JPS62172874A - Sound base band multiplex system - Google Patents

Abstract

PURPOSE:To obtain high stability, and to heighten profitability by performing a sound base band multiplex transmission with a digital signal process. CONSTITUTION:An input signal (a) in a data string of 1,350kbit/sec is time- compressed fitting for the V-BLK of a video signal, and the output (b) of the data string of 18,255Mbit/sec can be obtained. By converting the output (b) to a ternary, its output (c) becomes of 12.15Mbit/sec and of 2 bit. Since a sampling frequency conversion part 3 and a sound transmission matching filter 4 are constituted with one system of digital filter, its output (d) becomes a signal of 12.2mHz, and of 9 bits. The output (d) is multiplexed with the video signal from a MUSE encoder video processing part 15, and a signal (f) passing through a video transmission matching filter 6 becomes of 32.4mHz and of 10 bits, and becomes an analog signal at a D/A-LPF part 7, then being sent to a transmission line 30.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a high-definition television transmission system, more specifically, to a MUSE (Multiple Sub-N
yquist3aipling EnCOdiniJ
') system, it relates to the baseband multiplexing method for independent data transmitted by voice and other audio channels, and aims to unify the audio multiplexing method for broadcasting and related peripheral technologies, such as package media such as video discs and VTRs. That is.

(Prior art and problems) High-definition television transmission system, more specifically MUSE
A method for multiplexing independent data (hereinafter collectively referred to as audio) transmitted by digital audio and other audio channels in the system. Its achievements have been recognized through experiments and exhibitions at the 1985 Science Expo held in Tsukuba Gakuen City. However, RF multiplexing is difficult for package media such as video discs and VTRs, and baseband multiplexing has been considered for this purpose.

Regarding baseband multiplexing, see Ninomiya, for example.

Home video disc J using the FMUSE method by Otsuka et al.
TV technical report TEBS99-4. p15 (September 1984)
, Toyama et al., "Optical video disc for high-definition television using the rMUsE system," IEICE Technical Report I E84-73.
p. 9 (November 1984). These cited documents are based on the audio rate of current satellite broadcasting.
, 048kbit/sec is halved to 1,024kbit/sec (4-channel audio system using the current satellite broadcasting 32ktlz sampling, quasi-instantaneous compression/expansion method of 14bit → 10bit conversion) 2 channels of audio equivalent to A mode. vertical blanking period of the video signal (
Hereinafter, this will be abbreviated as V-BLK). However, V-81Kk: The time-compressed audio signal is a 21-bit NRZ code, and its sampling frequency is 16.2 MHz, which is the same as the sampling frequency of the MtJSE video signal.

This method (a) allows only two channels of i-voice to be multiplexed, and (b) the sampling frequency of 16.2 MHz has no margin in the time axis direction (jitter margin), making it difficult to use on video discs.

Although it is not preferable for VTRs etc., there is a margin in the amplitude direction (S/
There are drawbacks such as the fact that the method does not fully utilize its performance.

On the other hand, audio band compression methods have been announced both domestically and internationally, and audio pitlays can be compressed to 1,200 to 1,500 kbi.
With a t/sec density, it is possible to multiplex four channels of audio and independent data transmitted by the audio channels. (For example, Yu Takahashi's [Differential Compression PCM (DPCM) with Missing Pit Accumulation)') (Means for Solving the Problem) The inventor aimed to band-compress the audio, and increased the audio bit rate to the video sample rate of 16. 12 of 2MHz
It is equal to 1 / 1,350 kbit/sec, and the time is compressed to create a ternary NRZ code of 12.15 Mbau.
We investigated and experimented with a method of multiplexing V-BLK at 1/sec.

The audio rate is 1,350kbit/se.
We chose 4 audio channels and about 100 independent data channels for a.
In order to multiplex kbit/see, 1,200 to 1,
It is necessary to select 500 kb/Sec, and i, 350 kb it/Sec has a simple integer ratio (1:12) relationship with the image paste sample frequency of 16.2 MHz.

Furthermore, the reason why the ternary NRZ code is used is that the video signal is clamped at the gray level, and it has drawbacks such as requiring 8 GCs in 41 shifts.

The method of the present invention overcomes the drawbacks of the conventional method: (a) only two channels of i-voice can be multiplexed, (b) the audio baud rate is 16.2 MHz, which uses the video band to the limit, and (c) there is no margin for jitter. Not only does the four-channel multiplexing that is the current satellite broadcasting standard reduce the sound quality, but the jitter margin has also been increased, and the above-mentioned drawbacks have been improved.

When the audio baud rate is 12.15 Mbit/sec, the signal band is 6,075 MHz, and when the video sampling frequency is 16.2 MHz, the signal band is approximately 8 MHz.
fiz, which means that the audio signal falls within the band of the video signal. This means that D/A conversion in the encoder and A/D conversion in the decoder can be shared with the video system, and audio sampling frequency conversion, transmission matched filters, resampling, amplitude discrimination, etc. can all be performed using digital signal processing. It means that it can be done.

Therefore, in the conventional system, audio and video are separated and the audio system is performed by analog signal processing, but by changing to digital signal processing, high stability can be achieved, and economical efficiency can be improved by using IC. be able to.

(Embodiment) FIG. 1 shows a fundamental system diagram of the audio baseband multiplexing system of the present invention. Here, 1 is a time axis compression section, 2 is a binary → 3-value conversion section, 3 is a sampling frequency conversion section, 4 is an audio transmission matched filter section, 5 is a video multiplexing section, and 6 is a video transmission matched filter section , 7 is the D/A-LPF section, 8 is the L P
F-A/D section, 9 is an audio separation section, 10 is a sampling frequency conversion section, 11 is a resampling section, 12 is an amplitude discrimination section, 13 is a 3-value → 211iI conversion section, 14 is a time axis expansion section, 15 is an MUSE encoder video processing section, 16 is MU
In the SE decoder video processing section, 30 is a transmission path. In this method, the audio digital human power a is 1,350 kb.
This is a data string of i t/sec, and is time-compressed by the time axis compressor 1 in accordance with the V-BLK of the video signal to obtain an output signal that is a data string of 18.225 vbit/sec. When b is converted into ternary value by binary → ternary value converter 2, the output C is 12.15 Mbit/Sec,
It becomes a 2-bit signal.

The sampling frequency converter 3 and the audio transmission matched filter 4 are composed of one system of digital filters, and the output d thereof is a 16.2 Mllz, 9-bit digital signal. Output d of this audio transmission matched filter section 4
G, tMUsE The signal is multiplexed into a video signal by the video multiplexing section 5 together with the video signal from the encoder video processing section 15, and the output is e, and further passed through the video transmission matched filter section 6 having a cosine roll-off characteristic. “Ha3
2.4MHz, 10bit signal, D/A
- It becomes an analog signal in the LPF section 7 and is sent out to the transmission line 30. The above shows the configuration of the encoder,
The decoder does the opposite. The analog signal f''r (J is inputted to the decoder as h) sent from the encoder, and is converted to 16.2M H by the LPF-A/D section 8.
z, is converted into a 10-bit digital signal 1, and further separated into a video signal and an audio signal portion by an audio separation unit 9. (Note that the video signal is processed as a video signal by the MUSE decoder video processing section, but since it has no direct relation to the present invention, further explanation will be omitted.) In this case, the output J is 8b! (rounded to
It becomes a digital signal of 12, 15 Mbau /5 as a three-value (2 bit) signal by the amplitude discriminator 12.
It is demodulated into f30 audio signal 1. Subsequently, the 3-value → binary conversion unit 13 converts the data into 18.225 Mbit/se in step 1).
The data string l of c becomes 1, 3 by the time expansion unit 14.
The audio digital output n is 50 kbit/sec. However, it is assumed that the audio digital signal input to this system has already been interleaved and scrambled.

FIG. 2 shows in detail an embodiment of the encoder of the invention. In this part, the sampling frequency conversion section 3 and audio transmission matched filter section 4 shown in FIG. 1 are combined into one system of digital filters. 17 is a delay circuit using a D flip-flop, 18 is a coefficient unit, 19 is a switch, 20
21 is an adder circuit, 21 is a ROM section, and 22 is a quaternary counter. The sampling frequency conversion section is 12.15 MH
It converts No. 15 of Z into a 16.2MH2 signal,
The audio transmission matched filter section has a total frequency characteristic of 6.075M1 from the output of the binary to ternary conversion section to the sample section.
Matching is performed by setting lz to the cutoff frequency to obtain a cosine roll-off characteristic. First, 31a, 12
．． The audio data string of 15 Mbau/sea is 12.1
After converting into 3-level, 3-parallel, 4.05MH2 digital signals using a D flip-flop clocked at 5 MHz, each signal is delayed by a D flip-flop clocked at 4.05 MHz, and the ×1 to X8 signals are obtain. This signal is time-series xl, x2. ...x8
The order is as follows. Each signal of ×1 to ×8 is input to four coefficient multipliers 18 and weighted respectively. The four coefficient multipliers 18 are switched by the quaternary counter 22 using 16.2M1lz as a clock, so that 12.15
Mbau/sec audio data is filtered 1
This becomes a data string of 6.2 MHz. ×1 to ×8 are 8 bits
, the switch 19 is 2 bits, so if the ROM part 21 with 10 bit input is used, this part will have two ROs.
It can be realized with M and one adder, making it a very simple circuit. Note that the output d is 8 bits.

The coefficients α0 to α12 are 12.15MFf1. and 16.2M
This is the impulse response of an LPF with a cutoff frequency of 6.075M1lz sampled at the least common divisor of -48.6MHz, and the input/output r-J columns XI to X8 (+3 and yi
~y4 is expressed as shown in FIG.

FIG. 3 is a diagram showing an example of the relationship between the sampling frequency conversion used in the encoder of FIG. 2 and the input/output audio data string of the audio transmission matched filter.

Input audio data strings x1 to x8 are at 12.15MHz intervals,
The output audio data strings y1 to y4 are arranged in time series at intervals of 16.2M112. x1 to x8 and y1 to y
4 are connected by the coefficient relationship of equation (1). Here, the coefficient αO ~ α12 is 48.6M) tz sampling,
This is the impulse response of an LPF with a cutoff frequency of 6,075 MH2, and α0 is the coefficient of the center tap.

Equation (1) FIG. 4 shows in detail an embodiment of the decoder of the present invention. 21 is a ROM section, 22 is a quaternary counter, 23 is a ternary counter, 24 is a shift register, and 25 is an amplitude discriminator. This part is the sampling frequency converter 1 in Figure 1.
0 corresponds to the amplitude discrimination section 12. The 8-bit audio human input signal i is an LP with variable coefficient values controlled by a quaternary counter 22.
F (D flip 70 knob 17° coefficient unit 18. switch 1
9 and an adder 20), whose band is limited to 6,075 MHz.

Amplitude discrimination is performed by two amplitude discriminators (level comparators) 25 based on the output t of this LPF (one side of each amplitude discriminator 25a, 25b has a discrimination level p).
1 discrimination level q signal is applied), and 2bi
This output of t is a 16.2M H2 clock relay 1, but invalid data is entered once every 4 clocks, and by passing this through the shift register 24 in the subsequent stage, the lower part of the ternary counter 23 is -111111. resampled to 1
2.15MHz data rate, 31a (2
(which can be expressed in bits) becomes the audio data string 1.

FIG. 5 is a ψ diagram showing an example of the relationship between input and output data strings of the sampling frequency converter used in the decoder of FIG. 4. Input data string y1 to y8 is 16.2MHz
Input signal and output data sequence 21 arranged in time series at intervals
~z3 indicates output signals arranged in time series at 12.15 MHz intervals. y1-y8 and z1-13
are connected by the coefficient relationship of equation (2).

Here, the coefficients β0 to β8 are the impulse responses of an LPF having a cut-off frequency of 6,075 MHz sampled at 48.6 M)lz, and βO is the center tap.

Equation (2) (Effects of the Invention) The present invention performs audio baseband multiplex transmission in the MUSE system using digital signal processing. (b) Improved performance and efficiency, especially converting analog signal processing to digital signal processing to achieve higher performance and reliability; (C) Economical efficiency and simplified configuration, especially the feedback circuit of the present invention. Since the device does not have an IC, it is extremely easy to convert it into an IC, which leads to a simpler and more economical device.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the principle of an audio baseband multiplexing system according to the present invention. FIG. 2 is a diagram illustrating an audio baseband multiplexing encoder according to the present invention. FIG. 3 is a diagram showing the relationship between the input/output and voice data strings of the sampling frequency converter and voice transmission matched filter used in the encoder of the present invention. FIG. 4 is a diagram showing an audio baseband multiplexing decoder according to the present invention. FIG. 5 is a diagram showing the relationship between human output data strings of the sampling frequency converter used in the decoder of the present invention. 1... Time axis expansion unit 2... Binary → ternary value conversion unit 3.
...Sampling frequency conversion section 4...Audio transmission matched filter section 5...Video multiplexing section 6...Video transmission matched filter section 7...D/A-LPF section 8...LPF-A /,
Part 0 9...Speech separation unit 10...Sampling frequency conversion unit 11...Resampling unit 12...Amplitude discrimination unit 13.
...Three-value → binary conversion section 14... Time axis expansion section 15.
...MUSE encoder video processing section 16...MUSE
Decoder video processing unit 17...Delay circuit or D flip-flop 18...Coefficient 3 19...Switcher 20...Adder 21...ROM unit 22...Quadary counter 23...3 Advance counter 24...Shift register 25...Amplitude discriminator 3
0...Transmission line

Claims (1)

[Claims] 1. In a high-definition television transmission system based on the MUSE system, the bit rate of audio and/or independent data transmitted by an audio channel is compressed, and binary NR
An audio multiplexing system characterized in that a Z code is converted into a ternary NRZ code, and the ternary NRZ code is time-base compression multiplexed and transmitted during the vertical blanking period of a video signal. 2. The bit rate of the audio and independent data is 1,350
2. The audio multiplexing method according to claim 1, wherein the baud rate of the ternary NRZ code is 12.15 bau/sec. 3. Equipped with an audio multiplex encoder that converts the frequency of the ternary NRZ code into a video sampling frequency equal to the resampling clock rate of the high-definition television system using the MUSE system, and has a transmission matching circuit configured with a digital filter. 2. The audio multiplexing system according to claim 1, wherein the audio multiplexing method is characterized in that: 4. Claims characterized in that the frequency of the ternary NRZ code is 12.15 MHz, and the video sampling frequency equal to the resampling clock rate of a high-definition television system using the MUSE system is 16.2 MHz. The audio multiplexing method described in Section 3. 5. resampling the audio transmitted by converting it to the video sampling frequency and/or the independent data transmitted by the audio channel using a digital filter;
The audio multiplex system according to claim 1, further comprising an audio multiplex decoder that converts the audio data into a ternary NRZ code using a digital amplitude discriminator and demodulates the audio data. 6. The audio multiplexing system according to claim 5, wherein the sampling frequency of the video is 16.2 MHz, and the frequency of the ternary NRZ code is 12.15 MHz.