JP5817011B1 - Audio signal encoding apparatus, audio signal decoding apparatus, and audio signal encoding method - Google Patents

Abstract

An audio signal encoding apparatus, decoding apparatus, and encoding method capable of reducing adverse effects of noise, maintaining reproducibility of an original audio signal, and suppressing system damage. A pre-filter 5 of a speech signal encoding apparatus 1 having a pre-filter 5 having a characteristic of amplifying a high-frequency signal component is used together with an encoding unit 4 that encodes an audio signal. A first pre-filter 6 that is used together with the unit 4 and has a characteristic of amplifying a high-frequency signal component and a second pre-filter 7 that has a characteristic of attenuating a high-frequency signal component are connected in series. The first prefilter 6 is configured such that a specific frequency is a pole, and the second prefilter 7 is configured such that the specific frequency is a zero point. [Selection] Figure 1

Description

  The present invention relates to a technique for converting frequency characteristics of an audio signal in encoding of the audio signal.

  2. Description of the Related Art Conventionally, a technique for encoding an audio signal to be transmitted is known for transmitting an audio signal. It is known that various filters are used depending on the purpose of encoding and the nature of the audio signal to be encoded. For example, in an audio coding technique that compresses and encodes an audio signal in a telephone band with high efficiency, a configuration using an auditory weight filter that weights codes in order to improve auditory sound quality is known (for example, Patent Documents). 1).

Japanese Patent Laid-Open No. 10-63297

  Here, various noises may be mixed in the audio signal. However, although the invention described in Patent Document 1 can improve the audible sound quality of the audio signal itself, it cannot reduce the influence of mixed noise.

  Here, the sound component based on natural sound has signal components that are not constant depending on the frequency band, and in particular, the signal components in the high frequency band tend to be smaller than those in the low frequency band. As a result, the audio signal tends to have a smaller energy amount in the high frequency band than in the low frequency band. Then, when processing such as such that an audio signal is uniformly distorted over the entire frequency band, for example, encoding such as distortion data compression, or transmission such as wired communication or wireless communication, An audio signal in a high frequency band with a small amount of components and energy is more susceptible to noise caused by signal distortion. And the S / N ratio of such a frequency band falls and it contributes to sound quality degradation.

  Therefore, before encoding, the pre-filter is used to amplify the high frequency band of the audio signal from the original signal, and then perform processing such as encoding, transmission, and decoding. It can be considered that the influence of noise is reduced by attenuating the amplified high frequency band of the audio signal and returning it to the original state.

  However, in such a configuration, the pole of the pre-filter (the signal amplification factor is theoretically infinite) at a specific frequency (for example, the highest frequency that can be taken by the audio signal processed by the system). ) Is often formed. In this case, if noise with a frequency near or near the pole is mixed in the audio signal before filtering, the noise is amplified to almost infinite in the pre-filter, making it difficult to reproduce the original audio signal and damaging the system. There is a problem that can be given.

  The present invention has been made in view of the above problems, and an audio signal encoding device capable of reducing the adverse effect of noise on an audio signal while maintaining the reproducibility of the original audio signal and suppressing system damage, An object of the present invention is to provide an audio signal decoding device and an audio encoding method.

  In order to solve such a problem, the invention according to claim 1 is an audio signal code having a prefilter which is used together with an encoding means for encoding an audio signal and has a characteristic of amplifying a high frequency signal component. The prefilter is used together with an encoding unit, and has a first prefilter having a characteristic of amplifying the high frequency signal component and a characteristic of attenuating the high frequency signal component. And a second prefilter having a specific frequency as a pole, and the second prefilter is configured such that the specific frequency becomes a zero point. It is characterized by that.

  According to a second aspect of the present invention, in addition to the configuration of the first aspect, the first pre-filter is configured such that a maximum frequency is the pole, and the second pre-filter is the maximum frequency. Is configured to be the zero point.

  According to a third aspect of the invention, in addition to the configuration of the first or second aspect, when the first prefilter performs a Z-transform on the transfer function, only the denominator of the transfer function after the Z-transform Z is configured as a fractional value in which Z exists, and when the transfer function of the first pre-filter after Z conversion is multiplied by the transfer function of the second pre-filter after Z conversion, It is configured as a fractional value in which only Z exists.

The invention described in claim 4 is the audio signal decoding apparatus for decoding an audio signal encoded by the speech signal coding apparatus according to any one of claims 1 to 3, decrypt the audio signals used with decrypt hand stage of performing, with a post-filter having a characteristic of attenuating the signal components of the high frequency, post-filter is configured as an inverse filter of the first pre-filter of the speech signal encoder It is characterized by that.

The invention according to claim 5 is an audio signal encoding method having a frequency characteristic conversion procedure having a characteristic of amplifying a high frequency signal component, which is used together with an encoding procedure for encoding an audio signal , In the frequency characteristic conversion procedure, a first conversion procedure used in the first pre-filter having the characteristic of amplifying the high-frequency signal component, which is used together with the encoding procedure, and the high-frequency signal And a second conversion procedure performed in the second prefilter having the characteristic of attenuating the component, and the first conversion procedure is processed in the first prefilter having a specific frequency as a pole. The second conversion procedure is performed in the second prefilter configured so that the specific frequency becomes a zero point. .

  According to the first and fifth aspects of the present invention, the second prefilter whose specific frequency is zero is connected in series to the first prefilter whose specific frequency is a pole. The second prefilter attenuates the audio signal and noise amplified to an almost infinite signal value in the pole of the first prefilter and in the vicinity of the pole. For this reason, signal values and noise are output as infinite signal values, and the original audio signal cannot be reproduced, or the system is prevented from being damaged by an infinite signal value. . As a result, the reproducibility of the original audio signal can be maintained, and damage to the system can be suppressed while reducing the adverse effect of noise on the audio signal.

  According to the second aspect of the present invention, the second pre-filter having a simple configuration is amplified with respect to the first pre-filter having a simple configuration, which is amplified to an infinite signal value at or near the maximum frequency of the audio signal. With the prefilter, it is possible to accurately attenuate signals and noise amplified to almost infinite.

  According to the invention described in claim 3, the transfer function after the Z conversion of the first prefilter and the multiplication result of the transfer functions after the Z conversion of the first prefilter and the second prefilter are the denominator. By configuring as a fractional value in which Z is present only, the first pre-filter and the second pre-filter can be formed so as to be easily controlled with a simple filter configuration.

  According to the fourth aspect of the present invention, since the post filter can be configured such that the transfer function is formed as the reciprocal of the first pre-filter, the post filter has a simple attenuation characteristic corresponding to the amplification characteristic of the pre-filter. With the post filter having the configuration, the audio signal in which the signal component of the high frequency is amplified can be reliably returned to the audio signal having the original characteristic.

It is a functional block diagram which shows the whole structure of the audio | voice encoding / decoding apparatus which concerns on embodiment of this invention. It is a figure which shows the frequency characteristic of the pre filter of transfer function f _e (z) and the 1st pre filter of transfer function f _s (z) in a speech coding / decoding apparatus same as the above. It is a figure which shows the frequency characteristic of the 2nd pre filter of transfer function _ft (z) in an audio | voice encoding / decoding apparatus same as the above. In the high frequency speech coding and decoding apparatus is a diagram showing the frequency characteristic of the post-filter of the transfer function f _{a (z).}

[Basic configuration]
1 to 4 show an embodiment of the present invention.

  FIG. 1 is a functional block diagram showing the overall configuration of the encoding / decoding apparatus according to this embodiment.

  The encoding / decoding device 1A according to this embodiment is used for encoding and decoding audio signals. In this embodiment, an audio signal generated as a digital signal is applied as a time-series signal processed by the encoding / decoding device 1A. Note that the encoding / decoding device 1A includes an analog / digital conversion unit (not shown), converts a plurality of audio signals supplied as analog signals into digital signals, performs compression processing, and is decompressed. The digital audio signal may be converted into an analog signal and output.

  The encoding / decoding device 1 </ b> A includes an audio signal encoding device 1 and an audio signal decoding device 2. The audio signal encoding device 1 and the audio signal decoding device 2 can mutually transmit and receive signals and data by the communication unit 3.

  The audio signal encoding device 1 includes at least one control unit such as a CPU, and performs an audio signal encoding process, for example, various processes for reducing the signal amount of the audio signal. The audio signal decoding device 2 includes at least one control unit such as a CPU, and performs various processes for decoding the audio signal encoded by the audio signal encoding device 1 and restoring the audio signal before encoding. . The communication unit 3 is a transmission path and transmission equipment such as a shield, a cable, and a wireless communication device used for transmission and reception of signals and data, and a communication interface of the audio signal encoding device 1 and the audio signal decoding device 2. Signals and data are transmitted and received between the audio signal encoding device 1 and the audio signal decoding device 2.

  The audio signal encoding apparatus 1 includes an encoding unit 4 and a pre-filter 5 as “encoding means”. The encoding unit 4 encodes an audio signal. Specifically, the input audio signal is output with a code amount reduced by a predetermined method.

  The pre-filter 5 performs filtering of the audio signal input to the audio signal encoding device 1. That is, the frequency characteristic of the input audio signal is converted and output, and the audio signal is supplied to the encoding unit 4. The prefilter may be realized by executing a program in the CPU, or may be realized by hardware logic or the like.

  The prefilter 5 includes a first prefilter 6 and a second prefilter 7. The first pre-filter 6 and the second pre-filter 7 are configured so that the function of the pre-filter 5 is realized in such a manner that the respective filter characteristics are connected in series.

  The audio signal decoding device 2 includes a decoding unit 8 as “decoding means” and a post filter 9. The decoding unit 8 decodes the audio signal. Specifically, the code amount of the audio signal encoded by the encoding unit 4 is set to the same (or substantially equivalent) code amount as that of the original audio signal, and processing for restoring the audio signal before encoding I do. The post filter 9 performs filtering of the audio signal decoded by the decoding unit 8 and outputs the filtered audio signal to the outside of the audio signal decoding apparatus 2 that has been filtered. The post filter may be realized by executing a program in the CPU, or may be realized by hardware logic or the like.

この伝達関数に示す特性を有することにより、第一のプレフィルタ６は、Ｚ平面上で、ナイキスト周波数に相当するＺ＝−１に極を持つ。そして、第一のプレフィルタ６は、入力された音声信号の高域周波数の信号成分を入力時よりも増幅して出力させる特性を有する。具体的には、図２に示す、伝達関数ｆ_ｓ（ｚ）（点線表示）である第一のプレフィルタ６の周波数特性１０２は、「０．５」の位置の示す、音声信号符号化装置１で処理可能な音声信号の最も高い周波数（つまり、Ａ／Ｄ処理時のサンプリング周波数の半分の周波数）である最高周波数（以下単に「最高周波数」と称する。本明細書において同じ。）において、増幅度が理論上無限大になるように構成される。つまり、第一のプレフィルタ６のみによって音声信号を増幅した場合、最高周波数近傍においては、音声信号や混入したノイズは、ほぼ無限大の値に増幅されて出力されてしまう。 [Filter characteristics of pre-filter]
The first prefilter 6 constituting the prefilter 5 is a primary digital filter. The transfer function f _s (z) of the first prefilter 6 is expressed by the following formula (1).

By having the characteristics shown in this transfer function, the first prefilter 6 has a pole at Z = −1 corresponding to the Nyquist frequency on the Z plane. The first prefilter 6 has a characteristic of amplifying and outputting a high frequency signal component of the input audio signal as compared with the input. Specifically, the frequency characteristic 102 of the first pre-filter 6 that is the transfer function f _s (z) (shown by a dotted line) shown in FIG. 1 is the highest frequency (that is, the frequency that is half the sampling frequency at the time of A / D processing) (hereinafter, simply referred to as “the highest frequency”; the same applies in this specification). The degree of amplification is theoretically infinite. That is, when an audio signal is amplified only by the first prefilter 6, the audio signal and mixed noise are amplified to an almost infinite value and output near the highest frequency.

この伝達関数に示す特性を有することにより、第二のプレフィルタ７は、Ｚ＝−１に零点を有する。また、第二のプレフィルタ７は、ローパスフィルタとして機能する。そして、図３に示すように、伝達関数ｆ_ｔ（ｚ）である第二のプレフィルタ７の周波数特性１０３は、大半の周波数帯（同図における、横軸の値が０〜４．５あたりまで）は増幅量（縦軸の値）が「１．０」で一定であり、高域（同図における、横軸の値が４．５あたり〜５．０まで）において増幅量が減少し、「０．５」の位置の示す最高周波数において、増幅度が０になるように構成される。 On the other hand, the second prefilter 7 constituting the prefilter 5 has a constant τ shown in the following formula (2).
τ∈ (0,1) (2)
Is a digital filter whose parameter is. The transfer function f _t (z) of the second prefilter 7 is expressed by the following equation (3).

By having the characteristic indicated by this transfer function, the second prefilter 7 has a zero at Z = -1. The second prefilter 7 functions as a low pass filter. As shown in FIG. 3, the frequency characteristic 103 of the second pre-filter 7 that is the transfer function f _t (z) is in most frequency bands (in the figure, the value on the horizontal axis is around 0 to 4.5. The amount of amplification (value on the vertical axis) is constant at “1.0”, and the amount of amplification decreases in the high range (in the figure, the value on the horizontal axis is around 4.5 to 5.0). , The amplification degree is set to 0 at the highest frequency indicated by the position of “0.5”.

Incidentally, the second pre-filter 7, the higher the higher the cut-off frequency close constant τ 1, is characteristic of the original signal near the Nyquist frequency is maintained, increased power of while the filtered signals Therefore, it is desirable to set the constant τ to an appropriate value depending on the configuration and application of the prefilter 5. .

これにより、プレフィルタ５は、双１次ディジタルフィルタを形成することになる。図２に、伝達関数ｆ_ｅ（ｚ）のプレフィルタ５の周波数特性１０１を示す。同図に示すように、プレフィルタ５の周波数特性１０１は、第一のプレフィルタ６の周波数特性１０２と同様、入力された音声信号の高域周波数の信号成分を入力時よりも増幅して出力させる特性を有する。ただし、図２に示す通り、プレフィルタ５（周波数特性１０１参照）は、第一のプレフィルタ６（周波数特性１０２参照）よりも高域周波数帯の増幅度が低くなり、最高周波数においても、増幅度が理論上無限大に至ることがないように構成されている。 The prefilter 5 is configured by connecting a first prefilter 6 and a second prefilter 7 in series. Therefore, the transmission of the prefilter 5 function f _{e (z)} becomes a transfer function f _{t (z)} and a value obtained by multiplying the first transmission of a pre-filter function f _s and _(z) a second pre-filter 7, That is, it is represented by the formula (4).

Thereby, the pre-filter 5 forms a bilinear digital filter. FIG. 2 shows the frequency characteristic 101 of the pre-filter 5 of the transfer function f _e (z). As shown in the figure, the frequency characteristic 101 of the pre-filter 5 amplifies and outputs the signal component of the high frequency of the input audio signal as compared with the input frequency, similar to the frequency characteristic 102 of the first pre-filter 6. It has the characteristic to make it. However, as shown in FIG. 2, the prefilter 5 (refer to the frequency characteristic 101) has a lower amplification factor in the high frequency band than the first prefilter 6 (refer to the frequency characteristic 102), and is amplified even at the highest frequency. The degree is theoretically infinite.

  By having such a frequency characteristic 101, the audio signal filtered by the pre-filter 5 amplifies the signal component of the high-frequency than the input at the same time as when only the first pre-filter 6 is filtered. However, unlike the case of being filtered only by the first pre-filter 6, the signal component is not amplified to almost infinite even at the highest frequency.

Note that the degree of amplification of the prefilter 5 can be adjusted as appropriate by changing the value of the constant τ shown in Equation (4). In FIG. 2, the frequency characteristic 101 of the transfer function f _e (z) in which the constant τ shown in the equation (4) is set to 15/16 is displayed as the optimum value in the speech signal encoding device 1 of this embodiment. However, the constant τ may be above or below this value.

  In this embodiment, the audio signal encoded by the encoding unit 4 is filtered by the above-mentioned pre-filter 5 to improve the S / N ratio in the high frequency band where the signal component and energy amount of the audio signal are small. As a result, it is possible to prevent a situation where the reproduction of the original audio signal becomes difficult or the system is damaged due to noise or the like whose theoretical value is amplified to almost infinite.

  Further, the pre-filter 5 is configured as a simple bilinear digital filter of the formula (4), thereby having a simple configuration and reducing the processing load of the audio signal encoding device 1. However, the filtering process can be performed.

[Filter characteristics of post filter]
On the other hand, the post filter 9 is an inverse filter of the first pre-filter 6. The transfer function f _a (z) of the post filter 9 is expressed by the following equation (5).
f _a (z) = 1 + z ⁻¹ (5)
This post filter 9 also functions as a low-pass filter. Specifically, as shown in FIG. 4, in the frequency characteristic 104 of the post filter 9 of the transfer function f _a (z), the frequency increases (as it goes to the right side in the figure), and the degree of amplification gradually decreases. The degree of amplification is configured to be 0 at the highest frequency indicated by the position “”.

  The post filter 9 is a first-order filter as shown in the above equation (5), and is realized by simple addition (a value corresponding to “Z”). Therefore, the audio signal decoding device 2 at the time of filtering is used. The processing load is small. Further, since it is realized by simple addition, it can be applied even after the value of the decimal point or less of the audio signal to be filtered is rounded to an integer. Noise can be removed, and the sound quality of the filtered audio signal can be improved.

[Processing procedure]
Next, a processing procedure of the encoding / decoding device 1A according to this embodiment will be described.

First, the audio signal inputted to the encoding and decoding apparatus 1A is supplied to the audio signal coding apparatus 1, similar speech signal encoding apparatus 1 Ru is carried out steps in the "speech signal encoding method".

  In the audio signal encoding device 1, the frequency characteristics of the audio signal are converted by the pre-filter 5 (frequency characteristic conversion procedure). Specifically, the first prefilter 6 constituting the prefilter 5 amplifies the high frequency signal component (first conversion procedure), and the second prefilter 7 converts the high frequency signal component. Attenuate (second conversion procedure). Since the prefilter 5 has the transfer function shown in the equation (4), the audio signal that has undergone the frequency characteristic conversion procedure of the prefilter 5 is pre-processed in a state in which the signal component having a higher frequency than the original audio signal is amplified. Output from the filter 5.

  The audio signal output from the pre-filter 5 is encoded by the encoding unit 4 to generate an audio signal with a small signal amount (encoding procedure).

  The encoded audio signal is output from the audio signal encoding device 1 and supplied to the audio signal decoding device 2 via the communication unit 3.

  In the audio signal decoding device 2, the encoded audio signal is decoded in the decoding unit 8, and an audio signal having a signal amount before encoding is generated (decoding procedure).

  The decoded audio signal is supplied to the post filter 9. In the post filter 9, the frequency characteristics of the audio signal are converted and the high frequency signal component is attenuated, and the frequency characteristics are substantially the same as those when input to the encoding / decoding device 1A. The audio signal is output to the outside from the encoding / decoding device 1A.

  By such processing, in the encoding / decoding device 1A of this embodiment, the signal component of the high frequency band is emphasized by the filtering of the pre-filter 5, and the signal component of the high frequency band is attenuated by the post filter 9. Thus, by restoring the signal components of the audio signal almost as they were, it is possible to improve the S / N ratio of the audio while suppressing adverse effects caused by the occurrence of noise or the like whose theoretical value is amplified to almost infinite. . In addition, since both the pre-filter 5 and the post-filter 9 are composed of simple primary filters, filtering can be performed with a small amount of computation, so that the encoding / decoding device 1A can be small.

  As described above, in this embodiment, the first prefilter 6 having a specific frequency is a pole, and the second prefilter 7 having a zero specific frequency is connected in series. The second pre-filter 7 attenuates the audio signal and noise amplified to a nearly infinite signal value at or near the pole of the filter 6. For this reason, signal values and noise are output as infinite signal values, and the original audio signal cannot be reproduced, or the system is prevented from being damaged by an infinite signal value. . As a result, the reproducibility of the original audio signal can be maintained, and damage to the system can be suppressed while reducing the adverse effect of noise on the audio signal.

  In this embodiment, the second pre-filter 7 having a simple configuration is compared with the first pre-filter 6 having a simple configuration, which is amplified to an infinite signal value at or near the maximum frequency of the audio signal. Therefore, it is possible to accurately attenuate signals and noise amplified to almost infinite.

  In this embodiment, the transfer function after the Z conversion of the first prefilter 6 and the multiplication result of the transfer functions after the Z conversion of the first prefilter 6 and the second prefilter 7 are only denominators. The first prefilter 6 and the second prefilter 7 can be formed so as to be easily controlled with a simple filter configuration.

  In this embodiment, the post filter 9 can be configured such that the transfer function is formed as the reciprocal of the first pre-filter 6, so that it has a simple configuration with an attenuation characteristic corresponding to the amplification characteristic of the pre-filter. By the post filter, the sound signal in which the signal component of the high frequency is amplified can be reliably returned to the sound signal having the original characteristic.

  In the above embodiment, the audio signal is a digital signal, and the pre-filter 5 and the post-filter 9 are each configured as a digital filter. However, the present invention is not limited to this, and the audio signal to be processed by the encoding / decoding device 1A is not limited thereto. An analog signal may be used, and the prefilter 5 and the postfilter 9 may be configured as analog filters.

  In the above embodiment, the signal processed by the encoding / decoding device 1A is an audio signal. However, the signal is not limited to this, and any signal such as a moving image signal may be used.

  In the above embodiment, the audio signal is communicated from the audio signal encoding device 1 to the audio signal decoding device 2 via the communication unit 3. However, the present invention is not limited to this, for example, the audio signal encoding device 1 and the audio signal encoding device 1. The audio signal decoding device 2 is connected to a database or the like, the audio signal input from the outside and recorded in the database is processed by the audio signal encoding device 1, and the audio signal extracted from the database and output to the outside is decoded by the audio signal It is good also as a structure processed with the apparatus 2. FIG.

  In the above embodiment, the audio signal is transmitted in one direction from the audio signal encoding device 1 to the audio signal decoding device 2 via the communication unit 3. However, the present invention is not limited thereto, and the audio signal encoding device 1 is not limited thereto. And a structure having both functions of the speech signal decoding device 2 are connected by the communication unit 3, and a signal transmitted to the communication unit 3 is processed by the configuration of the speech signal encoding device 1 and received from the communication unit 3. May be processed by the configuration of the audio signal decoding device 2.

  In the above-described embodiment, the poles of the first prefilter 6 and the poles of the second prefilter 7 have the highest frequencies, but the present invention is not limited to this. The pole and the zero point of the second prefilter 7 may be a specific frequency other than the highest frequency (that is, the first prefilter 6 and the second prefilter 7 are respectively a pole and a zero point. It may be a band-pass filter configuration having the same frequency). In this embodiment, the first pre-filter 6 and the second pre-filter 7 are each a primary digital filter, and the pre-filter 5 is a bilinear digital filter. For example, the first prefilter 6 and the second prefilter may be secondary digital filters, and the prefilter 5 may be a bisecondary digital filter.

  The above embodiment is an exemplification of the present invention, and it is needless to say that the present invention is not limited to the above embodiment.

DESCRIPTION OF SYMBOLS 1 ... Audio | voice signal encoding apparatus 2 ... Audio | voice signal decoding apparatus 4 ... Encoding part (encoding means)
DESCRIPTION OF SYMBOLS 5 ... Prefilter 6 ... 1st prefilter 7 ... 2nd prefilter 8 ... Decoding part (decoding means)
9: Post filter 101, 102, 103, 104 ... Frequency characteristics

Claims (5)

An audio signal encoding device having a prefilter having a characteristic of amplifying a high frequency signal component, used together with an encoding means for encoding an audio signal,
The pre-filter is used together with an encoding unit, and a first pre-filter having a characteristic of amplifying the high-frequency signal component and a second pre-filter having a characteristic of attenuating the high-frequency signal component Are connected in series,
It said first prefilter specific frequency becomes extremely, the second prefilter, features and be Ruoto voice signal coding device that the specific frequency is configured to be zero.   The first prefilter is configured such that the maximum frequency of the audio signal that can be processed by the audio signal encoding device is the pole, and the second prefilter is configured such that the maximum frequency is the zero point. The speech signal encoding apparatus according to claim 1, wherein the speech signal encoding apparatus is configured as follows.   The first pre-filter and the second pre-filter are a transfer function of the first pre-filter after Z conversion, a transfer function of the first pre-filter after Z conversion, and after the Z conversion. The speech signal encoding apparatus according to claim 1, wherein the multiplication result of the transfer function of the second pre-filter is configured as a fractional value in which Z exists only in the denominator. An audio signal decoding device for decoding an audio signal encoded by the audio signal encoding device according to any one of claims 1 to 3,
Used with decrypt hand stage to perform decrypt the audio signal, comprises a post-filter having a characteristic of attenuating the signal components of the high frequency,
The speech signal decoding apparatus, wherein the post filter is configured as an inverse filter of the first pre-filter of the speech signal encoding apparatus. A speech signal encoding method having a frequency characteristic conversion procedure having a characteristic of amplifying a high-frequency signal component, which is used together with an encoding procedure for encoding a speech signal ,
In the frequency characteristic conversion procedure, a first conversion procedure used in the first pre-filter having the characteristic of amplifying the high-frequency signal component, which is used together with the encoding procedure, and the high-frequency signal A second conversion procedure performed in a second prefilter having the characteristic of attenuating the component,
The first conversion procedure is performed in the first prefilter having a specific frequency as a pole, and the second conversion procedure is configured such that the specific frequency becomes a zero point. A speech signal encoding method, wherein processing is performed in a second prefilter.

Images