JPS6168775A - Television signal digital recording and reproducing device - Google Patents

Abstract

PURPOSE:To minimize expansion of error by selecting the bit number of conversion block to an integer number of multiple of a modulation block with the use of a block code. CONSTITUTION:From a TV signal quantized to N bits via an A/D converter 102, L sets of samples are subjected to blocking by a blocking device 104 of a high-efficiency coder 103, the samples are subjected to compression conversion from NXL bits into MXL bits by a Hadamard converter 105 ad converted into a data word of P-bit block modulation code by a modulator 108 via an error correction coder 107 and a quantizer 103 so as to be matched to the recording density and recorded/reproduced. In selecting the bit number to satisfy the relation of L.M=S.P, where S is a positive integer, the error delivery range after Hadamard inversion is limited to one Hadamard conversion block not exceeding two Hadamard conversion blocks and the spread of error is minimized.

Description

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

Conventional Structure and Problems If a general television signal is digitized and the digital signal is to be recorded as is in a recording device, the recording rate will be approximately 100 Mbps, which is extremely high.

Conventionally, high-efficiency encoding methods such as L, DPCM, or Hadamard transform have been used as a band compression method for reducing the recording rate to about % while maintaining the input image quality.

However, if such a method is used to reduce the recording rate, when a bit error occurs during playback,
The problem is error propagation. In particular, in the case of DPCM, during decoding, newly reproduced prediction error signals are successively added based on past reproduced data, so error propagation extends over a wide range.

On the other hand, in the case of orthogonal transform such as Hadamard transform, the range of error propagation is determined by the size of the transform block, and if the block size is not very large, error propagation can be within a relatively small range. Therefore, the Hadamard transform is suitable for recording devices such as digital VTRs, which have a relatively high bit error rate of 10 to 10 2 during reproduction. By the way, when recording a digital signal on a recording medium, modulation is often used to adapt the digital signal to be recorded to the transfer characteristics of the recording/reproducing system, to improve the recording density, or to lower the bit error rate during reproduction. It is common to do so.

FIG. 1 shows a schematic block diagram of a recording and reproducing system of a digital VTR when Hadamard transform is used as a high-efficiency encoding method.

In FIG. 1, an analog signal (considering an NTSC color TV signal as an example) inputted from an input terminal 101 is inputted to an A/D converter 102 at, for example, 4fSC (fS
C is sampled at a frequency (color subcarrier frequency), quantized to 8 bits per sample, and then input to the next high efficiency encoder 103.

This high-efficiency encoder 103 consists of a block generator 104, a Hadamard transformer 106, and a quantizer 106, where the number of pits is reduced to about 4 to 6 bits per sample while maintaining the input image quality. . That is, originally 114
Mbps bit rate is 57Mbps to 72Mbps
In other words, it is reduced to

Then, in the next error correction encoder 107, parity bits necessary for detecting and correcting bit errors during reproduction are added. The modulator 108 adapts the digital data signal output from the error correction encoder to the transmission characteristics of the tape head system before recording it on the magnetic tape.
In order to achieve higher density, data conversion is performed according to certain rules. The signal output from the modulator is finally recorded on tape 110 via recording head 109. During reproduction, the output signal taken out from the reproduction head 111 is first subjected to data conversion in the demodulator 112 that is completely opposite to the data conversion performed by the modulator during recording. Then, in the next error correction decoder 113, the parity bits added during recording are used to detect and correct bit errors that occur during reproduction. Next, high efficiency decoder 11
4 is composed of an inverse quantizer 116, a Hadamard inverse transformer 116, and an inverse blocker 117.L, the high-efficiency encoder 103 at the time of recording is completely inversely converted into an 8-bit signal per sample. be returned.

Finally, the signal is converted back to the original analog signal by the D/A converter 118 and output from the output terminal 119.

Now, in Fig. 1, the Hadamard transform is the adjacent −
(f!,: positive integer) samples are treated as one block, and transformation is performed in units of this block.This block will be referred to as a Hadamard transformation block. On the other hand, there are many known recording codes created by the modulator shown in FIG. 1, which can be broadly classified into bit conversion codes and block codes. A bit conversion code converts each bit based on certain rules, and a block code converts each bit into blocks (data words) of a predetermined length and then converts them into recording words based on certain rules. It is. As the former code, the M code, interleaved NRZI code, etc. are well known, and as the latter code, the 8-10 transform block code is well known. Now, the case where a block code is used as this recording code will be explained below.

If we call this block of block code a modulation block, the difference between the size and number of pits of the modulation block before modulation and the size and number of pits of the Hadamard transform block after the conversion is an integer multiple. If there is no relationship, the following inconveniences will occur.

Now, as an example, the Hadamard transform is shown in Figures 2 and 3.
Consider a two-dimensional 8-order Hadamard transform with a block configuration as shown in the figure. Here, x4 to x8 are each sampled sample.

That is, the input column vector to the Hadamard transformer is
If the output column vector after Hadamard transformation is Y, then the following relationship holds between the two.

Y=H8 workX ・・・・・・・・・・・・・・・
...(2) Here, H8 is an 8th order Hadamard transformation matrix and is given by the following equation.

Below, remaining white Here “+” means +1. -” represents -1.

Now, suppose the number of bits is reduced from 8 bits per sample to 4.5 bits by Hadamard transform.
The size of the Hadamard transform block after conversion is 4.5 bits x 8 = 36 bits.

On the other hand, if an 8-10 conversion block code is used as the modulation code, the size of the modulation block before modulation is 8 bits. And the block code modulator can be easily configured using ROM.

FIG. 4 shows the configuration of a modulator a and a demodulator S for an 8-10 transform block code. In FIG. 4, 201 and 204
is a serial-to-parallel converter, 202 and 205 are 256 words x 10 bits and 1024 words x 8 bits, respectively) (7) R
OM, 203.206 are parallel to serial converters.

Now, when using a block code as a recording code, if an error occurs in even one pit in a recorded word (10 bits) during playback, the correct data word will be demodulated into a completely different data word during demodulation. Therefore, the error increases to the size of the modulation block (8 bits). Therefore, as shown in FIG. 2, the Hadamard transform block size (36 bits) is not an integral multiple of the modulation block size (8 bits), and the Hadamard transform block HB2. If an uncorrectable error occurs in the modulation block MBs that straddles the boundary of HB3, the error will spread over two Hadamard transform blocks (shaded area) in the television signal that was finally inversely transformed by Hadamard. That is, although originally only one Hadamard transform block error would be sufficient, there is a major drawback in that the error area doubles.

Purpose of the Invention Accordingly, the purpose of the present invention is to use orthogonal transform such as Hadamard transform to reduce the recording bit rate for an input TV signal, and furthermore, when a block code is used as a recording modulation code, if during reproduction, Even if an uncorrectable modulation block error occurs, the problem is to provide a relationship between the size of the two blocks and the number of bits so that the error can only be expanded by one Hadamard transform block.

Structure of the Invention The television signal digital recording and reproducing apparatus of the present invention includes:
means for sampling an input television signal at a predetermined frequency and quantizing each signal/pull into N bits; dividing the sample into L adjacent blocks; and applying a high-efficiency block code to each block; means for reducing the number of bits per sample to M bits per sample (M≦N), and recording a data word of P bits to Q bits (Q≧P) for the data with the reduced number of bits per sample. between L, M, and P, L, M=S, P
(S is a positive integer) is established, so even if a modulation block error occurs, the expansion of the error can be kept to a minimum.

DESCRIPTION OF THE EMBODIMENTS The general structure of the recording and reproducing system of a digital VTR according to an embodiment of the present invention is the same as that shown in FIG.

In this embodiment, Hadamard transform, which is a type of orthogonal transform code, is used as high-efficiency block coding.

In this embodiment, an input TV signal is quantized into N bits, L adjacent samples are formed into blocks, and Hadamard transform is applied to reduce the signal to M bits per sample. In other words, from LxN bits per diblock to Lx
Reduce to M bits.

If the number of bits of a data word of a block modulation code is P, then L, M = S salary P ・・・・・・・・・・・・・・・・・・
・Construct so that the relationship (4) holds (S is a positive integer) Figure 3 shows the case where S = 4, that is, the sample is set to f) M = 4 bits by Hadamard transformation. show. In this case, it is clear that the range of error propagation after Hadamard inverse transform is limited to one Hadamard transform block, no matter which modulation block is erroneous.

In this embodiment, only Hadamard transform has been described as high-efficiency block coding, but it goes without saying that other block coding such as orthogonal transform or vector quantization can have exactly the same effect.

Furthermore, it is clear that the present invention is effective not only for the 8-10 transform block code but also for other block modulation codes as a modulation code.

As described in detail, the present invention performs high-efficiency block encoding, uses a block code such as an 8-10 transform block code as a recording modulation code, and converts the transform block (size and number of bits) into a modulation block. Since the size is selected to be an integral multiple of the number of bits), the expansion of errors can be kept to a minimum.

[Brief explanation of drawings]

Figure 1 is a block diagram showing the outline of the digital V and TR recording and reproduction system when Hadamard transform is used as high-efficiency encoding, and Figure 2 shows the size and number of bits of the Hadamard transform block and the modulation block. size and focus)
A diagram illustrating the expansion of errors when and is not an integer multiple,
FIG. 3 is a diagram showing the expansion of error 9 when the Hadamard transform block size and number of bits) and the modulation block size and number of bits are in an integral multiple relationship in one embodiment of the present invention; FIG. 4 is a diagram showing the configuration of a modulator/demodulator for an 8-10 transform block code. 104...Blocker, 105...
Hadamard transformer, 106...Quantizer, 10
7...Error 9 correction encoder, 108...Modulator, 112...Demodulator, 113...
Error correction decoder, 115... Inverse quantizer, 11
6... Hadamard inverse transformer, 117...
・Deblocker, 201.204...Serial to parallel converter, 203.206...Parallel to serial converter, 2
02゜205...ROM. Figure 2-~) Deaf q-F-false one tun~m-tun tun Mimi ζ (ζζmiξ Figure 4 ta)

Claims (4)

[Claims] (1) A means for sampling an input television signal at a predetermined frequency and quantizing each sample into N bits, and dividing the sample into L adjacent blocks and performing high-efficiency block encoding on each block. means to reduce the number of bits per sample to M bits (M≦N), and convert the data word of P bits into a recording word of Q bits (Q≧P) for the data with the reduced number of bits per sample. It has a modulation means for converting, and between L, M, and P, L・M=S・
1. A television signal digital recording and reproducing device characterized in that the following relationship P (S is a positive integer) is established. (2) The television signal digital recording and reproducing apparatus according to claim 1, wherein the high-efficiency block encoding is orthogonal transform encoding. (3) The television signal digital recording and reproducing apparatus according to claim 1, wherein the orthogonal transform encoding is Hadamard transform. (4) The television signal digital recording and reproducing apparatus according to claim 1, characterized in that P=8 and Q=10.