KR100880995B1 - Audio encoding apparatus and audio encoding method - Google Patents

Abstract

The present invention provides an audio encoding apparatus and an audio encoding method capable of appropriately selecting the length of a block while reducing the throughput. The power calculator 402 calculates the power change ratio from the input signal, the predicted gain change ratio calculator 406 calculates the predicted gain change ratio from the input signal, and the block length determination unit 407 determines the power change ratio and the predicted gain. From the change ratio, it is determined whether to encode by long block or by short block. Based on this determination, the long block MDCT change unit 409 or the short block MDCT change unit 410 is input. Discrete cosine transform is performed on the signal.

Framer, Pre-Echo, Quantizer, Power Variable Ratio, Power Calculator, Selector

Description

AUDIO ENCODING APPARATUS AND AUDIO ENCODING METHOD}

The present invention relates to an audio encoding apparatus and an audio encoding method for encoding an audio signal.

In recent years, communication fields such as the Internet and satellite broadcasting have been rapidly spreading. In addition, AV equipment such as DVD is also rapidly spreading. With the dissemination of these devices, there is a growing demand for audio coding that efficiently compresses audio signals. In the recent audio coding apparatus, the adaptive conversion audio coding apparatus using human auditory characteristics is mainstream. The basic encoding process of the adaptive transform audio encoding apparatus is as follows.

In this encoding process, the audio signal in the time domain is converted into the frequency domain. The signal on the frequency axis is then divided into frequency bands corresponding to hearing frequency resolution capability. The optimal amount of information necessary for encoding in each frequency band is calculated using human auditory characteristics.

Then, the signal on the frequency axis is quantized in accordance with the amount of information allocated to each frequency band. Among the adaptive conversion audio encoding apparatuses, there is a Moving Picture Expens Group (MPEG) -2 Advanced Audio Coding (AAC) scheme standardized by the International Organization for Standardization (ISO) / International Electrotechnical Commission (IEC). This method is also employed in BS digital broadcasting. This method is recently attracting attention as an audio encoding apparatus capable of realizing high sound quality at a low bit rate.

<First Prior Art>

Fig. 10 is a block diagram showing the configuration of an encoder of MPEG-2 AAC, which is the first conventional technology. Hereinafter, the technique shown in FIG. 10 is called 1st prior art. The detail of an AAC encoder is described in detail in the following nonpatent literature 1, for example.

The AC encoder divides the input signal into frames consisting of a predetermined number of samples.

The AAC encoder then performs encoding processing for each frame. There are two types of frame lengths of the AAC system: long blocks (1024 samples) and short blocks (128 samples). Here, the length of one frame and one block is the same. The following description is the processing procedure of the AAC encoder shown in FIG.

(1) First, an input signal is input to the framer 1001. The framer 1001 divides the input signal into frames (long blocks) consisting of a predetermined number of samples. The signal output from the framer 1001 is input to a long block discrete cosine transform unit (hereinafter simply referred to as an MDCT transform unit) 1002 and a short block MDCT transform unit 1003.

The long block MDCT conversion unit 1002 performs MDCT conversion of 1024 points on the input signal. Then, the long block MDCT converter 1002 calculates the MDCT coefficient MDCT1. In addition, the block MDCT conversion unit 1003 performs MDCT conversion of 128 points on the input signal. Then, the short block MDCT converter 1003 calculates the MDCT coefficient MDCT2. In addition, since there are eight blocks per frame, eight sets of MDCT2 are generated.

(2) Next, the framer 1001 outputs the divided input signal to the psychoacoustic hearing analyzer 1004 for the long block. In addition, the psychoacoustic hearing analysis unit 1004 for the long block obtains the masking threshold value Th1 for the long block and the psychoacoustic entropy PE1 from the input signal. Here, the calculation method of Th1 and PE1 is a well-known method as described in the psychoacoustic model part of nonpatent literature 1. Similarly, the framer 1001 outputs an input signal divided into frames to the psychoacoustic auditory analyzer 1005 for a short block. The psychoacoustic auditory analysis unit 1005 for the short block obtains the masking threshold value Th2 and the psychoacoustic entropy PE2 for the short block from the input signal.

Psychological auditory entropy is an information amount indicating a minimum number of bits necessary for quantizing a signal. In addition, masking refers to the phenomenon that a human cannot perceive the error, if it is below the reference | standard with the error at the time of quantizing a signal by a quantization part. In addition, a reference value indicating a limit of error that cannot be perceived by humans is called a masking threshold.

(3) PE1 and Th1 obtained from the long block and PE2 and Th2 obtained from the short block are input to the block length determination unit 1006. The block length determination unit 1006 determines which of the long block and the short block to quantize.

In general, it is desirable to quantize normal signals with little change in properties into long blocks. However, when a signal whose amplitude changes abruptly in a block is quantized into a long block, noise called a pre-echo that does not exist in the input signal occurs. The occurrence of this noise is caused by sound quality deterioration. 11 is a schematic diagram showing an example of pre-echo. FIG. 11A is a schematic diagram showing an input signal before encoding, and FIG. 11B is a graph showing a decoding sound when only a long block is encoded. At the head of FIG. 11B, noise that is not present in the input signal occurs before the attack sound.

This noise is called pre echo. The pre echo can be eliminated by shortening the quantization block length. Therefore, in the AAC system, the block length determination unit 1006 determines the nature of the input signal. The block length determination unit 1006 then determines a block length that is optimal for quantization. Specifically, the block length determining unit 1006 selects the long block if PE1> PE1_thr, and otherwise selects the short block. PE1_thr is a predetermined threshold value (constant).

(4) The determination result of the block length determination unit 1006 is output to the selector 1007 for selecting MDCT. In addition, the masking threshold value selected by the block length determination unit 1006 is output to the spectrum quantization unit 1008. In other words, when the block length determination unit 1006 selects the long block, the MDCT1 and Th1 are input to the spectrum quantization unit 1008. In addition, when the block length determination unit 1006 selects a short block, MDCT2 and Th2 are input to the spectral quantization unit 1008.

(5) The spectral quantization unit 1008 quantizes MDCT coefficients for each frequency band according to the input masking threshold value. The spectral quantization unit 1008 then outputs a quantization code 1.

(6) The quantized code 1 output from the spectral quantizer 1008 is input to the Huffman encoder 1009. The Huffman encoder 1009 converts the quantized code 1 into the quantized code 2 from which redundancy has been removed more than that of the quantized code 1.

(7) The quantization code 2 is output from the Huffman coding unit 1009 to the quantization control unit 1011. The quantization control unit 1011 then calculates the total number of bits of the finally output bit stream from the input quantization code 2. In addition, the range enclosed by the dotted line in FIG. 10 is the range which the quantization control part 1011 can control.

(8) When the calculated total number of bits exceeds the number of bits allowed for the current block, the quantization control unit 1011 repeats the processes (5) to (7) so that the spectral quantization unit 1008 and the Huffman coding unit ( 1009). In addition, the quantization control unit 1011 outputs the quantization code 2 from the Huffman encoder 1009 to the bitstream generator 1010 when the calculated total number of bits is less than the number of bits allowed for the current block. In the following description, the quantization controller 1011 controls the bitstream generator 1010 to output a bitstream.

Here, the detail of the quantization process of AAC system is demonstrated.

(a) The AAC method sets the exponent part of the MDCT spectrum to an initial value.

(b) ACC modifies the MDCT spectrum into mantissa and exponent. That is, the AAC scheme transforms the MDCT spectrum into floating point representation. The AAC method quantizes the mantissa (MDCT quantization).

(c) The ACC method obtains the number of bits (total number of bits) required when Huffman coding of the mantissa and exponents quantized in (b).

(d) The ACC method terminates quantization if the total number of bits obtained in (c) is equal to or less than the number of quantization bits (allowed bits) allowed in the current frame. The AAC method determines that the exponent part set in (a) is inadequate when the total number of bits is more than the allowable number of bits. In the AAC method, the exponent part is changed to repeat the processes of (b) to (d). Then, the AAC method determines the exponent part such that the total number of bits is equal to or less than the allowable number of bits.

That is, the AAC method temporarily fixes the exponent part first. The AAC method determines the mantissa and quantizes the MDCT spectrum. The AAC method finds the total number of bits in which the quantization error when the MDCT spectrum is transformed into the exponent part and the mantissa part becomes less than the allowable error. Then, the AAC method determines that the total bit number is larger than the preset bit rate is inappropriate. In the AAC system, the exponent part is changed, and the fixed part of the exponent part of the MDCT spectrum and the quantization process of the mantissa part are performed again. The ACC method determines an optimal exponent part and mantissa part whose quantization error is equal to or less than the tolerance and the total number of bits is equal to or less than the set bit rate.

As described above, the AAC method calculates the required total number of bits after performing quantization and Huffman coding. Then, the AAC method determines the optimal exponent part and mantissa part whose total number of bits is equal to or less than the allowable number of bits allowed in the current frame. Here, "optimal" means "quantization error is below the tolerance."

As described above, the first conventional technique selects an optimal block length from long blocks and short blocks. Therefore, the first prior art can obtain good sound quality with little pre-echo. By the way, the 1st prior art performs MDCT transformation and psychoacoustic analysis for long block and short block, respectively. Therefore, the first prior art has a high throughput.

<2nd prior art>

As a method for solving the problems of the first prior art, a method of first determining the block length by investigating the properties of the input signal before MDCT transformation and psychoacoustic analysis is known. As a method of examining the property of an input signal, for example, there is a method disclosed in Patent Document 1 below. This is a known method.

Hereinafter, the method disclosed in patent document 1 is called 2nd prior art. And the structure of this method is shown in FIG. It is a block diagram which shows the structure of a 2nd prior art. This second prior art divides one frame into shorter short blocks.

(1) First, an input signal is input to the frame unit 1201. The framer 1201 divides the input signal into frames (long blocks) made up of a predetermined number of samples. The signal output from the framer 1201 is output to the power calculator 1202, the selector 1204, and the psychoacoustic hearing analyzer 1208.

The power calculator 1202 calculates a power and a power variation ratio from the input signal. The power calculating section 1202 outputs the calculated power fluctuation ratio to the block length determining section 1203.

The block length determination unit 1203 determines whether a long block or a short block is to be used, based on the input power fluctuation ratio. The block length determination unit 1203 then outputs the determination result to the selector 1204 and the selector 1207. Each selector 1204 and 1207 select whether to use a long block or a short block based on the determination result of the block length determination unit 1203.

The long block MDCT converter 1205 performs 1024 MDCT transforms on the input signal. And the long block MDCT converter 1205 calculates an MDCT coefficient MDCT1.

In addition, the block MDCT conversion unit 1206 performs MDCT conversion of 128 points on the input signal. The block MDCT converter 1206 then calculates an MDCT coefficient MDCT2. In addition, since there are eight blocks per frame, eight sets of MDCT2 are generated.

(2) Next, the psychoacoustic auditory analysis unit 1208 obtains a masking threshold value from the input signal. The masking threshold value obtained from the input signal is input to the spectrum quantization unit 1209.

(3) The spectral quantization unit 1209 quantizes MDCT coefficients for each frequency band according to the input masking threshold value. The spectrum quantization unit 1209 outputs a quantization code 1 obtained by quantizing the MDCT coefficients.

(4) The quantized code 1 output from the spectral quantizer 1209 is input to the Huffman encoder 1210. The Huffman encoder 1210 converts the quantized code 1 into the quantized code 2 with redundancy removed from the quantized code 1.

(5) This quantization code 2 is input to the quantization control unit 1212. The quantization control unit 1212 calculates the total number of bit streams finally output, based on the input quantization code 2. In addition, the range enclosed by the dotted line in FIG. 12 is the range which the quantization control part 1212 can control.

(6) The quantization control unit 1212, when the calculated total number of bits exceeds the number of bits allowed in the current block, causes the spectral quantization unit 1209 and the Huffman encoder to repeat the processes (3) to (5). 1210). In addition, the quantization control unit 1212 outputs a quantization code 2 from the Huffman encoder 1210 to the bitstream generator 1211 when the calculated total number of bits is less than the number of bits allowed for the current block. The quantization controller 1212 controls the bitstream generator 1211 to output a bitstream.

FIG. 13 is a conceptual diagram illustrating an example of dividing a frame into short blocks in the second prior art. FIG. FIG. 13 shows a case where one frame is divided into four short blocks. The second prior art finds the input signal powers P (1), P (2), P (3), and P (4) of each block. In the second prior art power variable costs between the short blocks adjacent _{Δ p (1, 2),} Δ p (2, 3), Δ p (3, 4) is obtained. Here, Δ _p (i, j) is the power variable costs between the short block and the short block i j. Δ _p (i, j) can be obtained by the following equation.

The power fluctuation ratio is large when the input signal is rapidly increased. In contrast, the power fluctuation ratio is small when the input signal is drastically small. Therefore, when the power fluctuation ratio is hardly changed, the block length determination unit 1203 selects the long block. In addition, the block length determination unit 1203 selects a short block when the power fluctuation ratio rapidly increases or decreases. By this process, the second prior art can select the optimum window length.

The second prior art also determines the block length prior to MDCT transformation and psychoacoustic analysis. For this reason, the second prior art performs MDCT transformation and psychoacoustic analysis only on either the long block or the short block. Therefore, the second conventional technique can encode an audio signal with a smaller throughput than the first conventional technique.

By the way, when the property of an input signal changes even if the power fluctuation ratio does not change, the 2nd prior art may be unable to detect the change of the property of an input signal. For example, when the sinusoidal wave is input and the frequency of the sinusoidal wave is changed while the power is constant, the second prior art cannot detect the point of change of the signal by using only the power variation ratio.

Here, examples of the input signal, power fluctuation ratio, and predictive gain fluctuation ratio will be described with reference to FIG. 14 is a graph illustrating examples of input signals, power fluctuation ratios, and predictive gain fluctuation ratios. FIG. 14A is a graph showing an input signal before encoding, FIG. 14B is a graph of power fluctuation ratio, and FIG. 14C is a graph of predicted gain fluctuation ratio. Section B or section C of FIG. 14 is changed from a silent part to a sound part. In this case, the power fluctuation ratio is also greatly changed. Therefore, the second prior art can detect the point of change of the signal in this section.

However, in section A, the nature of the input signal changes from the normal to the transient. In this case, the power fluctuation ratio hardly changes. Therefore, in this case, the second prior art cannot detect a change in the signal. Therefore, in this case, the second prior art selects a long block. However, as in this second conventional technique, when the portion in which the signal is rapidly changed is treated as a long block, pre-echo occurs. Therefore, in the second prior art, sound quality deteriorates.

Patent Document 1: Japanese Patent Laid-Open No. 7-66733

<Non-Patent Document 1> PART 7, "Advanced Audio Coding (ACC)" of ISO / IEC13818-7

[Initiation of invention]

[Problem to Solve Invention]

As described above, the first conventional technique performs MDCT transformation and psychoacoustic analysis for long blocks and short blocks, respectively. Therefore, the first conventional technique has a problem that the throughput increases as compared with the case of processing only a long block or a short block.

In addition, even if the property of the input signal is changed, the second prior art cannot detect the change in the property of the signal unless the power fluctuation ratio is changed. Therefore, the second prior art has a problem that it may not be possible to select an appropriate block length.

An object of the present invention is to provide an audio encoding apparatus and an audio encoding method capable of appropriately selecting a block length while reducing throughput.

[Means for solving the problem]

The audio encoding device of the present invention,

An audio encoding device having a long block mode for dividing an input signal into a frame having a predetermined number of samples, encoding an input signal of one frame, and a short block mode for dividing the frame into short blocks and encoding the short block. as,

Power calculating means for calculating a power fluctuation ratio from the input signal;

Calculating means for calculating a predicted gain variation ratio from the input signal;

Block length determination means for judging whether to perform long block encoding or short block encoding is provided from the power fluctuation ratio and the predicted gain fluctuation ratio.

In addition, the audio encoding apparatus of the present invention,

The block length determining means,

If one of the power fluctuation ratio and the predictive gain fluctuation ratio is greater than a predetermined threshold, encoding by short blocks is selected, and other than when either one of the power fluctuation ratio and the predictive gain fluctuation ratio is greater than a predetermined threshold. Choose encoding by long block.

In addition, the audio encoding apparatus of the present invention,

Threshold determination means for changing a threshold value for determining the block length at the time of encoding used by said block length determination means in response to the determination result of said block length determination means.

In addition, the audio encoding apparatus of the present invention,

The threshold value determining means,

When the determination result of the block length determination means indicates coding by a short block, the threshold value is set to a value larger than the initial value.

In addition, the audio encoding apparatus of the present invention,

The calculating means,

The power calculating means makes one block using a predetermined number of blocks for calculating power, and calculates the predicted gain variation ratio of the one block.

In addition, the audio encoding apparatus of the present invention,

The power calculation means,

The calculation means sets one block by using a predetermined number of blocks for calculating the prediction gain, and calculates the power variation ratio of the one block.

In addition, the audio encoding apparatus of the present invention,

A long block mode for dividing an input signal into frames consisting of a predetermined number of samples and encoding an input signal of one frame;

An audio encoding device comprising a short block mode that divides the frame into short blocks and encodes the short blocks.

Power calculating means for calculating a power fluctuation ratio from the input signal;

Calculating means for calculating a predicted gain variation ratio from the input signal;

Block length determination means for judging whether to perform long block coding or short block coding from the power fluctuation ratio and the predicted gain fluctuation ratio;

In the case where encoding by long blocks is selected by the block length determining means, first transform means for obtaining a first coefficient by discrete cosine transforming an input signal in units of long blocks and short blocks by means of the block length determining means. Second encoding means for dividing cosine transforming the input signal in units of short blocks to obtain a second coefficient when encoding is selected;

Selecting means for selecting the first coefficient or the second coefficient as a third coefficient in response to a determination result of the block length determining means;

Psychoacoustic analysis means for obtaining a masking threshold value from the input signal;

Quantization means for spectral quantizing the third coefficient according to the masking threshold to obtain a first code;

Huffman coding means for obtaining a second code by Huffman coding the first code,

Quantization control means for calculating the total number of bits of the output bitstream from the second code, and instructing the output of the bitstream based on the calculation result;

Bitstream generation means for generating a bitstream from the second code and outputting the bitstream based on an instruction of the quantization control means.

In addition, the audio encoding apparatus of the present invention,

The block length determining means,

If at least one of the power fluctuation ratio and the predicted gain fluctuation ratio is greater than a predetermined threshold, encoding by short blocks is selected, and at least one of the power fluctuation ratio and the predictive gain fluctuation ratio is greater than a predetermined threshold. Otherwise, encoding by long block is selected.

In addition, the audio encoding apparatus of the present invention,

Threshold value determining means for changing the threshold value for determining the block length at the time of encoding used by said block length determination means in response to the determination result of said block length determination means.

In addition, the audio encoding apparatus of the present invention,

The threshold value determining means,

When the determination result of the block length determination means indicates coding by a short block, the threshold value is set to a value larger than the initial value.

In addition, the audio encoding apparatus of the present invention,

The calculating means,

The power calculating means makes one block using a predetermined number of blocks for calculating power, and calculates the predicted gain variation ratio of the one block.

In addition, the audio encoding apparatus of the present invention,

The power calculation means,

The calculation means sets one block by using a predetermined number of blocks for calculating the prediction gain, and calculates the power variation ratio of the one block.

In addition, the audio encoding method of the present invention,

An audio encoding having a long block mode for dividing an input signal into frames having a constant number of samples, encoding an input signal of one frame, and a short block mode for dividing the frame into short blocks and encoding the short blocks. As a way,

A power calculation step of calculating a power fluctuation ratio from the input signal;

A calculating step of calculating a predicted gain variation ratio from the input signal;

And a block length determination step of determining whether to perform long block coding or short block coding from the power fluctuation ratio and the predicted gain fluctuation ratio.

In addition, the audio encoding method of the present invention,

A long block mode for dividing an input signal into frames consisting of a predetermined number of samples and encoding an input signal of one frame;

An audio encoding method having a short block mode for dividing the frame into short blocks and encoding the short blocks,

A power calculation step of calculating a power fluctuation ratio from the input signal;

A calculating step of calculating a predicted gain variation ratio from the input signal;

A block length determination step of determining whether to perform long block coding or short block coding from the power fluctuation ratio and the predicted gain fluctuation ratio;

A first conversion step of obtaining a first coefficient by discrete cosine transforming an input signal in units of long blocks when encoding by long blocks is selected in the block length determination step;

A second conversion step of obtaining a second coefficient by discrete cosine transforming the input signal in units of short blocks when encoding by short blocks is selected in the block length determination step;

A selection step of selecting the first coefficient or the second coefficient as a third coefficient in response to the determination result of the block length determination process;

A psychoacoustic auditory analysis step of obtaining a masking threshold value from the input signal;

A quantization step of spectrally quantizing the third coefficient according to the masking threshold to obtain a first code;

A Huffman encoding step of Huffman encoding the first code to obtain a second code;

A quantization control process of calculating the total number of bits of the output bitstream from the second code and instructing the output of the bitstream based on a result of the calculation;

And a bitstream generation step of generating a bitstream from the second code and outputting the bitstream based on the instructions in the quantization control step.

The audio encoding apparatus and the audio encoding method of the present invention determine whether to perform long block coding or short block coding from the power fluctuation ratio and the prediction gain fluctuation ratio. Therefore, the audio coding apparatus and the audio coding method of the present invention do not need to perform both long block coding and short block coding. Therefore, the audio encoding apparatus and the audio encoding method of the present invention can reduce the throughput and determine the block length to be encoded by using both the power fluctuation ratio and the predictive gain fluctuation ratio. Encoding can be performed.

In addition, the audio encoding apparatus and the audio encoding method of the present invention change the block length determination threshold used for the block length determination in response to the determination result of the block length, so that, for example, encoding by the short block is frequently selected. Can be prevented and the sound quality of the output sound can be reduced.

The audio encoding apparatus and the audio encoding method of the present invention can reduce the throughput by setting one block using a predetermined number of power calculation blocks and calculating the predicted gain variation ratio of the one block.

In addition, the audio encoding apparatus and the audio encoding method of the present invention can reduce the throughput by making one block by using a predetermined number of blocks for calculating the prediction gain and calculating the power fluctuation ratio of the one block.

[Effects of the Invention]

As described above, according to the present invention, an audio encoding apparatus and an audio encoding method capable of appropriately selecting a block length while reducing throughput are provided.

1 is a schematic diagram of an audio encoding apparatus of the present invention.

2 is a conceptual diagram of an example of a long block and a short block used in the audio encoding apparatus of the present invention.

3 is a conceptual diagram of a method for calculating a predicted gain variation ratio in an audio encoding apparatus of the present invention.

4 is a configuration diagram of a first embodiment of an audio encoding device of the present invention.

Fig. 5 is a flowchart of the operation of the block length determination method performed by the first embodiment of the audio encoding device of the present invention.

6 is a configuration diagram of a second embodiment of an audio encoding device of the present invention.

Fig. 7 is a graph showing a threshold value control operation in the threshold value determining unit of the second embodiment of the audio encoding device of the present invention.

Fig. 8 is a conceptual diagram of a method for obtaining a prediction gain variation ratio and a power variation ratio in the third embodiment of the audio encoding apparatus of the present invention.

9 is a conceptual diagram illustrating a method for calculating a power fluctuation ratio in a fourth embodiment of an audio encoding device of the present invention.

Fig. 10 is a configuration diagram showing the configuration of an encoder of MPEG-2 AAC which is the first conventional technology.

11 is a schematic diagram showing an example of pre-echo.

12 is a configuration diagram showing a configuration of a second conventional technology

FIG. 13 is a conceptual diagram illustrating an example of dividing a frame into short blocks in a second prior art. FIG.

14 is a graph showing examples of input signals, power fluctuation ratios, and predictive gain fluctuation ratios.

[Description of the code]

101: framing unit

102: power calculation unit

103: calculating unit

104: block length determination unit

105: selector

106: MDCT transformation for chapter blocks

107: MDCT conversion unit for the short block

108: selector

109: psychological hearing analysis

110: quantization unit

111: Huffman encoder

112: bitstream generator

113: quantization control

401: framing unit

402: power output unit

403: autocorrelation calculation unit

404: k parameter calculator

405: predictive gain calculator

406: predicted gain variation ratio calculating unit

407: block length determination unit

408: selector

409: MDCT transformation for chapter blocks

410: MDCT conversion unit for the short block

411: selector

412: psychological hearing analysis

413: quantization unit

414: Huffman encoder

415: bitstream generator

416: quantization control

601: frame unit

602: power output unit

603: autocorrelation calculation unit

604: k parameter calculation unit

605: prediction gain calculator

606: predicted gain variation ratio calculating unit

607: block length determination unit

608: threshold determination unit

609: selector

610: MDCT transformation unit for long blocks

611: MDCT conversion unit for the short block

612: selector

613: psychological hearing analysis

614: quantization unit

615: Huffman encoder

616: bitstream generation unit

617: Quantization Control

Best Mode for Carrying Out the Invention

SUMMARY OF THE INVENTION

EMBODIMENT OF THE INVENTION Hereinafter, the best form for implementing this invention with reference to drawings is demonstrated. First, the outline | summary of the audio coding apparatus and audio coding method of this invention is demonstrated. 1 is a schematic diagram of an audio encoding apparatus of the present invention. The following description also serves as an overview of the audio encoding method of the present invention. In FIG. 1, the framer 101 divides an input signal into an input signal frame (long block) consisting of a predetermined number of samples. Next, the long block MDCT converter 106, the short block MDCT converter 107, the power calculator 102, and the calculator 103 divide one frame into short blocks shorter than the long block. 2 is a conceptual diagram of an example of a long block and a short block used in the audio encoding apparatus of the present invention. 2 shows a case where one frame (long block) is divided into four short blocks. The following description will be made based on the example shown in FIG. 2. However, the present invention can be similarly implemented even when one frame is divided into n pieces (n> O).

(1) The power calculator 102 calculates input signal powers P (1), P (2), P (3), and P (4) for each block. Next, the power calculation unit 102, a power variable costs between adjacent blocks _{Δ p (1, 2),} Δ p (2, 3), Δ p (3, 4) is obtained. Here, Δ _p (i, j) can be determined by power variable costs between the short block and the short block i j, in the above-described equation (1).

(2) Next, the calculation unit 103 performs LPC analysis (linear prediction analysis) on the input signal of the short block to obtain a k parameter. 3 is a conceptual diagram of a method of calculating a predicted gain variation ratio in the audio encoding apparatus of the present invention. In the present invention, the calculation method of the k parameter is arbitrary. However, the present invention can use a method of obtaining an autocorrelation function from an input signal, for example, and calculating a k parameter from an autocorrelation function by a known method such as a Levinson algorithm.

(3) Next, the calculation unit 103 obtains the prediction gain G (i) from the k parameters k (i, m) and (m = 1, ... p) obtained from the block i using the following equation. Where p is the predicted order.

(4) Next, the calculation unit 103 obtains the predicted gain variation ratio Δ (i, j) from the predicted gains G (i) and G (j) obtained from the blocks i and j using the following equation.

(5) Next, the power variable costs Δ _p (i, j) is input to a block length determining section (104). In addition, the variable cost prediction gain _G Δ (i, j) is input to a block length determining section (104). The block length determination unit 104 then determines which of the long block and the short block is quantized. As the determination method by the block length determination unit 104, the following method can be used. In addition, in the following description, that the block length determination unit selects the long block means that the block length determination unit selects encoding by the long block. Similarly, the block length determining unit selecting the short block means that the block length determining unit selects encoding by the short block. That is, that the block length determination unit selects a block means that the block length determination unit selects encoding by the block.

A) The block length determination unit 104 sets the threshold TH _p and the predicted gain variation ratio TH _G with respect to the power variation ratio.

B) Next, if the block length determining section (104) _{Δ p (1, 2),} Δ p (2, 3), Δ p (3, 4) in the threshold value TH _P than any great that one of the short blocks If no, go to the next C).

C) Next, the block length determining section (104) Δ _G (1, 2), Δ _G (2, 3), Δ _G (3, 4), the threshold TH _G all, if any great that one end block at one If not, select the chapter block.

In other words, the block length determination unit 104 selects a short block only when either one of the power fluctuation ratio and the predicted gain fluctuation ratio in the frame exceeds a preset threshold, and otherwise selects the long block.

(6) When the block length determination unit 104 selects the long block, the determination result is output to the selector 105 and the selector 108. The selector 105 and the selector 108 select a block based on the determination result of the block length determination unit 104. Therefore, when the block length determination unit 104 selects the long block, the selector 105 and the selector 108 select the long block.

Then, the input signal output from the framer 101 is input to the long block MDCT converter 106. And the long block MDCT converter 106 outputs MDCT1.

In addition, when the block length determination unit 104 selects a short block, the determination result is output to the selector 105 and the selector 108. The selector 105 and the selector 108 select a short block.

Then, the input signal output from the framer 101 is input to the short block MDCT converter 107. Then, the short block MDCT converter 107 outputs MDCT coefficients by the number of short blocks. That is, when one frame is divided into four short blocks, the short block MDCT converter 107 outputs four sets of MDCT coefficients.

(7) Next, the psychoacoustic auditory analyzer 109 calculates a masking threshold value from the input signal. Here, when the block length determination unit 104 selects a long block, the psychoacoustic analysis unit 109 obtains a masking threshold value for the long block. In addition, when the block length determination unit 104 selects a short block, the psychoacoustic analysis unit 109 calculates a masking threshold value for the short block.

In the present invention, the calculation method of the masking threshold value can use any method. For example, the psychoacoustic hearing analyzing unit 109 can use the method disclosed in Non Patent Literature 1. In other words, the psychoacoustic analysis unit 109 performs FFT analysis on the input signal. The psychoacoustic analysis unit 109 then obtains an FFT spectrum. Then, the psychoacoustic hearing analyzing unit 109 calculates a masking threshold value from the FFT spectrum.

(8) Next, the MDCT coefficients and the masking threshold value are input to the quantization unit 110. The quantization unit 110 quantizes MDCT coefficients for each frequency band according to the input masking threshold. The quantization unit 110 outputs a quantization code 1 in which the MDCT coefficients are quantized.

(9) Next, the quantization code 1 is input to the Huffman coding unit 111. The Huffman encoder 111 converts the quantized code 1 into the quantized code 2 with redundancy removed from the quantized code 1.

(10) Next, the Huffman coding unit 111 outputs the quantization code 2 to the quantization control unit 113. The quantization control unit 113 calculates the total number of bits of the bitstream finally output from the input quantization code 2. In addition, the range enclosed by the dotted line in FIG. 1 is the range which the quantization control part 113 can control.

(11) If the total number of bits calculated exceeds the number of bits allowed for the current block, the quantization control unit 113 repeats the processes (8) to (10) so that the quantization unit 110 and the Huffman coding unit 111 are repeated. ). In addition, the quantization control unit 113 outputs the quantization code 2 from the Huffman encoder 111 to the bitstream generator 112 when the calculated total number of bits is less than the number of bits allowed for the current block. The quantization control unit 113 controls the bitstream generation unit 112 to output a bitstream. As a result, the audio encoding apparatus shown in FIG. 1 realizes quantization. In addition, since the process of quantization in this invention is the same as the detail of the AAC system quantization process demonstrated in the column of the prior art mentioned above, the detailed description is abbreviate | omitted.

Next, an embodiment of the present invention will be described with reference to the drawings. The structure of the following embodiment is an illustration, and this invention is not limited to the structure of embodiment. In addition, description of each following embodiment is demonstrated using the audio coding apparatus which encodes an audio signal as an example. In addition, description of each embodiment of the audio encoding device of the present invention described below also serves as a description of each embodiment of the audio encoding method of the present invention.

<First Embodiment>

4 is a configuration diagram of a first embodiment of an audio encoding device of the present invention. In FIG. 4, the framer 401 divides the input signal into an input signal frame (long block) consisting of a predetermined number of samples.

Next, the short block MDCT converter 410, the power calculator 402, and the autocorrelation calculator 403 divide the input one frame into short blocks. The division of the frame in the present embodiment will be described with reference to FIG. 2 described above. 2 is a conceptual diagram illustrating an example of a long block and a short block. The example shown in FIG. 2 divides one frame (long block) into four short blocks. The following description will be made based on such an example. However, this embodiment is similarly performed even when one frame is divided into n (n is a non-negative integer).

(1) First, the power calculating section 402 obtains input signal powers P (1), P (2), P (3), and P (4) for each block. Then, the power calculation unit 402 the power variable costs between adjacent blocks _{Δ p (1, 2),} Δ p (2, 3), Δ p (3, 4) is obtained. Here, Δ _p (i, j) is the power fluctuation ratio between the short block i and the short block j. This power fluctuation ratio can be obtained from the above equation (1).

(2) Next, the autocorrelation calculation unit 403 obtains autocorrelation from the input signal of the short block. The autocorrelation calculator 403 then outputs this autocorrelation to the k parameter calculator 404.

Next, the k parameter calculator 404 calculates the k parameter from a autocorrelation function by a known method such as a Levinson algorithm. In addition, the k parameter calculator 404 may obtain the LPC coefficient from the autocorrelation function, and the k parameter calculator 404 may convert the LPC coefficient into the k parameter.

(3) Then, the prediction gain calculator 405 calculates the prediction gain G (i) from the k parameters k (i, m) and (m = 1, ..., p) obtained from the block i by the following equation. Where p is the predicted order. The predicted gain G (i) is input to the predicted gain variation ratio calculation unit 406.

(4) Next, the prediction gain variable cost calculation unit 406 is the short blocks i, from a single block prediction gain calculated by j G (i), G (j), the prediction gain variable costs shown by the following equation Δ _G (i, j ) Here, the autocorrelation calculator 403, the k parameter calculator 404, the predictive gain calculator 405, and the predicted gain variation ratio calculator 406 are part of the functions of the calculator 103 shown in FIG. 1. You may also

(5) Next, the power variable costs Δ _P (i, j) and the variable cost prediction gain _G Δ (i, j) is input to a block length determining section (407). The block length determination unit 407 then determines which of the long block and the short block is quantized. As the determination method used by the block length determination unit 407, the following method can be used. Hereinafter, the determination method performed by the block length determination unit will be described with reference to FIG. 5. 5 is a flowchart of the operation of the block length determination method performed by the first embodiment of the audio encoding apparatus of the present invention. In addition, in the following description, as described above, that the block length determination unit selects the long block means that the block length determination unit selects encoding by the long block. Similarly, the block length determining unit selecting the short block means that the block length determining unit selects encoding by the short block. That is, that the block length determination unit selects a block means that the block length determination unit selects encoding by the block.

(A) The block length determination unit 407 sets the threshold TH _p for the power fluctuation ratio and the threshold TH _G for the predicted gain fluctuation ratio.

(B) a block length determining section 407 selects the short blocks if _{Δ p (1, 2),} Δ p (2, 3), Δ p (3, 4) from the threshold TH _p than any large that one (S501, S502, S503, S508), if no, go to (C).

(C) a block length determining section 407 if _{Δ G (1, 2),} Δ G (2, 3), Δ G (3,4) in the threshold value TH _G larger than any one to select the short blocks, and (S504, S505, S506, S508) If not, the long block is selected (S507).

In other words, the block length determination unit 407 selects a short block only when either one of the power fluctuation ratio and the predicted gain fluctuation ratio in the frame exceeds a preset threshold value, and selects a long block elsewhere.

(6) The determination result of the block length determination unit 407 is input to the selector 408 and the selector 411. Each selector 408 and selector 411 select the block length to be used based on the determination result of the block length determination unit 407.

When the block length determination unit 407 selects the long block, the input signal is input to the long block MDCT converter 409. The long block MDCT converter 409 outputs MDCT coefficients.

In addition, when the block length determination unit 407 selects a short block, an input signal is input to the short block MDCT converter 410. Then, the short block MDCT converter 410 outputs a few MDCT coefficients of the short block. That is, when one frame is divided into four short blocks, the short block MDCT converter 410 outputs four sets of MDCT coefficients.

(7) Next, the psychoacoustic auditory analyzer 412 obtains a masking threshold value from the input signal. The psychoacoustic auditory analyzer 412 receives an input signal output from the framer 401. Here, when the block length determination unit 407 selects a long block, the psychoacoustic hearing analyzer 412 obtains a masking threshold value for the long block. In addition, when the block length determining unit 407 selects the short block, the psychoacoustic hearing analyzing unit 412 calculates a masking threshold value for the short block.

In this embodiment, the calculation method of a masking threshold value can use arbitrary methods. For example, the psychoacoustic hearing analyzing unit 412 can use the method disclosed in Non Patent Literature 1. That is, the psychoacoustic hearing analyzing unit 412 performs FFT analysis on the input signal. The psychoacoustic analysis unit 412 then obtains an FFT spectrum. Then, the psychoacoustic hearing analyzer 412 calculates a masking threshold value from the FFT spectrum.

(8) The MDCT coefficients and masking thresholds are input to the quantization unit 413. The quantization unit 413 quantizes the MDCT coefficients for each frequency band according to the input masking threshold. The quantization unit 413 outputs a quantization code 1 obtained by quantizing the MDCT coefficients.

(9) Next, the quantization code 1 is input to the Huffman encoder 414. The Huffman encoder 414 converts the quantized code 1 into the quantized code 2 with redundancy removed from the quantized code 1.

(10) Next, the Huffman coding unit 414 outputs the quantization code 2 to the quantization control unit 416. The quantization control unit 416 calculates the total number of bits of the bitstream finally output from the input quantization code 2. In addition, the range enclosed by the dotted line in FIG. 4 is the range which the quantization control part 416 can control.

(11) If the total number of bits calculated exceeds the number of bits allowed for the current block, the quantization control unit 416 repeats the processes (8) to (10) so that the quantization unit 413 and the Huffman encoder 414 are repeated. ). In addition, the quantization control unit 416 outputs the quantization code 2 from the Huffman encoder 414 to the bitstream generator 415 when the calculated total number of bits is less than the number of bits allowed for the current block. The quantization controller 415 controls the bitstream generator 415 to output a bitstream. This embodiment realizes quantization. In addition, since the quantization process in this embodiment is the same as the detail of the AAC system quantization process demonstrated in the above-mentioned prior art column, the detailed description is abbreviate | omitted.

In this embodiment, the case where one frame is divided into four short blocks has been described in the example. The present invention can be similarly realized even when one frame is divided into an arbitrary number (for example, eight blocks).

As described above, in the present embodiment, since the block length is determined before the MDCT conversion, the audio signal of high quality can be encoded with a small throughput compared to the first conventional technology. Further, in the present embodiment, since the block length is determined using the power fluctuation ratio and the predicted gain fluctuation ratio, the block length is more precisely determined than in the second prior art. Coding is possible.

In other words, the present embodiment determines the block length for encoding before MDCT transformation and psychoacoustic analysis. Therefore, this embodiment can encode high quality with a small throughput compared with the first prior art. In this embodiment, the block length determining means uses the power fluctuation ratio and the predicted gain fluctuation ratio. For this reason, the present embodiment can accurately determine the block length as compared with the second prior art.

The effect of this embodiment is demonstrated in more detail using FIG. 14 mentioned above. 14 is a graph showing the results of calculating the power fluctuation ratio and the predicted gain fluctuation ratio. In the input signal shown in FIG. 14A, the value of the power fluctuation ratio hardly changes to 0 in the section A (FIG. 14B). In contrast, in the input signal shown in FIG. 14A, the predicted gain variation ratio is greatly changed in the section A (FIG. 14C).

This embodiment calculates both a power fluctuation ratio and a prediction gain fluctuation ratio. Then, in this embodiment, if either one of the power fluctuation ratio and the predicted gain fluctuation ratio exceeds the threshold, the short block is selected. Therefore, the present embodiment can accurately determine the block length even in an input signal such as the section A shown in FIG.

In addition, in the sections B and C shown in Fig. 14, the predicted gain fluctuation ratio is hardly changed. On the other hand, in the sections B and C shown in Fig. 14, the power fluctuation ratio varies greatly. Therefore, the present embodiment can detect the change point of the signal in the sections B and C as in the second conventional technique.

<2nd embodiment>

6 is a configuration diagram of a second embodiment of an audio encoding device of the present invention. Compared with the first embodiment, the present embodiment differs from the threshold TH _P for the power fluctuation ratio and the portion for dynamically changing the threshold TH _G for the predicted gain fluctuation ratio. Since other parts are common to the first embodiment, description thereof is omitted.

In general, the short block is often selected from a rapidly changing portion such as an attack sound. Attack sounds have large amplitudes in the MDCT spectrum over a wide frequency range. Therefore, when the attack sound is encoded, a large number of quantized bits is required.

If a short block is selected in succession, the sound quality may be extremely degraded due to insufficient number of quantized bits. Therefore, in order to encode an audio signal at a low bit rate, it may be necessary to control such that short blocks are not selected continuously.

Therefore, in the present embodiment, once a short block is selected, the threshold value TH _P and the threshold value TH _G are increased for a certain time thereafter. As a result, in this embodiment, the possible short blocks are not selected continuously.

Here, the configuration of the second embodiment of the audio encoding device of the present invention will be described. The configuration of this embodiment is shown in FIG. Since the operation of blocks other than the block length determining unit 607 and the threshold value determining unit 608 in each block shown in FIG. 6 is the same as that of the corresponding respective block shown in FIG. Omit.

That is, the operation of the framer 601 shown in FIG. 6 is the same as the operation of the framer 401 shown in FIG. 4, and the operation of the power calculator 602 is the power calculator 402 shown in FIG. 4. The operation of the autocorrelation calculation unit 603 is the same as the operation of the autocorrelation calculation unit 403 shown in FIG. 4, and the operation of the k parameter calculation unit 604 is shown in FIG. 4. Similar to the operation of the k parameter calculator 404, the operation of the predictive gain calculator 605 is the same as the operation of the predictive gain calculator 405 shown in FIG.

In addition, the operation of the predicted gain variation ratio calculating unit 606 is the same as the operation of the predicted gain variation ratio calculating unit 406 shown in FIG. 4, and the operation of the selector 609 is the operation of the selector 408 shown in FIG. 4. Similarly, the operation of the long block MDCT converter 610 is the same as the operation of the long block MDCT converter 409 shown in FIG.

In addition, the operation of the short block MDCT converter 611 is the same as the operation of the short block MDCT converter 410 shown in FIG. 4, and the operation of the selector 612 is the selector 411 shown in FIG. 4. The operation of the psychoacoustic analysis unit 613 is similar to that of the psychoacoustic analysis unit 412 shown in FIG. 4, and the operation of the quantization unit 614 is similar to the operation of the quantization unit ( Similar to the operation of 413, the operation of the Huffman encoder 615 is the same as the operation of the Huffman encoder 414 shown in FIG. 4, and the operation of the bitstream generator 616 is the bit shown in FIG. 4. Similar to the operation of the stream generator 415, the operation of the quantization control unit 617 is the same as the operation of the quantization control unit 416 shown in FIG. 4. In addition, the range enclosed by the dotted line in FIG. 6 is the range which the quantization control part 617 can control.

On the other hand, the block length determination unit 607 shown in FIG. 6 receives the threshold value determined by the threshold value determination unit 608. The block length determination unit 607 further outputs the block length determination result to the selector 609, the selector 612, and the threshold value determination unit 608. The threshold value determination unit 608 determines the threshold value based on the determination result output from the block length determination unit 607. That is, the threshold value determination unit 608 outputs the increased threshold value when the determination result output from the block length determination unit 607 is a determination result of selecting a short block. In addition, the block length determination unit 607 performs the determination processing based on the threshold value received from the threshold determination unit 608. The determination processing in the block length determination unit 607 is the same as the case shown in FIG. 5 described above except that the threshold value can be changed, and thus detailed description thereof is omitted. The threshold value determination unit 608 may be part of the function of the calculation unit 103 shown in FIG. 1.

Fig. 7 is a graph showing a threshold value control operation in the threshold value determination unit of the second embodiment of the audio encoding device of the present invention. In the graph shown in FIG. 7, when a block is selected, the threshold value TH _G is changed to TH _G + α. Here, α> 0. Similarly, if a short block is selected, the threshold value TH _P is changed to TH _P + β. Here, β> 0.

After that, when the predetermined time Δt elapses, the threshold value is changed to the original value (initial value) TH _G , TH _p . That is, in the present embodiment, once a short block is selected, the threshold value TH _P and the threshold value TH _G are increased for a predetermined time thereafter so that the possible short blocks are not selected continuously.

As described above, the present embodiment can obtain the same effects as those of the first embodiment described above. In addition, in this embodiment, when a short block is selected, the threshold value is controlled so that the short block is not selected for a certain time thereafter. Therefore, in this embodiment, the sound quality deterioration which arises by selecting a continuous block continuously can be reduced.

In addition, as a modification of this embodiment, the following method can also be implemented. Also in the following modifications, the same effects as in the second embodiment of the audio encoding device of the present invention can be obtained.

(1) In the modification of this embodiment, after the short block is selected, the short block is not selected for a predetermined time.

(2) In the modification of the present embodiment, after the block is selected, α or β is sufficiently large. However, in the modification of the present embodiment, it is necessary to examine the TH _G or TH _P range in advance.

(3) In the modification of the present embodiment, when the short block is selected and the short block is selected again when the threshold value is TH _G + α or TH _P + β, the threshold value is set to TH _G + α + α or TH _P + β + β. However, the modified example of this embodiment returns a threshold value to an original value after a fixed time.

Third Embodiment

Next, a third embodiment of the audio encoding device of the present invention will be described. The structure of this embodiment is the same as that of 1st embodiment shown in FIG. However, the third embodiment differs from the above-described first embodiment in that the prediction gain variation ratio is obtained in units of frames. In other words, in this embodiment, one block is calculated using a predetermined number of power calculating blocks, and the predicted gain variation ratio of the one block is calculated.

In the first embodiment, LPC analysis is performed for each block. Therefore, the first embodiment can calculate the predicted gain variation ratio precisely. However, in the first embodiment, since the number of times of performing LPC analysis increases, the throughput also increases. In this embodiment, one LPC analysis is performed for one block. Therefore, this embodiment can reduce the amount of computation as compared with the first embodiment.

8 is a conceptual diagram of a method for obtaining a predicted gain variation ratio and a power variation ratio in the third embodiment of the audio encoding apparatus of the present invention. In the first embodiment, the prediction gain is obtained from the k parameter obtained by performing LPC analysis for each block. In the first embodiment, the predicted gain variation ratio is calculated based on the ratio of the predicted gains obtained in the same manner in one previous block.

In contrast, in this embodiment, as shown in Fig. 8A, LPC analysis is performed on an input signal of one long block (n-th frame) to obtain a k parameter. In other words, the k parameter calculation unit performs LPC analysis on the input signal of one long block (n-th frame) to obtain the k parameter. In this embodiment, the prediction gain G (n) is calculated from the k parameter. Next, the embodiment using the prediction gain power G (n-1) and G (n) obtained in the same manner as in the one previous frame (the n-1 frame), prediction gain variable costs by the following formula: Δ _G ( calculate n).

On the other hand, as the present embodiment is shown in (b) of Figure 8, as in the first embodiment, each short block power variable costs _{Δ p (1, 2),} Δ p (2, 3), Δ p (3 , 4). Next, the present embodiment determines the optimum block length from the calculated predicted gain variation ratio and power variation ratio. This determination operation will be described below.

(1) Δ _G (n) block length determining section is larger than the threshold value TH _G predetermined selects the short blocks.

(2) Next, the block length determining section Δ _p (1, 2), Δ _p (2, 3), Δ _p (3, 4) from, only one block, even if there is pre-large determined than the threshold value TH _P Choose.

(3) And if the short block is not selected in either (1) or (2), a block length determination part selects a long block. In this embodiment, the configuration and processing contents after the block length is selected are common to the first embodiment. Therefore, description is abbreviate | omitted about the structure and content after selecting the block length of this embodiment.

As explained above, this embodiment can acquire the effect similar to 1st embodiment of this invention mentioned above. In addition, in this embodiment, the LPC analysis is performed only once per long block, so that the block length can be selected with a smaller throughput than in the first embodiment. However, in the present embodiment, the block for calculating the prediction gain is not limited to the case of using a block of one frame, and the block for calculating the power is used as an arbitrary number of blocks. The prediction gain may be calculated. Even in such a case, the present embodiment can obtain the same effects as described above.

<4th embodiment>

Next, a fourth embodiment of the audio encoding device of the present invention will be described. The configuration of this embodiment is the same as that of the first embodiment. However, the present embodiment differs from the first embodiment in the method of calculating the power fluctuation ratio which is performed by dividing one frame into eight blocks. That is, in this embodiment, one block is calculated using a predetermined number of blocks for calculating the prediction gain, and the power fluctuation ratio of one such block is calculated.

9 is a conceptual diagram illustrating a method for calculating a power fluctuation ratio in a fourth embodiment of an audio encoding device of the present invention. As shown in Fig. 9, the present embodiment divides one frame into eight short blocks and calculates a power fluctuation ratio. However, this embodiment does not calculate the power fluctuation ratio per one block as in the first embodiment.

In other words, the present embodiment differs from the first embodiment in that the power variation ratio is obtained from a plurality of adjacent single blocks. The calculation method of the power variation ratio of the present embodiment is described below.

In this embodiment, the power P (1) is obtained from the first and second end blocks. In this embodiment, the power P (2) is obtained from the third and fourth end blocks. In this embodiment, the power P (3) is obtained from the fifth and sixth end blocks. In this embodiment, the power P (4) is obtained from the seventh and eighth end blocks.

Next, the present embodiment calculates the power variable costs Δ _P (1, 2) from P (1) and P (2). In addition, this embodiment calculates a power variable costs Δ _P (2, 3) from P (2) and P (3). In addition, the present embodiment calculates the power variable costs Δ _P (3, 4) from the P (3) and P (4).

As described above, the present embodiment differs from the first embodiment in that power of two short blocks is obtained. In other words, in the first embodiment, eight prediction gain variation ratios and eight power variation ratios are calculated. In this embodiment, only eight prediction gain variation ratios and four power variation ratios are calculated. That is, in this embodiment, the number of prediction gain fluctuation ratios and power fluctuation ratios calculated within one frame may be different. Since it is the same as that of 1st Embodiment except the above-mentioned part of this embodiment, description is abbreviate | omitted.

Thus, this embodiment can acquire the effect similar to the 1st Embodiment of this invention mentioned above. In addition, in this embodiment, by calculating the power of two short blocks, the calculation amount of the power calculation process can be reduced as compared with the first embodiment. In addition, this embodiment is a block which calculates electric power, It is not limited to using two single blocks, You can calculate electric power using three or more arbitrary numbers of short blocks. Also in this case, an effect similar to the above effect can be obtained.

Claims (14)

Power calculating means for calculating a power fluctuation ratio based on an input signal; Calculating means for calculating a predicted gain variation ratio based on the input signal; Based on the power fluctuation ratio and the predicted gain fluctuation ratio, encoding using a long block mode for dividing an input signal into frames each consisting of a predetermined number of samples and encoding each of the frames; Block length determining means for dividing into short blocks and selecting one of encoding using a short block mode for encoding each of the short blocks; And Threshold determination means for changing a threshold value for determining the block length used by the block length determination means at the time of encoding according to the selection result of the block length determination means. Including; The block length determining means selects coding using the short block mode when any one of the power fluctuation ratio and the predictive gain fluctuation ratio is larger than a predetermined threshold value, and one of the power fluctuation ratio and the predictive gain fluctuation ratio is The audio encoding apparatus which selects the coding using the long block mode except when it is larger than a predetermined threshold. The method of claim 1, The threshold value determining means, And the threshold value is set to a value larger than an initial value when the selection result of the block length determination means indicates the selection of the encoding using the short block mode. The method of claim 1, The calculating means, And an audio encoding apparatus for calculating the predicted gain variation ratio of a single block, each of which is a combination of a predetermined number of blocks used by the power calculating means for calculating power. The method of claim 1, The power calculation means, An audio encoding apparatus for calculating the power variation ratio of a single block, each of which is a combination of a predetermined number of blocks used by the calculating means for calculating a prediction gain. Power calculating means for calculating a power fluctuation ratio based on an input signal; Calculating means for calculating a predicted gain variation ratio based on the input signal; Based on the power fluctuation ratio and the prediction gain fluctuation ratio, encoding using a long block mode for dividing an input signal into frames each consisting of a predetermined number of samples and encoding each of the frames, and blocks each of the frames Block length determining means for dividing into and selecting one of encodings using a short block mode for encoding each of the short blocks; Threshold value determining means for changing a threshold value for determining the block length used by said block length determination means at the time of encoding according to a selection result of said block length determination means; A first transforming means for obtaining a first coefficient by performing a modified discrete cosine transform on the input signal in units of long blocks when the block length determining means selects encoding using a long block mode; Second conversion means for obtaining a second coefficient by transforming discrete input cosine transforming the input signal in units of short blocks when the block length determining means selects encoding using a short block mode; Selecting means for selecting one of the first coefficient and the second coefficient as a third coefficient according to a selection result of the block length determining means; Psychoacoustic analysis means for obtaining a masking threshold value from the input signal; Quantization means for obtaining a first code by spectral quantizing the third coefficient according to the masking threshold; Huffman coding means for obtaining a second code by Huffman coding the first code; Quantization control means for calculating the total number of bits constituting the output bitstream from the second code, and instructing the output of the bitstream based on the calculation result of the total number of bits; And Bitstream generating means for generating a bitstream from the second code and outputting the bitstream based on an instruction of the quantization control means Including; The block length determining means selects an encoding based on use of the short block mode when any one of the power fluctuation ratio and the prediction gain fluctuation ratio is larger than a predetermined threshold, and selects any one of the power fluctuation ratio and the prediction gain fluctuation ratio. Audio encoding apparatus for selecting encoding based on the use of the long block mode when is not larger than the predetermined threshold. Calculating a power fluctuation ratio based on the input signal; Calculating a predicted gain variation ratio based on the input signal; Based on the power fluctuation ratio and the prediction gain fluctuation ratio, encoding using a long block mode for dividing an input signal into frames each consisting of a predetermined number of samples and encoding each of the frames, and blocks each of the frames. A block length determining step of dividing into and selecting one of encodings using a short block mode for encoding each of the short blocks; And A threshold value determining step of changing a threshold value for determining the block length used in the block length determining step in accordance with a selection result of the block length determining step when encoding; Including; In the block length determination step, when any one of the power fluctuation ratio and the predictive gain fluctuation ratio is greater than a predetermined threshold value, encoding is selected using the short block mode, and one of the power fluctuation ratio and the predictive gain fluctuation ratio is An audio encoding method for selecting encoding using a long block mode when it is not larger than a predetermined threshold. Calculating a power fluctuation ratio based on the input signal; Calculating a predicted gain variation ratio based on the input signal; Based on the power fluctuation ratio and the prediction gain fluctuation ratio, encoding using a long block mode for dividing an input signal into frames each consisting of a predetermined number of samples and encoding each of the frames, and blocks each of the frames A block length determining step of dividing into and selecting one of encodings using a short block mode for encoding each of the short blocks; A threshold value determining step of changing a threshold value for determining the block length used in the block length determining step when encoding, according to a selection result of the block length determining step; A first transforming step of obtaining a first coefficient by performing transformed discrete cosine transforming the input signal on a long block basis when encoding using a long block mode is selected in the block length determining step; A second transforming step of obtaining a second coefficient by transforming discrete cosine transforming the input signal in units of short blocks when encoding using a short block mode is selected in the block length determining step; A selecting step of selecting one of the first coefficient and the second coefficient as a third coefficient according to a selection result of the block length determining step; Psychoacoustic analysis step of obtaining a masking threshold value from the input signal; A quantization step of obtaining a first code by spectral quantizing the third coefficient according to the masking threshold; A Huffman encoding step of obtaining a second code by Huffman encoding the first code; A quantization control step of calculating a total number of bits constituting the output bitstream from the second code and indicating an output of the bitstream based on a result of the calculation of the total number of bits; And A bitstream generation step of generating a bitstream from the second code and outputting the bitstream based on an instruction in the quantization control step Including; In the block length determination step, when any one of the power fluctuation ratio and the prediction gain fluctuation ratio is greater than a predetermined threshold value, the encoding is selected based on use of the short block mode, and any one of the power fluctuation ratio and the prediction gain fluctuation ratio is selected. Audio encoding method for selecting encoding based on the use of the long block mode when is not larger than the case where the predetermined threshold value is larger than the predetermined threshold value. delete delete delete delete delete delete delete

Images