JP3518737B2 - Audio encoding device, audio encoding method, and audio encoded signal recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an audio encoding method that uses window functions to conduct overlap encoding. SOLUTION: In the audio encoding method that uses at least 4 kinds of window functions of a long window, a short window, a start window, an a stop window to conduct overlapping encoding, the long window or the start window of the window function at the start of encoding is used by taking the matching performance of codes into account, the long window or the stop window is used for the window function at the end of encoding to decide the window function, the window function is multiplied, its output is converted from a time axis into a frequency axis, data are encoded so that a virtual buffer value of a frame at the start of encoding and a virtual buffer value of a frame at the end of encoding are the same specified value so as to enhance the matching performance of the codes.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio (voice) encoding method and apparatus for performing overlap encoding using a plurality of window functions.

[0002]

2. Description of the Related Art A conventional VC has been used as an audio encoding system.
MP, which is an encoding method used in D (video CD), etc.
Linear PCM1152 for EG1-LayerII system
Encoding is performed in sample units (frames) 1152.
The code amount generated for each sample is constant. In such a case, even if the coded streams that have been signal-coded separately are connected, if the coding rate is constant as shown in “Stream connection in MPEG1 Layer II” in FIG. Since the code amount (Nbits) generated in units is constant (fixed), overflow or underflow did not occur in the virtual buffer indicating the buffer amount at the time of decoding.

However, M standardized in April 1997
In the case of the PEG2-AAC (Advanced Audio Coding) coding method, coding is performed in PCM1024 units (frames), and even if the coding rate is constant, "MPEG" in FIG.
As shown in "2-AAC stream connection", the number of bits per frame is variable, and the code amount generated for each frame is different.

An example of the audio encoding / decoding apparatus of the MPEG2-AAC encoding system will be described with reference to the drawings. FIG. 14 is a block diagram showing an example of an audio encoding / decoding device. This audio encoding / decoding device includes a window function processor 11, a time frequency converter 12, a quantizer 13,
Auditory model 14, inverse quantizer 15, frequency-time converter 1
6 and a window function processor 17.

First, the window function processor 11 multiplies the input PCM signal by a window function.
This is necessary for obtaining the frequency component with high accuracy in the next time frequency converter 12. This window function is shown in Figure 1.
As shown in FIG. 6, an overlapping area is provided to create a frame, and the process of multiplying the window function is performed for each frame.

The time-frequency converter 12 uses the FFT and MD
Conversion from the time axis to the frequency axis is performed using CT (Modified Discrete Cosine Transform) or the like. The auditory model 14 is supplied with the input PCM signal, calculates the bit allocation amount for quantization by calculating the masking level based on the psychology of hearing, and supplies the bit allocation amount to the quantizer 13. The quantizer 13 quantizes based on the bit allocation amount, forms a bit stream, and outputs it.

On the decoding side, the inverse quantizer 15 decomposes the transmitted bit stream and performs inverse quantization. The frequency-time converter 16 includes an inverse quantizer 15
Is supplied to perform conversion from the frequency axis to the time axis.

The output of the frequency-time converter 16 is supplied to the window function processor 17, and the window function used at the time of encoding is further multiplied by the frame, and the influence of the window function is removed by adding the preceding and succeeding frames. And the PCM signal is decoded.

In the case of the MPEG2-AAC system,
As shown in “Window function and overlap processing in MPEG2 AAC” in FIG. 4, encoding is performed by overlapping in units of 1024 samples. In this case, 204
After multiplying the window function every 8 samples, MDCT (ModifiedD
Iscrete Cosine Transform) is used to perform conversion on the frequency axis for encoding.

For the window function, in order to perform adaptive block length switching in accordance with the block length for performing the MDCT operation, a long window, as shown in FIG. Short window, start
There are four types of functions, windows and stop windows.

FIG. 16 is a diagram showing a sine window that is often used when MDCT is used in the time-frequency converter 12 of FIG. 14 and a frame structure with 50% overlap. The audio signal (PCM data) is divided at a certain fixed period, multiplied by a sine window, MDCT is executed, and signal processing is performed (frame 1). The next frame (frame 2) is composed of an audio signal (overlap area) in the latter half of the frame 1 and a new audio signal, and is processed in the same manner.

However, in the case of an audio signal containing an attack sound such as castanets, since quantization noise is easily detected in frames with a constant interval, frames having a shorter interval than a normal frame (short frame) are detected.
By configuring, it is difficult to perceive the quantization noise.
In that case, processing is performed using a window function corresponding to a short frame (short window function) and a window function (start window function) for shifting to a short frame.

FIG. 6 shows four types of window shapes: long, short, start, and stop. start,
The stop window is used for connecting the long window and the short window and has an asymmetrical shape. FIG. 17 shows a state transition diagram of the window function. Normally, the long window is used repeatedly, but when a transient signal such as an attack sound is detected, the next frame transits to the start window, and the next frame transits to the short window. .

As it returns to a steady signal, it goes from the short window to the stop window and back to the long window. Further, there is a possibility that the start window transits to the stop window without passing through the short window. However, when the coded stream is edited, the frames are not always connected in a form following the state transition of FIG. 17, for example, the long window is followed by the long window or the start window.

[0015]

Here, at the start of encoding, as shown in the "encoding start frame" of FIG.
It is common to use silence data for the overlap portion. Therefore, when the bit stream is simply connected at the time of editing by using the silent data in the overlapping portion, the reproduced sound of the connecting portion is as shown in “Example of decoding result of connecting portion” in FIG. There is a problem that normal playback sound cannot be obtained. In particular, when the short window or the start window is used at the end of encoding, the start window is used even when the short window is used (FIG. 15A) as shown in each addition result column of FIG. If there is (Fig. 15
Also in (b), since the level of the reproduced sound sharply drops in a short time and reaches the silent data level, there is a problem that the output sound becomes uncomfortable.

When the material A and the material B are coded by the MPEG2-AAC coding method, when the material A is coded, A1, A2, A3, A4 bits are set for each frame.
Is generated, and in the case of the material B, B1,
B2, B3, B4 bits are generated. As for these, the amount of code that can be used for each material A and B is controlled, and encoding is performed so that the virtual buffer becomes stable.
As shown in "Stream Joining with PEG2AAC", simply connecting the material A in the first half and the material B in the second half results in the connected bit streams A1, A2, B3,
Since the code is B4 bits, there is a problem that the virtual buffer causes overflow or underflow.

[0017]

In order to solve this problem, the present invention uses at least four kinds of window functions of a long window, a short window, a start window and a stop window. In an audio coding method for performing overlap coding, the window function at the start of coding is either the long window or the start window in consideration of the splicing of codes, and the window function at the end of coding is the long window. The window function is determined as either a window or a stop window, the window function is multiplied, the output is converted from the time axis to the frequency axis, and the virtual buffer value of the frame at the start of encoding and the encoding are performed. An audio encoding method is provided, wherein encoding is performed such that the virtual buffer values of the frames at the end are the same specified value to improve the connectivity of the codes. In an audio encoding method for performing overlap encoding using at least four types of window functions of a long window, a short window, a start window, and a stop window, the window at the start of encoding is considered in consideration of code splicing. The function determines the window function as either the long window or the start window, multiplies the window function, then converts the output from the time axis to the frequency axis, and the virtual buffer of the frame at the start of encoding. The code string encoded so that the value becomes a specified value is the first code string, and the window function at the end of encoding determines the window function as either the long window or the stop window, and the window function After the function is multiplied, the output is converted from the time axis to the frequency axis, and the code string encoded so that the virtual buffer value of the frame at the end of the encoding becomes the specified value is the second code string. age 4. An audio encoding method, comprising: connecting a start portion of the first code string and an end portion of the second code string so as to improve splicing. The invention of
In an audio encoding device that performs overlap encoding using at least four types of window functions of a long window, a short window, a start window, and a stop window, the window function at the start of encoding considers the splicing of codes. The long window or the start window, and the window function at the end of encoding is either the long window or the stop window, and the window function processing means 1 and 8 for determining the window function, and the window function. Time-frequency conversion means 2 for converting the output of the processing means from the time axis to the frequency axis, and the output of the time-frequency conversion means for the virtual buffer value of the frame at the start of encoding and the virtual buffer value of the frame at the end of encoding. To provide an audio encoding device characterized by comprising quantizing means 3, 4, 9 for encoding such that
According to a fourth aspect of the present invention, there is provided an audio coded signal recording medium in which an audio signal subjected to overlap coding using at least four window functions of a long window, a short window, a start window and a stop window is recorded. , The window function at the start of encoding is either the long window or the start window in consideration of the splicing of codes, and the window function at the end of encoding is either the long window or the stop window, and further, Audio encoding characterized by recording an audio signal encoded such that the virtual buffer value of the frame at the start of encoding and the virtual buffer value of the frame at the end of encoding have the same specified value. A signal recording medium is provided.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of an audio encoding / decoding apparatus of the present invention will be described below with reference to the drawings. FIG. 1 shows a block diagram of an embodiment of an encoding / decoding apparatus when used in the MPEG2-AAC system to which the present invention is applied.

As shown in the block diagram of FIG. 1, the audio encoding / decoding apparatus according to the embodiment of the present invention has a window function processor 1, a time frequency converter 2, a quantizer 3, an auditory model 4, The inverse quantizer 5, the frequency time converter 6, the window function processor 7, the window function designator 8, the bit allocation designator 9, and the coding start / end time detector 10 are included.

First, an embodiment of the audio encoding device of the present invention will be described. The input PCM data is supplied to the window function processor 1, where the window function is multiplied.

The MPEG2-AAC system, as shown in "One example of the shape of window function" in FIG.
Two types of window functions, a long window and two types of window functions, a start window and a stop window, which are used to switch the window function between the short window and the long window are defined. In Unit 1, they are multiplied and overlapped. The start window and the stop window are responsible for connecting the long window and the short window, and their shapes are asymmetrical, as shown in FIG.

Normally, the long window is repeatedly used, but when it detects that a transient signal such as an attack sound is input, the next frame transits to the start window and further the next frame to the short window. I will do it. From a short window to a stop window,
Then go back to the long window.

FIG. 4 is a diagram for explaining the "window function and overlap processing" of the long window and the short window in the MPEG2-AAC system. The window function multiplication is performed on 2048 samples of the audio signal (PCM signal) divided every 1024 samples (frame), and the PCM signal of the previous frame (PCM signal) and the PCM signal of the previous frame. .

In normal encoding, the shape of the window actually used for encoding one frame before and the input P after one frame
The window shape determined by the CM signal determines the window shape used in the current multiplication.

At the start of encoding according to the present invention, the encoding start / end time detector 10 detects the encoding start time, supplies the detected signal to the window function designator 8, and is determined here. The window function information is supplied to the window function processor 1, and the window function used in the frame one frame before is set as a long window. Further, the encoding is performed using the window function determined by the shape of the window determined from the input data after one frame. That is, if the input data of the frame after one frame is suitable for the long window, use the long window,
If it is suitable for the short window, the start window is used for encoding.

As shown in "when the previous frame is a long window and the next is a short window" in FIG. 8, for example, the PCM signal of the processing frame after one frame is encoded by the long window before one frame. If is suitable for coding with a short window, the start window is chosen as the window function for the junction.

Further, as shown in "when the previous frame is a long window and the next is a long window" in FIG. 9, the frame one frame before is encoded by the long window and the frame one frame later is the long window. If suitable for the coding used, a long window is chosen as the window function for the splice.

Further, as shown in the "coding start frame" of FIG. 5, silent data is used for the overlap portion.

Normally, the window function is determined by the window function used for the encoding one frame before and the window function determined from the input data one frame later. In the case of the last frame, the window function designator 8 determines the window function only by the window function used for the encoding one frame before.

That is, according to the present invention, in the case of the last frame, the window one frame before is the window function designator in the case of the long window as shown in "When the window before the last frame is the long window". 8 allows the final window to be encoded using a long window. Further, the present invention is
In the case of the final frame, the window one frame before is used as a stop window by the window function designator 8 in the case of a short window as shown in "when the window before the final frame is a short window". Encode the final frame.

Further, according to the present invention, in the case of the final frame, when the window one frame before is the start window as shown in "When the window before the final frame is the start window" in FIG. Thus, the final frame is encoded using the stop window.

Further, according to the present invention, in the case of the final frame, when the window one frame before is the stop window as shown in "When the window before the final frame is the stop window" in FIG. Thus, the long window is used to encode the final frame.

That is, the present invention is a stop window, or
When the final frame is encoded using the long window, the bitstream is simply connected and then decoded and reproduced, compared with the case where the short window or the start window is used for the final frame. , The change in the level of the reproduced sound becomes gradual, and the deterioration of the sound quality can be made less likely to be felt. That is, as shown in each addition result column in FIG. 7, the level of the reproduced sound is changed regardless of whether the long window is used (FIG. 7A) or the stop window is used (FIG. 7B). The change is gradual, and it is possible to make it difficult to feel the deterioration of sound quality.

FIG. 7 also shows a state in which the code processed by the present invention is edited, and FIG. 7 (a) shows a case where a code sequence terminated in the long window and a code sequence started in the long window are connected. FIG. 7B shows a case where a code string that ends in the stop window and a code string that starts in the long window are connected. It is shown that even if the start part and the end part of the separately encoded bit streams are simply connected (edited), the deterioration of the sound quality of the connection part can be suppressed to the minimum.

Besides this, although not shown here,
The same effect can be obtained even when the code string that ends in the long window and the code string that starts in the start window are connected, or if the code string that ends in the stop window and the code string that starts in the start window are connected. It is a thing.

As described above, the window function processor 1
Then, after multiplying the window function every 2048 samples, MDCT (Modified Discrete Cosine Transf
orm) to convert to the frequency axis.

In the quantizer 3 to which the PCM signal converted to the frequency axis by the time-frequency converter 2 is supplied, the input PCM signal is allocated from the auditory model 4 via the bit allocation designator 9. Quantization is performed based on the bit amount of bit allocation.

The bit allocation designator 9 normally performs the bit allocation from the auditory model 4 as it is to the quantizer 3, and when the coding start frame is quantized and the end frame is quantized. , The detection signal from the encoding start / end time detector 10 is supplied to the bit allocation designator 9, and bit allocation is performed on the quantizer 3 so that the virtual buffer values have the same specified value, and the encoding is performed. To output a bitstream as an encoder output.

Next, an embodiment of the audio decoding apparatus of the present invention will be described below. As shown in FIG. 1, the decoding device side includes an inverse quantizer 5 to which a bit stream is supplied, a frequency time converter 6, and a window function processor 7 that outputs a PCM.

The bit stream from the encoder side is first supplied to the inverse quantizer 5 to be subjected to inverse quantization processing.

The output of the inverse quantizer 5 is the frequency-time converter 6
And converted from the frequency axis to the time axis signal. The conversion output of the frequency-time converter 6 is the window function processor 7
, The window function processor 7 multiplies the window function by this, and outputs PCM data as a decoder output.

The window function processor 7 includes a frequency-time converter 6
Is supplied, the effect of the window function can be removed by further multiplying the frame by the window function used at the time of encoding and adding the preceding and succeeding frames, and the PCM signal is decoded.

In this way, by controlling the code amount so that the virtual buffer values at the start of encoding and at the end of encoding are set to the same specified value, even if bit streams encoded separately are simply combined. , The virtual buffer can be normally decoded without causing overflow or underflow. By the above processing, the bit streams encoded separately can be combined, and the deterioration of the sound quality of the combined part can be minimized.

[0044]

Industrial Applicability The present invention considers the splicing of codes in an audio coding method for performing overlap coding using at least four types of window functions of a long window, a short window, a start window and a stop window. Then, the window function at the start of encoding is either the long window or the start window, and the window function at the end of encoding is either the long window or the stop window, and determines the window function, Converting the output from the determination of the window function from the time axis to the frequency axis,
Coding is performed so that the virtual buffer value of the frame at the start of encoding and the virtual buffer value of the frame at the end of encoding have the same specified value, so that the cohesiveness of the code can be improved, and thus the codes can be encoded separately The integrity of the combined code is guaranteed.

Further, according to the present invention, even if the bit streams encoded separately are simply combined, the deterioration of the sound quality of the connection portion can be suppressed to the minimum.

Further, according to the present invention, the code amount control is performed so that the virtual buffer values at the start of encoding and the virtual buffer value at the end of encoding are set to the same specified value, so that the separately encoded bit streams are simply combined. However, the decryption can be performed normally without causing overflow or underflow.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an audio encoding / decoding device of the present invention.

FIG. 2 is a diagram for explaining stream joining in MPEG1.

FIG. 3 is a diagram for explaining stream joining in MPEG2-AAC.

FIG. 4 is a diagram illustrating “window function and overlap processing” of a long window and a short window in the MPEG2-AAC system.

FIG. 5 is a diagram for explaining a coding start frame.

FIG. 6 is a diagram for explaining four types of window functions used in MPEG2, that is, a short window, a long window, a start window, and a stop window.

FIG. 7 is a diagram showing an example of a decoding result of a connection part in the case of the present invention.

FIG. 8 is a diagram for explaining a case where the frame before the present invention is a long window and the next is a short window.

FIG. 9 is a diagram for explaining a case where the frame before the present invention is a long window and the next is a long window.

FIG. 10 is a diagram for explaining a case where a window before the final frame according to the present invention is a long window.

FIG. 11 is a diagram for explaining a case where a short window is in front of the final frame according to the present invention.

FIG. 12 is a diagram for explaining a case where a start window is before the final frame according to the present invention.

FIG. 13 is a diagram for explaining a case where a stop window is provided before the final frame according to the present invention.

FIG. 14 is a block diagram showing an example of a conventional audio encoding / decoding device.

FIG. 15 is a diagram showing an example of a decoding result of a connection part in the conventional case.

FIG. 16 is a diagram showing a frame structure using MDCT.

FIG. 17 is a state transition diagram showing a normal state transition.

[Explanation of symbols]

1 Window function processor (window function processing means) 2 time frequency converter (time frequency conversion means) 3 Quantizer (quantization means) 4 Auditory model 5 Inverse quantizer 6 Frequency time converter 7 Window function processor 8 Window function designator 9-bit allocation designator 10 Encoding start / end time detector

Claims (4)

(57) [Claims] 1. An audio encoding method for performing overlap encoding using at least four types of window functions of a long window, a short window, a start window, and a stop window, in consideration of code splicing, The window function at the start of encoding is either the long window or the start window, and the window function at the end of encoding is either the long window or the stop window, the window function is determined, and the window function is After multiplying
The output is converted from the time axis to the frequency axis, encoded so that the virtual buffer value of the frame at the start of encoding and the virtual buffer value of the frame at the end of encoding have the same specified value, and the code An audio encoding method characterized by improving the splicing property of the audio. 2. An audio encoding method for performing overlap encoding using at least four types of window functions of a long window, a short window, a start window, and a stop window, in consideration of code splicing. The window function at the start of encoding determines the window function as either the long window or the start window, multiplies the window function, and then converts the output from the time axis to the frequency axis to start the encoding. The code string encoded so that the virtual buffer value of the frame becomes the specified value is the first code string, and the window function at the end of encoding is the window function determined as either the long window or the stop window. After multiplying the window function, the output is converted from the time axis to the frequency axis, and the coded code string is coded so that the virtual buffer value of the frame at the end of coding becomes the specified value. Second As a code sequence, and it connects the termination portion of said the start of the first code string second code string, audio encoding method is characterized in that so as to improve the bonding properties. 3. An audio encoding device for performing overlap encoding using at least four types of window functions of a long window, a short window, a start window and a stop window,
The window function at the start of encoding is either the long window or the start window in consideration of the splicing of codes, and the window function at the end of encoding is either the long window or the stop window. Window function processing means for making a decision, time frequency conversion means for converting the output of the window function processing means from the time axis to the frequency axis, and the output of the time frequency conversion means for the virtual buffer value of the frame at the start of encoding And an quantizing means for performing encoding so that the virtual buffer values of frames at the end of encoding have the same specified value. 4. An audio coded signal recording medium in which an audio signal subjected to overlap coding using at least four window functions of a long window, a short window, a start window and a stop window is recorded. The window function at the start of encoding is either the long window or the start window in consideration of the splicing of codes, and the window function at the end of encoding is either the long window or the stop window, and further the encoding is performed. Audio coded signal recording characterized by recording an audio signal encoded such that the virtual buffer value of the frame at the start and the virtual buffer value of the frame at the end of encoding have the same specified value Medium.