JPS61278290A - Digital recording and reproducing device for compound color video signal - Google Patents

Abstract

PURPOSE:To visually make the deterioration of a picture quality invisible as much as possible by shuffling with the combination of samples of the number as least as possible to be required which makes possible the color separation after applying the shuffling and distributing a sample point to be corrected on a television display screen even when an error incapable of being corrected by an error correction code is generated during reproducing. CONSTITUTION:The data divided into three channels by a channel separator 12 during recording is divided into groups of continuous four samples in the next shuffling circuits 13a-13c. The shuffling is applied so as to distribute the continuous groups on a display screen to a wide range. During reproducing, in deshuffling circuits 23a-23c, the operation absolutely reverse to the shuffling is operated. Thus, even when an error incapable of being corrected by an error correction code during reproducing is generated, the data to be corrected is distributed on the display screen by the group unit of the four samples. Even if it is replaced by an adjacent sample of low correlation, there is no problem in a visual deterioration of the picture quality.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a digital recording and reproducing apparatus for composite color video signals.

Conventional technologyDigital VTR that digitizes and records video signals
In such recording devices, sample errors during playback are the biggest cause of image quality deterioration. This sample error is usually
Errors that are protected by an error correction code but cannot be corrected by this error correction code are corrected. This correction can be done by replacing the erroneous sample with a sample that is spatially and temporally adjacent, or by performing an operation between multiple samples that are spatially and temporally adjacent to the erroneous sample, and using the result of the calculation. It interpolates erroneous samples. Furthermore, when considering a helical scan type digital VTRQ, multiple fields of signals are reproduced during one field period during special playback such as slow, still, and search, so in particular, composite color signals such as NTSC color television signals are reproduced. In the case of video signals, the continuity of subcarriers is a problem.

Next, these points will be explained in some detail using an example of a digital VTR-f as a recording/reproducing apparatus.

FIG. 2a shows an example of a plot diagram of a recording system of a digital VTR, and FIG. 2b shows an example of a block diagram of a reproduction system.

In FIG. 2 &, 30 is a video input terminal to which, for example, an NTSC color television signal is input. 31, an A/D converter samples the input signal at a frequency of, for example, four times the subcarrier frequency fso (C4fsc), and quantizes each sample to 8 bits. In this case, the recording bit rate is 114 Mbps, and since the rate is too high to record on one channel, the data is divided into three channels and recorded. 32 is a channel separator for performing all channel divisions. Fifth
The figure shows how channels of a television signal are divided during a 1H period (one horizontal scanning period). That is, in the case of an NTSC television signal, the following relationship exists between the subcarrier frequency fB0 and the line frequency fg.

Therefore, when sampling is performed with ', /'sc, there are 910 sample points in the 1H period. Of these, it is assumed that all 768 samples are actually recorded as a video signal.

In the above channel separator, 1
H minute data (768S) is 0H-A.

C) H minutes (2
563). Next, 333°33b,
330 is an error correction encoder. In digital VTRs that record digital video signals and sounds as they are, bit errors during playback are the biggest cause of deterioration in image quality. Therefore, even when this bit error occurs, it is very important to efficiently detect and correct this error to restore the original data. The error correction encoder uses 0RCC (
Cyclic Redundancy Check C
The purpose is to add parity data for correction. As an example of an error correction code, consider, for example, a chain code as shown in FIG. Here, each sub-block is divided into %H (12
ss) data [5YNC, ID.

This is a data unit configured by adding OR and CC, respectively.

This entire chain code consists of one field of each channel (
252H). At the right end of each row, horizontal parity is added at a rate of 1 block for every 12 blocks, and at the bottom end of each column, vertical parity is added at a rate of 1 block for every 21 blocks.

These parities were determined by the following equation. For example, horizontal parity of the first row -8825,5B2
6 is calculated with, and the vertical parity of the first column -8B547 is S
B, ■5B27■・・・・・・■SB5□1=SB54
□ ...... Calculated by (3). (2) + (
In equation 3), the symbol ■ represents the exclusive OR between corresponding bits in each block. In the chain code constructed in this way, if there is an error of one block in each of the horizontal and vertical directions, it is possible to correct it using the parity described above. However, for this purpose, it is necessary that the block containing the error has been detected by cucc in advance.

aRca and horizontal by error correction encoder.

The signal to which vertical parity has been added is sent to the next recording processors 342L, 34b, and 34C. In this recording processor, an 8-bit signal quantized by an A/D converter is converted into, for example, 10 bits. Among these 10 bits, there are six different binary values "1'" and six "0", so that the data after conversion does not include a DC component. This is because DC components cannot pass through digital VTRs because they pass through a rotary transformer during the recording and reproducing process and because the reproducing process is essentially a differential characteristic. Furthermore, in this recording processor, 5YNC and ID in one subblock in FIG. 5 are added.

The SYN C signal is a signal indicating the start of this sub-block, and the ID signal is the sub-block number shown in FIG.
Alternatively, it consists of channel distinction, frame even/odd signals, etc., which will be explained in detail later, and are required during special playback. The signal converted into serial data by the recording processor is sent to the next recording amplifier 351.
L, 36b.

35ci are recorded on the chip via the recording heads 11L, 1b, and IC'i.

Next, the regeneration process shown in FIG. 2 will be explained.

Signals taken out from the playback heads 11, 1b, 10 are manually inputted to playback processors 37&, 37b, 37C via playback amplifiers 36&, 36b, 36C'ii.

In the reproduction processor, the reproduced analog signal is converted back into a digital signal consisting of two binary values ++1+l, +(011), and then the 10-pit signal is transformed into an 8-pit signal by performing the completely opposite conversion to that of the recording processor. signal, then 38&.

3Bb and 3BO are TBCs that remove fluctuations in the time axis that occur during the recording and reproducing process. Reference numeral 39 denotes an inter changer, which is not necessary for normal playback, but for special playback, since data on a different channel than when recording is also played back at the same time, it is necessary to adjust data between channels.

4C1, 401), 400 is an error correction decoder, which detects error υ in subblock units using cRca added to each subblock, and based on this information, horizontally and
Errors are corrected using vertical image parity. 411L, 41b.

41Ct is a buffer memory, one for each channel.
Field (a bit x 4fa0÷1/6o÷
It is a memory with a capacity of 1/3 to 2640 Kb. This will be explained in detail later, but it is necessary for special playback. Reference numeral 42 denotes a channel mixture, which combines data divided into three channels by a channel separator during recording and returns it to one channel. Reference numeral 43 denotes an error corrector, which interpolates errors that cannot be corrected by the error correction code using spatially and temporally adjacent highly correlated samples. Then, the signal is converted back to the original analog signal by the final D/MU converter 44 and taken out from the output terminal 46.

Problems to be Solved by the Invention Consider the case where an error that cannot be corrected by an error correction code occurs in the digital VTR described above. In this case, the error corrector replaces the erroneous sample with surrounding highly correlated samples.
As described above, the unit of error detection using <cRcc is one subblock (128S, X/6H minutes). therefore,
The unit of the above replacement is one subblock, which extends to the length of %H on a TV screen, and if there is a high correlation between the error subblock and the subblock to be replaced, it will not be a big problem, but if the correlation is If it is low, there will be a visually large deterioration in image quality. FIG. 7 shows the state of the modified sub-blocks on the television screen. Therefore, from the viewpoint of error correction, it is desirable that the size of the sub-block is smaller, but making the sub-block unit smaller means that the proportion occupied by cRcc will increase relatively. There is naturally a limit to the size of a subblock due to the redundancy allowed for the entire error correction code.

The best way to solve all of the above problems is to shuffle the data sample by sample before the error correction encoder during recording to increase the distance between successive temples. By doing this, even if an error like the one shown in Figure 7 occurs during playback, the sample to be corrected will be
For example, as shown in Figure 8, it is distributed on the screen in sample units,
Even when replacing with a sample with low correlation, the image quality deterioration becomes visually less noticeable.

However, there is one problem here. This is the case when special playback such as slow, still, and search is executed in a helical scan type digital VTH. In the case of special playback, as is clear from FIG. 5, the video head scans over a plurality of tracks in one field period.

Therefore, the reproduced data becomes intermittent, and if left as is, band-like errors as shown in FIG. 9 will occur on the screen. Therefore, as shown in Figure 2, a buffer memo IJ'r is prepared for each channel, and the subblock is transferred to a predetermined address on the buffer memory according to the subblock number of the signal added to the subblock that has been reproduced. Write the data. No new writing is performed to the address corresponding to the sub-block that could not be reproduced.
The previous data will remain as is. If this is done, errors like the one shown in FIG. 9 will not occur.

However, the sub-blocks written to the buffer memory during one field period are not only from a single field but also from multiple fields, especially when the input signal is NTSC. In the case of composite color television signals, the following disadvantages occur. That is, in the case of NTSG, the subcarrier frequency fH and the frame frequency fr have the following relationship.

As a result, the phase of the subcarriers is inverted every frame. Therefore, if even-odd frame data is mixed in one field's worth of data read from the buffer memory, the colors will be reversed between those data, resulting in extremely serious deterioration in image quality. . Since the even-odd information of the frame is included in the ID signal, all ID signals of sub-blocks read from the buffer memory are always monitored, and when the data of 7 frames opposite to the original frame is read, the color It is necessary to completely separate only the signal and reverse the polarity. When a sub-block consists of a large number (128) of consecutive samples as in the conventional example, this color separation can be easily achieved using a digital filter or the like. However, as mentioned above, if sample-by-sample shuffling is performed before the error correction encoder during recording,
The samples within the subblock become discontinuous, and color separation is no longer possible.

Means for Solving All Problems The present invention provides composite color video signals at subcarrier frequency f.
Sampling is performed at a frequency of n times s0 (n = 3 or 4), consecutive n x m (m is a positive integer) samples are grouped, and the data is rearranged in time series for each group. a means for recording and reproducing all the data subjected to this rearrangement operation; and a means for performing the rearrangement and returning to the previous data arrangement by performing an operation completely opposite to the rearrangement at the time of reproduction. This is a digital recording and reproducing device for composite color video signals.

According to the present invention, shuffling is performed using a combination of the minimum number of samples necessary to enable color separation even after shuffling, and even if an error that cannot be corrected by an error correction code occurs during playback, By dispersing the sample points to be corrected over the entire television screen, it is possible to make image quality deterioration as visually inconspicuous as possible even if replacement is performed with samples with low correlation. Furthermore, even when special playback is performed, color signals can be separated and their polarities can be reversed, and significant image quality deterioration such as color reversal will not occur on the screen.

Example An embodiment of the invention is shown in FIGS. 1a and 1b.

In this figure, 10 is an input terminal, 11 is an A/D converter, 12 is a channel separator, and 13 &.

13b, 13C are shuffling circuits, 14&.

14t), 140 is an error correction encoder, 15&.

181), 15C is a recording processor, and 16a.

18t), 160 is a recording amplifier, 13.1 +, IC are a recording head and a reproducing head, 17&, 17t).

17C is a reproduction amplifier, 1 B & , 18b, 18
C is a playback processor, 191.19t), 190TB
C.

20 is interche 7 ja, 21&, 21+).

2IC is an error correction decoder, 22a, 22b.

22C is a buffer memory, 23a, 23b.

23C is a deshuffling circuit, 24 is a channel mixture, 26 is an error corrector, 26 is a D/person converter, 2
The end is the output terminal. In this figure, shuffling circuits 13+L, 13b, 13 (i and deshuffling circuit 23
Components other than a, 23b, and 23c and their operations are completely the same as those in FIG. 2, so their explanations are omitted here.

During recording, the data divided into three channels by the channel separator is divided into groups of four consecutive samples in the next shuffling circuit. Then, shuffling is performed so that these consecutive groups are widely dispersed on the screen. During reproduction, an operation completely opposite to shuffling is performed in the deshuffling circuits 23&, 23b, and 230VC.

In this way, even if an error that cannot be corrected occurs in the error correction code during playback, the data to be corrected will be distributed over the image in groups of 4 samples as shown in Figure 8. Therefore, compared to the modification in units of one subblock in the conventional example, even if adjacent samples with low correlation are replaced, the deterioration of image quality does not pose as much of a problem visually.

Next, in such a shirt ring of 4 samples,
Even if the even-odd frames are reversed during special playback,
This shows that color separation is possible and image quality deterioration due to color inversion is not a problem.

Since the sampling frequency during A/D conversion is four times the subcarrier frequency fsc, the angle is 0° with respect to the subcarrier frequency ISO as shown in FIG. 1o.

Sampling will be performed at 90°, 180°, and 2700 points. Therefore, the signal level of each sample point is set to Sl, S2, . S6. S4, the color video signal Sk (k=1.2, 3.4) is given by the following equation.

1.14 203s
in(JJsct =Y+DROO8ct+SCt+DBSinωSCt
・----fp) 1.14
203 Therefore, the band of signals (R-Y) and (B-Y) is 500kHz
, which is sufficiently low compared to the sampling frequency of 4 fB (3), i.e., the signal (R-Y).

Since the amplitude change of (BY) is sufficiently slow compared to the sampling period, DR, UDR3, DR2, DR4, etc.) are obtained. Therefore, if there is a sample output for one cycle, it is possible to completely separate the luminance signal and the carrier color signal from the sample output.

Therefore, even if the even-odd frames are reversed during special playback, the carrier color signal can be separated from the shuffled group of consecutive four samples by the process described above, and by reversing this polarity, the color reversal problem can be solved. can be resolved.

DETAILED DESCRIPTION OF THE INVENTION According to the present invention, even if an error that cannot be corrected by an error correction code occurs during playback and error correction is performed, the samples to be corrected are distributed on the screen in groups of four samples. Therefore, even if replacement is performed with adjacent pixels whose correlation is not very high, image quality deterioration does not visually cause much problem. Moreover, when special playback is taken into consideration, even if even-odd frames are reversed during playback, the color signals can be separated and inverted, and this does not lead to deterioration such as color signal inversion. These things can be achieved very easily by shuffling the data every four consecutive samples.

Note that the shuffling circuit and deshuffling circuit described here are both simple memory write operations such as RAM, etc.
This can be realized by appropriately controlling the read address, and can be configured extremely easily in terms of nodes. Further, in this embodiment, only the case where the sampling frequency is 4 fsc has been described, but the present invention is equally effective even in the case of 3 ft AC. However, in this case, shuffling may be performed in groups of three consecutive samples. It goes without saying that the present invention is also effective for composite color television signals other than NTSC, such as PAL. Further, in the present invention, a backup memory for one field is separately provided for special playback, but when an error correction code as explained in the conventional example is used, the error correction decoder It has a memory with a capacity of one field, and it is clear that it can be used for multiple purposes. Furthermore, in the above explanation, when sampling is performed at a frequency of nXfsc (n=3y4), only a group consisting of n consecutive samples is considered as a shuffling unit.
The present invention is equally effective even if the general i/i: is considered to be a group consisting of nxm (m is a positive integer) consecutive samples.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a block diagram of a conventional digital VTR, and FIG. 3 is a plan view showing the relationship between the tape, head, and cylinder of a helical scan digital VTR. , Figure 4 is a diagram showing the recording pattern on the tape and the trajectory of the head during playback, Figure 5 is a diagram showing the channel division and sub-block data structure for 1H of an NTSC color television signal, and Figure 6 is an error. A diagram showing an example of the structure of a code when a chain code is used as a correction code, FIG. 7 is a diagram showing an error correction unit in the conventional example, and FIG. 8 is a diagram showing an error correction unit in the present invention.
FIG. 9 is a diagram showing band-like errors during special playback, and FIG. 10 is 4f. FIG. 3 is a diagram showing the relationship between subcarriers and sample phases when sampling is performed. 12... Channel separator, 13a, 13
b. 13c...Shuffling circuit, 142L, 14
1). 140...Error correction encoder, 20...
... Interchanger, 21 & , 21 b
, 210---Error correction decoder, 22m, 2
zb, 22c...Buffer memory, 23a,
23b, 23c...Deshuffling circuit, 2
4...Channel mixture, 25...
...Error corrector. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
Figure (Ku) Cb) Figure 2 Figure 3 Figure 4 Jν, 3b JC Figure 5 Figure 6 Data Figure 7 Figure 8 Figure 9 Figure 10

Claims (1)

[Claims] The composite color video signal is transmitted at subcarrier frequency f_s_c
Sampling at a frequency n times (n = 3 or 4)
A means for chronologically rearranging data in groups of n×m (m is a positive integer) consecutive samples, and recording and reproducing data that has been rearranged. 1. A digital recording and reproducing apparatus for a composite color video signal, comprising: means for returning to the data arrangement before the rearrangement by performing an operation completely opposite to the rearrangement during reproduction.