedtransmission primaries in the following coding stages 71,72,73. Code conversion in stages 65 to 67 and 71 to 73 takes place by means of a switching signal which can be applied to terminal 74, so that the coding stages are always switched in the same direction.

Thus, in the case of an intensity-linear input signal, there is a linear matrixing, so that linear transmission primaries can be taken from the output of matrix 68. In the case of a logarithmated input signal, there is a logarithmic masking, so that logarithmictransmission primaries can be taken from the output. Thus, in each case logarithmic transmission primaries are produced in coding stages 71 to 73 and are then simultaneously processed in de-logarithmation stages 75,76,77 and gradation pre-emphasis stages 78,79,80. The de logarithmation stages 75 to 77 produce lineartrans- mission primaries, which are converted by means of the D/A converters 81,82,83 into analog transmission primaries R, G, and B available at the outputterminals 84,85,86. The logarithmated transmission primaries gamma-corrected in the gradation pre-emphasis stages 78,79, 80 are converted into analog, gradation pre-emphasized transmission primaries R1/T, GilT, B available atthe output terminals 91,92,93.

Thus, with the circuit arrangement according to the invention it is possible withoutsignificant additional expenditure and using a single matrix arrangement 68, to matrix the lineartransmission primaries or mask the logarithmated transmission primaries, whilst simultaneously deriving therefrom linear and grada tion-pre-emphasized transmission primaries.

Claims (9)

1. Acircuitfor processing transmission primaries from a video signal source ofthe kind in which the primaries are supplied to an electronic matrix for optimizing the colour reproduction wherein the matrix has seleetable coefficients whereby the transmission primaries may either by matrixed as intensity-linear signals or masked as logarithmated signals.

2. Acircuitaccording to Claim 1, wherein the coefficients ofthe electronic matrix are modified by means of control signals.

3. Acircuitaccordingto Claim 1 or 2, wherein the transmission primaries are quantization pre-equalized and digitized priorto matrixing.

4. A circuit according to Claim 3 wherein the quantization pre-equalization is represented by a characteristicaccordingtoy 1.16 x 1l30.16.

5. A circuit according to Claim 3 or 4, wherein the intensity-linear or logarithmated transmission primaries supplied to the matrix are produced by code conversion of the quantization pre-equalized signals.

6. A circuit according to any preceding claim, wherein the intensity-linear matrixed orthe logarithmic masked transmission primaries from the matrix are code converted to linear and gradation preemphasized transmission primaries.

7. A circuit according to Claims Sand 6, wherein code converters provided upstream and downstream of the matrix are switched according to whether intensity-linear matrixing or logarithmic masking is effected.

8. Acircuit according to Claim 1 wherein the three transmission primaries are each supplied to a respective quantization pre-equalization stage whose output is connected across a respective A/D converter to a digital processing circuit, wherein the outputs of the processing circuit are each connected to a coding stage for the code conversion of the pre-equalized digital transmission primaries into intensity-linear or logarithmated transmission primaries, wherein the outputs of the coding stages are connected tothe electronic matrix whose outputs are each connected to a further coding circuit for code conversion of the intensity-linearor logarithmated transmission primaries into linear or gradation pre-emphasized signals, and wherein the coding stages both upstream and downstream ofthe electronic matrix can be controlled by means of switching signals supplied to their control inputs for switching from linear into logarith- mic mode and vice versa.

9. A circuitfor processing transmission primaries frorn a video signal source, substantially as described herein with reference to Figure 3 ofthe accompanying drawings.