KR960012478B1 - Adaptable ms-stereo digital audio coder & decoder - Google Patents

Abstract

The device includes a middle MS matrix unit(310) for generating the second MS stereo digital audio signal, a similarity calculating unit(320) for calculating a value of mutual similarity of the audio signals by the first and second channel, a stereo mode selector(340) for selecting and generating one of the first and second MS stereo digital audio signal, a stereo mode selecting controller(330) for supplying the mode information from the stereo mode selector(340), the first and second coding unit(350,360) for coding the digital audio signals, and multiplexer(370).

Description

MS Stereo Digital Audio Encoding and Decoding Device Adaptive to the Mutual Similarity between Audio Signals

1 is a schematic block diagram of a conventional Middle Side (MS) stereo digital audio encoding apparatus.

2 is a schematic block diagram of a conventional MS stereo digital audio decoding apparatus.

3 is a block diagram showing an adaptive MS stereo digital audio coding apparatus of the present invention.

4 is a block diagram showing an adaptive MS stereo digital audio decoding apparatus of the present invention.

* Explanation of symbols for main parts of the drawings

310: MS matrix unit 320: mutual similarity calculation unit

330: stereo mode selection unit 340: stereo mode selection unit

350, 360: First and second encoders 370: Multiplexer

410: demultiplexer 420,430: first and second decoder

440: Adaptive MS selector 450,460: D / A converter

The present invention relates to an MS stereo digital audio encoding and decoding apparatus adapted to mutual similarity of audio signals, and in particular, according to mutual similarity of audio signals input to first and second channels, a middle side (MS) stereo or The present invention relates to an MS stereo digital audio encoding and decoding apparatus for determining an independent stereo mode and adaptively encoding and decoding accordingly to increase encoding and decoding efficiency.

Currently, high definition television (HDTV) information transmission systems, which are being developed for the purpose of reproducing sound quality signals such as compact discs (CDs) and digital tape recorders (DATs), which are in practical use, are relatively narrow. Since video and audio signals must be transmitted through a transmission line of about 6 MHz, an efficient signal compression technique has been required for audio signals as well as video signals.

To this end, it is a high-quality digital audio signal processing technology for HDTV that uses the adaptive transform coding method that reflects the human auditory characteristics to reproduce the above-mentioned sound quality of the digital audio device even with a relatively simple receiver at a low data rate. It is actively researched for possible algorithm and hardware implementation.

With the above-described adaptive conversion or encoding technique, the sum and difference of two-channel signals for transmitting stereo signals in the first and second two channels (or L and R channels) to the conventional conventional stereo digital audio encoding and decoding apparatus. There is an MS stereo system that transmits signals, that is, middle and side signals, and then reproduces the first, second, and two channel signals by separating them in a reverse process from the receiving side.

The conventional MS stereo encoding apparatus described above is shown in FIG.

As shown in FIG. 1, the A / D converters 110 and 120 respectively output analog audio signals x (t) and y (t) input to the first and second channels, respectively. n) and y (n)) to provide to the MS matrix calculator 130. The adder 131 of the matrix calculator 130 subtracts the digital signals x (n) and y (n) output from the A / D converters 110 and 120 and provides them to the multiplier 132. The subtractor 133 subtracts the digital signals x (n) and y (n) output from the A / D converters 110 and 120 and provides them to the multiplier 132. The multiplier 132 outputs a predetermined value to the output of the adder 131, for example. Is multiplied and provided to the first encoder 140, and the multiplier 134 outputs a predetermined value to the output of the subtractor 133, for example. Multiply by and give it to the second encoder. The first and second encoders 140 and 150 encode the outputs of the multipliers 132 and 134, respectively, and provide them to the multiplexer 160. The multiplexer 160 outputs the first and second encoders 140 and 150, respectively. The coded signals M and S are multiplexed and transmitted to the decoding side.

Meanwhile, as shown in FIG. 2, the MS stereo decoding apparatus reversely operates with respect to the above-described encoding apparatus, and includes the demultiplexer 210, the first and second decoders 220 and 230, and the MS matrix unit 240. ) And D / A converters 250 and 260.

The demultiplexer 210 demultiplexes the encoded bit stream as described above to provide the encoded signal M to the first decoder 220 and the encoded signal N to the second decoder 230. The first decoder 220 decodes the M signal provided from the demultiplexer 210 through the inverse process of the first encoder 140, and the second decoder 230 is provided from the demultiplexer 210. The signal is decoded through the reverse process of the second encoder 150 and provided to the MS matrix unit 24.

The MS matrix unit 240 adds the output of the first decoder 220 and the output of the second decoder 230 to the result. Multiplying and providing the result to the D / A converter 250, subtracting the output of the second decoder 230 from the output of the first decoder 220, Is multiplied and provided to the D / A converter 260. Therefore, the D / A converters 250 and 260 convert the analog signals to the L and R channels, respectively.

The conventional MS stereo coding apparatus configured and operated in this manner has several advantages over the method of encoding independently by concentrating and encoding the audio data into one channel when the digital audio signals inputted through the L and R channels have large similarities. There is this. However, if the audio signals inputted through the two channels are inferior in mutual similarity, the coding efficiency is reduced because the separated middle and side channel signals have similar data amounts when they are encoded by the MS stereo coding method.

However, the MS stereo encoding device and the encoding device having the above-described form are encoded after performing an MS matrix unconditionally regardless of mutual similarity of audio signals inputted through two channels, and decode the encoded audio signal. There is a problem of lowering the decoding efficiency.

Accordingly, an object of the present invention is to solve the above problems, and an object of the present invention is to measure mutual similarity between audio signals input to two (first and second or L and R) channels, and then, in response to mutual similarity, If the similarity is high, MS stereo mode is encoded. If the similarity is low, it is encoded in independent stereo mode and transmitted. Then, the MS stereo digital is adaptive to the mutual similarity of the audio signal which increases the encoding and decoding efficiency by decoding corresponding to each mode. An audio encoding and decoding apparatus is provided.

According to an embodiment of the present invention, the first and second independent stereo digital audio signals for the first and second channels, which are digitally converted and inputted at a predetermined sampling frequency, are added to and multiplied by a predetermined value to form the first MS stereo digital audio. A middle side for generating a signal and subtracting a second independent stereo digital audio signal from a first independent stereo digital audio signal for the first and second channels and multiplying it by a predetermined value to generate a second MS stereo digital audio signal; MS) matrix part; A similarity calculator for calculating mutual similarity values of the audio signals for the first and second channels by inputting the audio signals for the first and second channels; A stereo mode selector configured to select and output one of the first and second independent stereo digital audio signals directly input to the first and second channels or the first and second MS stereo digital audio signals provided by the MS matrix unit; Wow ; If the mutual similarity value is greater than a preset value in response to the mutual similarity value provided by the similarity calculator, the stereo mode selector causes the first and second MS stereo digital audio signals to be selected, and the mutual similarity value is a preset value. A stereo mode selection control unit providing all the predetermined information to the stereo mode selection unit to cause the stereo mode selection unit to select the first and second independent stereo digital audio signals when smaller or equal to each other; First and second encoders for encoding the first and second independent stereo digital audio signals or the first and second MS stereo digital audio signals selected and provided by the stereo mode selection unit; And a multiplexer for multiplexing and outputting the encoded audio signals provided from the first and second encoders and the mode information provided from the stereo mode selection controller.

According to another embodiment of the present invention, a demultiplexer for demultiplexing encoded MS stereo digital audio signals or independent stereo digital audio signals of first and second channels and mode information for determining the stereo encoding mode; First and second decoders for decoding the first and second channels of the encoded MS stereo digital audio signal or the independent stereo digital audio signal provided by the demultiplexer into an original digital audio signal; After determining whether the signal output from the first and second decoders is an MS stereo digital audio signal or an independent stereo digital audio signal based on the mode information, the decoded stereo digital audio output from the first and second decoders. If the signal is an MS stereo digital audio signal, the output of the first decoder and the output of the second decoder are added, and the result is multiplied by a first set value and outputted, and the second decoder is output from the output of the first decoder. Subtracts the output of the output signal and multiplies the result by the second set value, and outputs the first decoder and the first decoder when the decoded stereo digital audio signal output from the first and second decoders is an independent stereo digital audio signal. An adaptive MS selector for directly outputting the output of the decoder; And first and second digital / analog converters for converting the decoded digital audio signal provided by the adaptive MS selector into an analog signal and outputting the analog signal.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the present invention.

3 is a block diagram illustrating an MS stereo encoding apparatus adaptive to mutual similarities between audio signals according to the present invention. The encoding apparatus includes an MS matrix unit 310, a similarity calculating unit 320, and a mode selection controller 330. ), A mode selector 340, first and second encoders 350 and 360, and a multiplexer 370. By such a device, encoding efficiency can be improved by determining and encoding one of an independent stereo mode and an MS stereo mode. The MS matrix unit 310 performs the MS matrix in the same way as the digital audio signals x (n) and y (n) are applied to the first and second channels in the conventional MS stereo encoding apparatus described above. (n)) and side (S (n)) channel signals are generated and provided to the stereo mode selector 340 to be described later.

That is, the MS matrix unit 310 adds the digital audio signal x (n) of the first channel and the digital audio signal y (n) of the second channel, and then adds the result to the result. The stereo mode selector 340 is provided to multiply the value m (n), and the digital audio signal y (n) of the second channel is subtracted from the digital audio signal x (n) of the first channel. The stereo mode selector 340 is provided by multiplying the value s (n).

Meanwhile, the similarity calculator 320 uses both of the signals x (n) and y (n) by the technique described below with respect to the digital audio input signals x (n) and y (n) for the first and second channels. The similarity between the two is calculated and transmitted to the stereo mode selection controller 330 to be described later.

The mutual similarity calculation first calculates mutual similarity of an input signal composed of N samples (where N is a positive integer) in one frame for each of the first and second channels. ), assuming that the average of y (n) is 0, the mutual similarity (C) is obtained.

In other words, the mutual similarity (C) is calculated by [Equation 1].

[Equation 1]

Here, E {·} is an expectation value and means a stochastic mean.

The calculation of the expected value E {·} in Equation (1) can be approximated by using an arithmetic mean using finite number of data.

A method of calculating the mutual similarity of [Formula 1] using N pieces of data is shown in [Formula 2].

[Formula 2]

In the general case where the mean is not 0 for N samples, mutual similarity (C) is obtained by Equation 3.

[Equation 3]

Here, mean (·) represents an arithmetic mean, and can be obtained by [Equation 4].

[Equation 4]

If the average of x (n) and the average of y (n) are Mx and My, respectively, the mutual similarity (C) of the above-mentioned [Equation 3] can be expressed as [Equation 5].

[Equation 5]

The mutual similarity of Equation 5 theoretically has a value between -1 and +1, but may be out of this range because it is approximately calculated.

However, in the present invention, it is set to -1 if less than -1, and to +1 if greater than +1.

The large absolute value of the mutual similarity (C) calculated from the above [Equation 5] means that the mutual similarity of x (n) and y (n) is large, and the small absolute value of the mutual similarity (C) is small. It means that the mutual similarity of x (n) and y (n) is small. That is, in an exemplary embodiment of the present invention, if the absolute value of the mutual similarity (C) between the digital audio signals of the two channels obtained through Equation 5 is less than 0.5, the digital audio signals of the respective channels are encoded in an independent stereo mode. However, if the absolute value of the similarity is greater than 0.5, the mutual similarity is large, so that the digital audio signal of the two channels is encoded in MS stereo mode.

For example, when a digital audio signal sampled at 48 KHz for the first channel and the second channel is x (n) and y (n), if one frame is set to 1152 samples, the mutual similarity The value (C) can be expressed as shown in the following [Formula 6] by using the above-described formula (5).

[Equation 6]

Here, Mx and My are represented by the following [Formula 4] by the following [Formula 4].

[Formula 7]

Therefore, the stereo mode selection control unit 330 is a stereo for describing the following mode information (logic 1 or 0) such as [Equation 8] based on the mutual similarity (C) value calculated from the above [Equation 6] and [Equation 7]. The mode selector 340 and the multiplexer 370 are provided.

Namely, if C | 0.5, MS stereo,

If C | ≤ 0.5, independent stereo. Equation 8

The stereo mode selector 340 is directly provided with independent stereo digital audio signals x (n) and y (n) from the first and second channels, and the MS stereo audio signal m (n) is provided through the MS matrix unit 310. ), s (n) is provided. The stereo mode selector 340 selects the independent stereo digital audio signal x (n) or y (n) based on the mode information provided from the stereo mode selector 330, or the MS stereo audio signal m (n), s (n) is selected and output to the first and second encoders 360 to be described later. That is, when the stereo mode selector 340 selects the independent stereo digital audio signals x (n) and y (n), the independent stereo digital audio signals x (n) will be provided to the first encoder 350, and independent The stereo digital audio signal y (n) will be provided to the second encoder 360.

In addition, when the stereo mode selector 340 selects the MS stereo audio signals m (n) and s (n), the MS stereo audio signals m (n) will be provided to the first encoder 350, and the MS stereo audio signals s (n) will be provided to the second encoder 360. Accordingly, the first and second encoders 350 and 360 encode the stereo audio signals provided from the stereo mode selector 340 and provide them to the multiplexer 370.

The multiplexing unit 370 receives the encoded digital audio signals provided from the first and second encoders 350 and 360 and the mode information provided from the stereo mode selection control unit 330 and then fits the channel characteristics. Multiplex the output bit stream.

4 is a block diagram illustrating an adaptive MS stereo decoding apparatus according to the present invention, which includes a demultiplexer 410, first and second decoders 420 and 430, an adaptive MS selector 440, and a D / A. Converters 450 and 460. The demultiplexer 410 receives an encoded bit stream transmitted from an encoding apparatus as shown in FIG. 3, and then selects an MS stereo or independent stereo digital audio signal from the bit stream and selects two stereo methods selected from the above-described encoding apparatus. Demultiplex each piece of information. The demultiplexer 410 then provides an MS stereo or independent stereo digital audio signal to the first and second decoders 420 and 430, and provides selection mode information to the adaptive MS selector 440, respectively.

The first and second decoders 420 and 430 decode the encoded stereo digital audio signal provided from the demultiplexer 410 to the adaptive MS selector 440 by using a known decoding method.

The adaptive MS selector 440 determines the encoding mode of the decoded stereo digital audio signal output from the first and second decoders 420 and 430 from the stereo mode information, and then processes the signal according to the corresponding D / A converter. Provided at (450,460). That is, when the decoded stereo digital audio signals output from the first and second decoders 420 and 430 are MS stereo digital audio signals, the output of the first decoder 420 and the output of the second decoder 430 are added. After multiplying the result by t14 and providing it to the D / A converter 460, subtracting the output of the second decoder 430 from the output of the first decoder 420, and multiplying the result by t15 to obtain the D / A. To the converter 460. In addition, when the decoded stereo digital audio signal output from the first and second decoders 420 and 430 is an independent stereo digital audio signal, the output of the first decoder 420 is provided to the D / A converter 450. The output of the double decoder 430 is provided to the D / A converter 460.

Finally, the D / A converters 450 and 460 convert the digital audio signals for the L and R channels provided by the adaptive MS selector 440 into analog signals.

As described above, according to the adaptive MS stereo encoding and decoding apparatus according to the present invention, MS stereo and independent stereo modes are determined and encoded in response to mutual similarity values of audio signals input to the first and second two channels. And according to the decoding there is a great advantage to increase the encoding and decoding efficiency.

Claims (3)

A first MS stereo digital audio signal is generated by adding first and second independent stereo digital audio signals for the first and second channels which are digitally converted and inputted at a predetermined sampling frequency and multiplied by a predetermined value, and generating the first MS stereo digital audio signal. A middle side (MS) matrix unit 310 for generating a second MS stereo digital audio signal by checking the first and second independent stereo digital audio signals for the second channel and multiplying the first and second independent stereo digital audio signals by a predetermined value; A similarity calculator 320 for inputting audio signals for the first and second channels to calculate mutual similarity values of the audio signals for the first and second channels; Stereo mode selection for selecting and outputting one of the first and second independent stereo digital audio signals directly input to the first and second channels and the first and second MS stereo digital audio signals provided by the MS matrix unit 310. Section 340; In response to the mutual similarity value provided by the similarity calculator 340, when the mutual similarity value is greater than a preset value, the stereo mode selector 340 selects the first and second MS stereo digital audio signals. When the mutual similarity value is less than or equal to a preset value, the stereo mode selector 340 sends predetermined mode information to the stereo mode selector 340 so that the stereo mode selector 340 selects the first and second independent stereo digital audio signals. Providing a stereo mode selection control unit 330; First and second encoding (350,360) for encoding the first and second independent stereo digital audio signals or the first and second MS stereo digital audio signals selected and provided by the stereo mode selection unit (340); The mutual similarity between the audio signal provided by the first and second encoding units 350 and 360 and the audio signal including a multiplexer 370 which multiplexes and outputs mode information provided by the stereo mode selection control unit 330. Adaptive MS Stereo Digital Audio Coding Device. The method of claim 1, wherein the similarity calculation unit 320 is set to the similarity (C) value by the following equation (Equation 5), [Equation 5] Wherein Mx and My represent average values of the digital stereo audio signals x (n) and y (n), and xi and yi represent values for sequentially adding to the input digital stereo audio signal. MS stereo digital audio encoding apparatus adapted to mutual similarity between signals. A demultiplexer 410 for demultiplexing the encoded MS stereo digital audio signal or the independent stereo digital audio signal of the first and second channels and the mode information for determining the stereo encoding mode; First and second decoders 420 and 430 which decode the encoded MS stereo digital audio signal or the independent stereo digital audio signal of the first and second channels provided by the demultiplexer 410 into an original audio signal. )Wow ; After determining whether the signals output from the first and second decoders 420 and 430 are MS stereo digital audio signals or independent stereo digital audio signals based on the mode information, the first and second decoders 420 and 430. If the decoded stereo digital audio signal outputted from the N / A signal is an MS stereo digital audio signal, the output of the first decoder 420 and the output of the second decoder 430 are added, and the first set value is multiplied by the result. Output the first decoder 420 by subtracting the output of the second decoder 430 from the output of the first decoder 420, multiplying the second set value by the result, and outputting the first and second decoders 420. An adaptive MS selector 440 for directly outputting the outputs of the first decoder 420 and the second decoder 430 when the decoded stereo digital audio signal output from the 430 is an independent stereo digital audio signal; MS adaptive to mutual similarity between audio signals including first and second digital / analog converters 450 and 460 which convert the decoded digital audio signal provided by the adaptive MS selector 450 into an analog signal and output the analog signal. Stereo digital audio decoding device.