KR960012477B1 - Adaptable stereo digital audio coder & decoder - Google Patents

Abstract

The device includes a reference channel discriminator(100) for determinating a reference channel having large perceptual entropy in channels, the first encoder(110) for coding or coding of multiple by using a properties of auditory sense of human, the second encoder(120) for calculating an error of audio signal, a differential signal generator(130) for detecting and generating a differential signal, the third encoder(140) for calculating a perceptual error, an encoder selecting unit(160), a comparator(150) and MUX(multiplexing unit,170).

Description

Adaptive Stereo Digital Audio Coding and Decoding Device Using Cognitive Information Volume

1 is a block diagram showing a stereo digital audio encoding apparatus according to the present invention;

2 is a block diagram showing details of a first encoder shown in FIG. 1;

3 is a block diagram showing details of a second encoder shown in FIG. 1;

4 is a block diagram showing details of a third encoder shown in FIG. 1;

5 is a block diagram showing a stereo digital audio decoding apparatus according to the present invention;

* Explanation of symbols for main parts of the drawings

100: reference channel discriminator 110, 120, 140: first, second and third encoder

130: differential signal generator 150: comparator

160: encoder selection unit 170: multiplexer

510: demultiplexer 520,530: first and second decoders

540: adder 550: channel selector

560: decoder selection unit 570: digital / analog converter

The present invention relates to a digital audio encoder and a decoder. In particular, in a stereo digital audio encoding apparatus including first and second channels, a perceptual information amount for a digital audio signal input through two channels is calculated and calculated. A channel having a large amount of cognitive information is set as a reference channel and encoded, and a cognitive information amount that increases encoding and decoding efficiency and improves sound quality by encoding and decoding a difference signal between the other channel and the audio signal encoded and decoded. The present invention relates to an adaptive stereo digital audio encoding and decoding apparatus.

It is relatively narrow in high-definition television (HDTV) information transmission system, which is being developed for the purpose of reproducing sound quality signals such as compact discs (CDs) and digital audio tape recorders (DATs) that are currently in use. Since video and audio signals must be transmitted through a transmission line having a bandwidth of 6 MHz, an efficient signal compressor method has been required for audio signals as well as video signals.

To this end, high-quality digital audio signal processing technology for HDTV uses adaptive transform coding, which reflects the human auditory characteristics, to reproduce the sound quality of the aforementioned digital audio device even with a relatively simple receiver at a low data rate. Active algorithms and hardware implementation are being studied.

As the above-described adaptive change coding scheme, encoding methods for two-channel stereo audio signals are independently performed when the difference between the scale factors of the digital audio signals input through the L and R channels, that is, the two channels is large. A method of encoding or a method proposed by the Motion Picture Expert Group (MPEG) includes a method of transmitting and encoding size information of one channel when the size information of two channels is similar to each other. However, these coding schemes do not consider the signal characteristics of each channel sufficiently, and encode and decode by applying the same coding scheme for each channel, resulting in poor encoding and decoding efficiency and difficulty in realizing high sound quality.

Accordingly, an object of the present invention is to solve the above-described problem, and an object of the present invention is to calculate the amount of cognitive information of a digital audio signal input to the first and second channels, and to set and encode a channel having a large amount of cognitive information as a reference channel. In order to provide an adaptive stereo digital audio encoding and decoding apparatus using an amount of cognitive information that increases encoding and decoding efficiency and improves sound quality by encoding and decoding a difference signal between an audio signal obtained by encoding and decoding the other one channel and a reference channel, have.

According to an embodiment of the present invention to achieve the above object of the present invention, an apparatus for adaptively encoding and transmitting a digital audio signal input to the first and second channels, the first and second channels A reference channel discriminator for calculating the amount of cognitive information according to the auditory characteristics of the input digital audio signal, and determining, based on the calculated amount of cognitive information, a channel having the greater amount of cognitive information among the two channels as a reference channel; A first encoder for encoding and decoding the digital audio signal for the reference channel provided from the reference channel discriminator in consideration of human auditory characteristics; A second encoder that encodes and decodes a digital audio signal for a channel other than the reference channel provided from the reference channel discriminator, and calculates a recognition error through the digital audio signal for the other channel and the decoded digital audio signal. Wow ; A difference signal generator for detecting and outputting a difference signal between the digital audio signal decoded by the first encoder and the digital audio signal for the other channel; Encodes and outputs a difference signal provided from the difference signal generator, adds the difference signal and a decoded signal provided from the first encoder, and adds the added signal and an audio signal input to a channel other than the reference channel A third encoder for calculating cognitive error through; An encoder selection unit for selecting and transmitting the encoded digital audio signals provided by the second and third encoders; Compare the recognition error provided by the second and third encoders and provide a control signal corresponding thereto to the encoder selector so that the encoder selector receives one of the encoded digital audio signals input from the second and third encoders. A comparator to select from; And a multiplexer for multiplexing the channel selection information provided from the reference channel discriminator, the coded audio signal provided from the encoder selector and the encoder selection information provided from the comparator according to channel transmission characteristics.

According to another exemplary embodiment of the present invention, a compression-encoded digital audio signal provided from an adaptive stereo digital coding apparatus using an amount of cognitive information is received, and the digital audio signal of a reference channel and a digital audio signal of another channel and the reference and other channels are received. A demultiplexer for demultiplexing the channel selection information and the encoder selection information for each other; A first decoder configured to input a digital audio signal of the first channel detected through the demultiplexer and restore the original audio signal; A second decoder configured to input digital audio signals of the second and third channels detected through the demultiplexer and restore the original audio signals; An adder for adding the decoded digital audio signal provided by the first decoder and the digital audio signal provided by the second decoder; A channel selector for selecting and outputting an audio signal output from the first decoder, a digital audio signal output from the adder, or a digital audio signal output from the second decoder according to channel selection information provided by the demultiplexer; A decoder selector configured to switch according to encoder selector information output from the demultiplexer and to select and output an audio signal output from the adder and the second decoder; And a digital / analog converter for converting the digital audio signal output from the channel selector into an analog audio signal.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention.

1 is a block diagram showing an adaptive stereo digital audio encoding apparatus according to the present invention, which includes a reference channel discriminator 100, first, second and third encoders 110, 120 and 140, and difference. The signal generator includes a differential signal generator (DSG) 130, a comparator 150, an encoder selector 160, and a multiplexer (MUX) 170.

The reference channel discriminator 100 obtains a recognition information amount in units of frames of each channel by a method of describing digital audio signals input to the first and second channels, and then determines the reference channel in response to the recognition information amount. The amount of cognitive information is calculated first in order to calculate the amount of cognitive information of the audio signal input to the first and second channels, respectively. That is, the hearing characteristic applied by the first and a 2 N samples of the power density spectrum (S _xx (W)) of a frame of finite digital audio signal source, that is, x (n) one consisting input to the channel The masking threshold value M (w) is obtained by using the method, and the amount of cognitive information is measured by the following equation. That is, an approximate power density spectrum S _xx (w) for a finite digital audio signal source x (n) of one frame is obtained by the following equation (1).

In other words,

In addition, since M (w) described above corresponds to a power value of a region having a power density spectrum of S _xx (w) that cannot be detected by a human ear in a predetermined frequency component, If the signal is reproduced with an error value less than or equal to M (w), the area cannot be detected by the ear. Therefore, in order to obtain the power density spectrum (S _ee (w)) of the area, that is, the error signal, the input signal of the encoder Is x (n), the output signal is y (n), and the error signal e (n) is obtained by the following equation (2).

In other words,

e (n) = x (m) -Y (n) (2)

Therefore, by replacing S _ee (w) for N finite signal sources with M (w), the amount of cognitive information (R _PE ) for transmitting a signal x (n) that cannot detect errors as the human ear is given by Obtained by (3).

On the other hand, when the masking threshold M (w) uses the same band split coder having the same frequency and infinite frequency band every frame, the theoretically feasible minimum bit rate is the value obtained by the above equation (3).

However, in the case of configuring the encoder for the actual audio data, each audio parameter is analyzed for each frame time-domain signal having N samples, and then different quantization levels are used for each frequency band divided accordingly. Since the auditory parameters vary every N samples, the substantially necessary bit rate will be greater than the amount of cognitive information calculated by equation (3) above.

For example, when the input signal x (n) is divided into L frequency bands, where L is a positive integer, and encoded, the power density spectrum S _xx (i) and masking of the i th frequency band are masked. The threshold M (i) can be approximated by the following equations (4 and 5).

In other words,

here,

R ₁ is a frequency domain corresponding to the i-th division band, and S _xx (w _j ) is a value corresponding to the power density spectrum of the j-th frequency component in the N-point Discrete Fourier Transform (DFT).

Here, M (i) represents the minimum value of the masking threshold value M (w) belonging to the i-th division band.

For example, if a power density spectrum is obtained using a 1024 point DFT (that is, L = 1024) and then divided into 32 frequency bands (that is, L = 32), the cognitive information amount R _PE is expressed as follows. 6) can be obtained.

In other words,

here,

Next, the concept for calculating the amount of cognitive information of the entire frame of the first and second channels, that is, 1GOF, will be described. Recognition amount of information PE ₁ using the power density spectrum and masking threshold for the audio data of the i-th frame consisting of N samples, where i is a positive integer greater than 0 and smaller than the number of frames. Is obtained, and then the standard deviation PE indicating the change amount of the average recognition information amount PE _m for one frame group having a total of 2M frames for the first and second channels and the i-th recognition information amount PE ₁ for each PE _m described above. _std is obtained by the following expressions (7 and 8).

In other words,

In this way, the amount of cognitive information of the digital audio signal for the first and second channels is obtained, and a channel having a large amount of cognitive information is set as a reference channel and provided to the first encoder 110 to be described later. The audio signal is provided to the second encoder 120, the DSG 130, and the third encoder 140.

The first encoder 110 encodes (210) and decodes (220) a digital audio signal of a reference channel having a large amount of cognitive information among the inputted first and second channels by a conventional encoding method as shown in FIG. The encoded audio signal is provided to the multiplexer 170 to be described later, and the decoded audio signal is provided to the DSG 130 and the third encoder 140 to be described later.

The second encoder 120 encodes and decodes an audio signal (x ₁ (n)) for a channel having a small amount of cognitive information, not a reference channel, by using a method described below, and also encodes a digital audio input signal by a human auditory characteristic. An optimized masking threshold is obtained, and the recognition error of the difference signal between the audio input signal and the decoded audio signal is measured and transmitted to the comparator 150 to be described later. The encoded audio signal is further described below. Forward to 160.

Referring to FIG. 3, FIG. 3 shows a detailed block diagram of the second encoder 120 shown in FIG. 1, which is an encoder 310 encoding an audio input signal x (n), Decoded by a device that performs the same operation as the decoder configured in the digital audio decoding system compatible with the encoder 310 of the encoding system and the signal transmission is provided to the cognitive error measurer (PEM) 340 to be described later And an audio input signal x (n) to measure a recognition error between the audio input signal x ₂ (n) and the decoded audio signal y ₂ (n) as described above. A masking threshold calculation unit 330 for calculating a power density spectrum for the < RTI ID = 0.0 >),< / RTI > and calculating an optimized threshold value M (w) using human auditory characteristics, and an audio input signal x (n). ) and the decoded audio signal (y ₂ (n)) Seeking in the power density spectrum of the difference signal, and includes a PEM (340) for measuring a becomes equal to or greater than the threshold error optimize spectral components.

Referring back to FIG. 1, the DSG 130 is generated between the decoded audio signal y ₁ (n) provided from the first encoder 110 and the audio input signal x ₂ (n) for the second channel. The difference signal d (n) is extracted as shown in Equation (9) and provided to the third encoder 140.

In other words,

d (n) = y ₁ (n) -x ₂ (n) (9)

Referring to FIG. 4, FIG. 4 is a detailed block diagram of the third encoder 140 illustrated in FIG. 1, and includes an encoder 410, a decoder 420, an adder 430, and a masking threshold calculator 440. ) And PEM 450.

The encoder 410 encodes the difference signal d (n) provided by the difference signal generator 130 described above by a conventional encoding method. The decoder 420 is a device having the same function as the decoder shown in FIG. 2, and the adder 430 is the audio signal y ₁ (n) and the difference signal (decoded through the decoder 220 of the first encoder). d (n)) is added to the PEM 450 by adding the decoded and decoded audio signal y _d (n) as shown in Equation (10).

y ₃ (n) = y ₁ (n) + y _d (n) (10)

The masking threshold calculator 440 calculates the power density spectrum of the audio input signal to a channel other than the reference channel in the same manner as described above, and then uses this value to optimize the threshold value M (w). ) Is calculated and provided to the PEM 450. The PEM 450 calculates a power density spectrum and a masking threshold for the difference signal between the audio signal y ₃ (n) provided by the adder 430 and the audio input signal x ₂ (n) for another channel. The recognition error is measured and output in the same manner as described above using the optimized threshold value M (w) provided by the unit 440.

Referring back to FIG. 1, when the recognition error of the audio input signal of the reference channel is PE ₁ and the recognition error of the channel other than the reference channel is PE ₂ , the difference signal d (n) described above is the reference channel audio. The audio signal y ₁ (n) encoded and decoded from the input signal and the other channel audio input signal x ₂ (n) can be calculated by the above equation (9), and the calculated difference signal d If (n)) is encoded and decoded as y _d (n), the signal y ₃ (n) calculated by the above equation (10) actually decodes the audio input signal for the second channel. It becomes a signal. If the audio input signals for the first and second channels are highly correlated, the signal y ₃ (n) is the second channel audio input than the audio signal after encoding and decoding the second channel audio signal. More correlated with the signal.

The comparator 150 compares the cognitive errors PE ₂ and PE ₃ provided by the second and third encoders 120 and 140 described above, and outputs the encoded signals output from the encoders 120 and 140 which generate less cognitive error. The control signal and the second and third encoders 120 and 140 are output to the encoder selector 160 so that the audio signal can be selected and transmitted. For example, a recognition error provided by the second encoder 120 may be used. If PE ₂ is greater than the recognition error PE ₃ provided by the third encoder 140, the encoder selector 160, which will be described later, selects an audio signal encoded by the third encoder 140. It outputs '1', and in the opposite case, outputs logic '0'. Accordingly, the encoder selector 160 selects the encoded audio signals output from the second and third encoders 120 and 140 and provides them to the multiplexer 170 in response to the control signal provided from the comparator 150.

The multiplexer 170 multiplexes the coded audio signal provided from the encoder selector 160 and the encoder selection information provided from the comparator 150, converts the encoded audio signal into a bit stream suitable for channel characteristics, and outputs the converted bit stream.

5 is a schematic block diagram of a digital audio decoding apparatus according to the present invention. The decoding system includes a demultiplexing unit 510, first and second decoders 520 and 530, an adder 540, and a decoder selection. A unit 560, a channel selector 550, and a digital / analog (D / A) converter 570, 580 are included.

The demultiplexing unit 410 is an audio signal and reference channel selection information for the audio signal, i.e., the reference channel and other channels, before the multiplexing of the bit stream transmitted through the channel after encoding by the above-described encoding apparatus. The audio signal of the reference channel and other channels, which are main information, is provided to the first and second decoders 520 and 530 to be described later, and the reference channel selection information, which is additional information, is provided to the channel selector 550 to be described later. do.

The first and second decoders 520 and 530 decode by a well-known conventional decoding scheme. The audio signal of the first decoder 520 is provided to the digital / analog converter 450 and the adder 440. The audio signal decoded by the second decoder 430 described above is provided to the terminal 2 of the adder 540 and the decoder selector 560. The adder 540 adds the audio signals provided from the first and second decoders 520 and 530 to provide them to the terminal 1 of the decoder selector 560 described above, and the decoder selector 560 described above. The demultiplexer 510 described above in one of the composite audio signal to which the audio signals from the first and second decoders 520 and 530 output from the adder 540 are added and the audio signal provided from the second decoder 530. The same signal as the encoder selection information selected by the encoding apparatus is selected by the predetermined encoder selection information provided by the reference signal.

The channel selector 550 selects a signal output from the first decoder 520 and the decoder selector 560 based on the reference channel selection information output from the demultiplexer 510, and the D / A converter 570 to be described later. ) And 580.

Finally, the D / A converters 570 and 580 convert the digital audio signals provided from the first decoder 420 and the channel selector 550 into analog audio signals, respectively, and output the analog audio signals.

As described above, according to the adaptive stereo digital audio encoding and decoding apparatus according to the present invention, an audio signal input by setting a channel having a large amount of recognition information among the audio signals for the first and second channels as a reference channel is used. Encoding and transmitting, and after decoding the audio signal input to the other channel and the encoded audio signal again, by encoding, decoding and transmitting the resulting difference signal, a large effect to increase the encoding and decoding efficiency and improve the sound quality There is.

Claims (4)

An apparatus for adaptively encoding and transmitting digital audio signals input to first and second channels, wherein the digital audio signals input to the first and second channels are calculated according to human auditory characteristics. A reference channel discriminator 100 for determining a channel having a larger amount of recognition information among the two channels as a reference channel according to the calculated recognition information amount; A first encoder (110) for encoding and decoding the digital audio signal for the reference channel provided from the reference channel discriminator (100) in consideration of human hearing characteristics; A digital audio signal for a channel other than the reference channel provided from the reference channel discriminator 100 is encoded and decoded, and a cognitive error is calculated through the digital audio signal for the other channel and the decoded digital audio signal. A second encoder 120; A differential signal generator 130 for detecting and outputting a differential signal between the digital audio signal decoded by the first encoder 110 and the digital audio signal for the other channel; The difference signal provided from the difference signal generator 130 is encoded and outputted, and the difference signal and the decoded signal provided from the first encoder 110 are added, and the added signal and a channel other than the reference channel are added. A third encoder 140 that calculates a recognition error through an audio signal inputted into the third encoder; An encoder selector 160 for selecting and transmitting the encoded digital audio signals provided from the second and third encoders 120 and 140; The recognition errors provided by the second and third encoders 120 and 140 are compared with each other, and a control signal corresponding thereto is provided to the encoder selector 160 so that the encoder selector 160 may use the second and third encoders ( A comparator 150 for selecting one of the coded digital audio signals input from 120 and 140; Multiplexing the channel selection information provided from the reference channel discriminator 100, the coded audio signal provided from the encoder selecting unit 160, and the encoder selection information provided from the comparator 150 to suit channel transmission characteristics. Adaptive stereo digital audio encoding apparatus using the amount of cognitive information including a multiplexer 170. The digital audio signal of claim 1, wherein the second encoder 120 band-divides the digital audio signal for a channel other than the reference channel into a predetermined bandwidth, and masks the input digital audio signal by audible characteristics. An encoder 310 which measures and allocates and encodes a quantization bit in each of the divided bands by this value; A decoder (320) for decoding the encoded digital audio signal; A masking threshold calculator 330 for calculating a threshold optimized by the human auditory characteristics of the input digital audio signal; A cognitive error for measuring a cognitive error with respect to the difference signal generated between the decoded digital audio signal y ₂ (n) and the input digital audio signal x ₂ (n), using the calculated optimized threshold value. Adaptive stereo digital audio encoding apparatus using the amount of cognitive information including a measuring unit (340). 2. The apparatus of claim 1, wherein the third encoder (140) comprises: an encoder (410) for encoding the difference signal (d (n)) provided by the difference signal generator (130); A decoder 420 for decoding the encoded audio signal; An adder for adding the decoded digital audio signal y ₁ (n) provided by the decoder 220 of the first encoder 210 and the digital audio signal y _d (n) provided by the decoder 420. 430; A masking threshold calculator 440 for calculating a digital audio signal (x ₂ (n)) input to a channel other than the reference channel optimized by human auditory characteristics; Recognition error using the calculated optimized threshold for the difference signal generated between the added digital audio signal y ₃ (n) and the digital audio signal x ₂ (n) of a channel other than the reference channel Adaptive stereo digital audio encoding apparatus using the amount of cognitive information comprising a cognitive error measurer 450 for measuring the. Receives a compression-encoded digital audio signal provided from an adaptive stereo digital encoding apparatus using a cognitive information amount, and receives a digital audio signal of a reference channel and a digital audio signal of another channel, channel selection information and encoder selection information of the reference and other channels. A demultiplexing unit 510 for demultiplexing to detect a signal; A first decoder 520 for inputting the digital audio signal of the first channel detected through the demultiplexer 510 and restoring the original signal; A second decoder 530 for inputting the digital audio signals of the second and third channels detected by the demultiplexer 510 and restoring the original signals; An adder (540) for adding the decoded digital audio signal provided from the first decoder (520) and the digital audio signal provided from the second decoder (530); The digital audio signal output from the first decoder 520 and the digital audio signal output from the adder 540 or the second decoder 530 according to the channel selection information provided by the demultiplexer 510. A channel selector 550 for selecting and outputting a digital audio signal to be output; It is composed of a switching element which is switched according to the encoder selection information output from the demultiplexer 510 and selects an audio signal output from the adder 540 and the second decoder 530 and provides it to the channel selector 550. A decoder selector 560 to be used; And a digital / analog converter (570,580) for converting the digital audio signal output from the channel selector (550) into an analog audio signal.