JP2570603B2 - Audio signal transmission device and noise suppression device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for a transmitting device of a voice signal transmitting device. In particular, the present invention relates to a noise suppression device that suppresses noise superimposed on a high frequency band of a transmission device that transmits a broadband audio signal on which noise is superimposed via a line in an audio band.

[0002]

2. Description of the Related Art Conventionally, an audio signal transmission apparatus has transmitted without removing noise superimposed on a wideband audio signal.

[0003]

However, in such a conventional audio signal transmitting apparatus, since the noise superimposed on the audio signal is transmitted without being removed, the voice communication quality deteriorates. there were.

Here, a pitch is proposed by digitally processing a speech signal to obtain a pitch, a noise superimposed section of the speech signal is divided by the pitch, synchronously added, and the noise superimposed section of the speech signal is interpolated with the noise-removed speech thus obtained. There is (Japanese
No. 26123) does not suppress the noise of the high frequency component of the wideband audio signal.

The present invention solves the above problems,
It is an object of the present invention to provide an audio signal transmission device capable of removing noise superimposed on a wideband audio signal and preventing deterioration of audio quality.

[0006]

According to the present invention, there is provided an audio signal transmitting apparatus comprising: a transmitting apparatus for transmitting a wideband voice to be transmitted to a line; and a receiving apparatus for receiving a transmission signal of the transmitting apparatus from the line. An analog-to-digital converter for converting the wideband voice into a time-series digital signal, a conversion unit for converting a digital signal output from the analog-to-digital converter into a frequency-divided signal, and an output of the conversion unit. Compression means for thinning out the high frequency components of the signal and band-compressing the signal; inverse conversion means for converting the output signal of the compression means into a time-series digital signal; and conversion of the output signal of the inverse conversion means into an analog signal A noise suppression device including a digital-to-analog converter is provided.

Further, according to the present invention, the conversion means takes out a finite number of data from the digital signal from the analog-to-digital converter, performs frame processing and windowing processing, and converts the data into a frequency domain for data in the processed frame. The fast Fourier transform operation circuit for performing the conversion by the fast Fourier transform, the inverse transform means converts the output signal of the fast Fourier transform operation circuit into a time-series digital signal by an inverse fast Fourier transform, and performs overlap processing. It may include an inverse fast Fourier transform operation circuit for performing the operation.

Further, the present invention includes a compression means for increasing a thinning rate in a section where high frequency is thinned out, selecting a frequency component having a maximum power in the section as an effective component to be left, and compressing a band. Can be.

[0009] Another aspect of the present invention is a noise suppression device used in the audio signal transmission device. This device can be separately traded separately from the audio signal transmission device.

[0010]

[Function] A wide-band sound on which noise is superimposed is converted from analog to digital to a time-series signal of a wide-band sound, and the time-series signal is converted into a signal in the frequency axis by a fast Fourier transform to remove a high frequency component. Thin out. Furthermore, by inversely transforming the frequency component signal into a time-series signal by inverse fast Fourier transform and performing digital-to-analog conversion,
It is possible to remove noise superimposed on the wideband audio signal and prevent deterioration of audio quality.

[0011]

Embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of an audio signal transmission apparatus according to one embodiment of the present invention.

In FIG. 1, an audio signal transmitting apparatus includes a transmitting apparatus 40 including input means 20 for inputting broadband audio to be transmitted, and transmitting means 30 for transmitting an output of the input means 20 to a line, and transmitting from the line. A receiving device 50 that receives the transmission signal (noise suppression signal) of the device 40.

[0014] The filtrate come when the feature of the present invention herein, the transmission device 40, an analog-digital converter 11 for converting a wideband speech from the input unit 20 in a time-series digital signal, the output of the analog-digital converter 11 Conversion means for converting a digital signal into a frequency-divided signal; compression means 13 for thinning out a high-frequency component of the output signal of the conversion means and band-compressing the signal; converting the output signal of the compression means 13 into a time-series digital signal There is provided a noise suppressing device 10 including an inverse conversion means for converting, and a digital / analog converter 51 for converting an output signal of the inverse conversion means into an analog signal and outputting the analog signal to the transmitting means 30.

The conversion means takes out a finite number of data from the digital signal from the analog-to-digital converter 11 and performs frame processing and windowing processing to convert the data in the processed frame into the frequency domain at high speed. An FFT operation circuit 12 is included as a fast Fourier transform operation circuit for performing a Fourier transform, and the inverse transform means converts an output signal of the FFT transform operation circuit 12 into a time-series digital signal by an inverse fast Fourier transform to perform overlap processing. An inverse FFT operation circuit 14 is included as an inverse fast Fourier transform operation circuit to be performed.

Further, the compression means 13 includes means for increasing a thinning rate in a section where high frequency is thinned out, selecting a frequency component having a maximum power in the section as an effective component to be left, and compressing the band.

The operation of the audio signal transmitting apparatus having such a configuration will be described.

FIG. 2 is a diagram showing a spectrum of a wideband voice of the voice signal transmitting apparatus according to the present invention. The horizontal axis indicates frequency, and the vertical axis indicates signal power. FIG. 2A shows the spectrum of a broadband speech with noise added, and FIG. 2B shows the spectrum of the wideband speech with noise suppression.

In FIG. 1, a wideband audio signal to which noise has been added is converted into a time-series digital signal by an analog-to-digital converter 11 and the signal is converted to an FFT operation circuit 1
In step 2, the signal is converted into a signal on the frequency axis. The spectrum at this point is shown in FIG. Of these frequency components, the frequency is thinned out for high frequency components. As shown in FIG. 2A, in the section where the frequency is thinned, the thinning rate is increased as the frequency increases from the section where the thinning rate is 2: 1 to the section where the thinning rate is 4: 1.

The setting of the frequency component to be left as the effective component is to select the maximum power of the section to be thinned out. Frequency components shown in black in the figure are components to be left as effective components,
Components shown in white are components to be deleted by thinning. That is, the frequency component to be deleted is a portion that does not significantly affect the wideband audio signal even if the frequency component is deleted, and is regarded as a signal having a large specific gravity of the noise component. These thinning processes are performed by the compression means 13. Further, after the compressed frequency signal is returned to a time-series signal by the inverse FFT operation circuit 14, it is converted to an analog signal by the digital-to-analog converter 15 and output as a noise-suppressed wideband sound.

Among the embodiments, the noise suppressing device can be individually traded in the market as a target of the present embodiment, and the user can use it incorporated in the transmitting device of the voice signal transmitting device.

Next, a description will be given of an analog transmission system of a wideband audio signal in which the band is compressed and transmitted using the noise suppressing apparatus of the present invention, the transmission signal is received, and the band is extended.

In this analog transmission system, a wideband voice is band-compressed into a telephone band and transmitted analog, and the wideband voice is reproduced by band expansion on the receiving side. In the band compression, broadband speech is compressed into the telephone band by thinning out middle and high frequency components. In band expansion, pitch detection is performed from the low-frequency component of the compressed voice, and the middle-high frequency component is rearranged and expanded at a frequency position close to an integral multiple of the pitch. Using a subjective evaluation result of a five-stage MOS value (Mean Opinion Score) based on a computer simulation using a 12 kHz band voice, it is shown that 3.1 having a higher evaluation than the telephone band voice 2.4 can be obtained. According to this method, it is possible to realize a broadband audio conference terminal capable of making a call closer to the real voice using an existing analog telephone line.

In the telephone band voice coding of about 4 kHz band mainly for use in communication, in order to improve the efficiency of line use, 32 kbps ADPCM and 16 kbps MPC (Multi Pulse Code) are used from conventional 64 kbps PCM transmission.
c), encoding techniques such as the 8 kbps CELP (Codebook Excited LPC) scheme are being reduced in bit rate (International standard for multimedia encoding published by Maruzen, 1991.
5, edited by Hiroshi Yasuda). In coding for a band wider than the telephone band, SBADPCM (Sub-band ADPCM) for transmitting a 7 kHz band at 64 kbps is typical. Furthermore, recently, MPEG (Moti) capable of compressing and encoding an audio band of about 20 kHz to about 1/6 to 1/12 is used.
on Picture Expert Group) standardized by ISO (Television Society of Japan, Vol. 46, No. 9, pp. 1072-1075, 199).
2) Practical use in transmission terminals is also being studied. In the field of voice communication, there is an increasing demand for a system that transmits a band wider than the telephone band.

All of these compression techniques are digital encoding techniques, and to be applied to a subscriber transmission path, an ISD
Digital networks such as N are targeted. However, it cannot be said that the penetration rate of the digital line in the current subscriber line is high, and the communication through the conventional analog telephone line occupies most. On the other hand, when transmitting broadband voice over an analog telephone line, a method of transmitting digitally coded broadband voice using a high-speed modem is conceivable. However, since the current highest speed modem has a communication speed of about 20 kbps, its feasibility is low.

Here, a description will be given of a transmission system which enables a wideband voice call for an existing analog telephone line which can normally transmit only a band of about 4 kHz or less.
First, the principle of this transmission system will be described, and then, the effectiveness of this analog transmission system will be shown using subjective evaluation results of reproduced sound by simulation.

(A) Configuration of Analog Transmission System FIG. 3 is a block diagram of the analog transmission system, showing an example applied to a two-wire transmission device. In FIG. 3, on the transmitting side, an analog-to-digital converter 11 converts an analog signal of a wideband voice into a digital signal, and converts the analog signal into a digital signal.
In step 2, the time series signal is converted to the frequency domain. Next, the compression means 13 compresses the wideband signal in the frequency domain to the telephone band by a band compression process described later. Furthermore, inverse FF
After returning to the time domain again by the T operation circuit 14, the digital-to-analog converter 15 converts the digital signal into an analog signal and transmits it as analog compressed voice.

On the receiving side, the incoming compressed sound is converted from analog to digital, and the frequency loss of the line is compensated by an EQL (line loss compensation) circuit 52.
In step 3, conversion to the frequency domain is performed. The expansion unit 55 performs a band expansion process described later based on the pitch value detected by the pitch detection circuit 54 from this signal. Further, the signal is returned to the time domain again by the inverse FFT operation circuit 56, and is reproduced by the digital-to-analog converter 57 as expanded voice.

(B) Band Compression Processing FIG. 4 is a flowchart showing band compression processing of the analog transmission system. FIG. 5 is a diagram showing a divided section and a thinning rate in the band compression processing of the analog transmission method. In FIG. 4, band compression processing includes frame processing, windowing processing,
It comprises DFT (Discrete Fourier Transform, Discrete Fourier Transform) thinning processing, IDFT (Inverse DFT), and overlap processing.

First, frame processing for extracting a finite number of data from the time-series data is performed, and the data in the extracted frame is converted into the frequency domain by DFT. In performing DFT, it is necessary to multiply by a time window function so as not to lack continuous periodicity (window missing processing). There are many window functions in the time window function, and it is necessary to select a window function suitable for the purpose (Prentice-Hall, 1975, Theory a
nd applications of digital signal processing, LR
Rabiner and B. Gold). In this analog transmission system, a trapezoidal window or a triangular window is used in order to realize an overlap process for reducing frame boundary distortion by simple addition.

Further, when performing band compression and band expansion, it is necessary to determine a frequency thinning section and a thinning rate in consideration of the frequency characteristics of voice between the transmitting and receiving terminals. The long-term average frequency spectrum of speech is typically 800
Hz is almost flat, and 800 Hz or more is often approximated by frequency characteristics having a slope of -10 dB per octave (Research and practical use report, Vol.4, No.2, pp.245-262,19).
55.1, Instantaneous level distribution and spectrum of speech, T. Miura, J. Koshikawa). Applying this high-frequency attenuation characteristic, a thinning rate for each divided section as shown in FIG. 5 is used. The horizontal axis in FIG. 5 indicates the frequency, and the indicated fraction indicates the thinning rate. For example, in a division section in which 1/4 is described, the number of samples in the frequency domain is thinned out to 1/4, and 1/4
To obtain a compressed audio sample of. This operation indicates compression to 1/4 of the bandwidth. In the low-frequency divided section in which 1/1 is described, the thinning rate is increased from the middle range to the high range without thinning.

The moving position for each frequency component in the band compression is determined between the transmission and the reception by the division and the thinning rate. For example, in a division section with a thinning rate of 1/4, one frequency position in the compression band is determined for a frequency range of 4 sample widths in the extension band.
In this analog transmission system, the arrangement is performed in an ascending order without inverting the frequency value by moving the frequency component.

In the decimation process, first, from the values of the real part and the imaginary part obtained by the DFT operation, the sum of the squares is calculated to obtain the spectral power for each frequency component. Frequency components. Next, after the effective component is subjected to a phase correction, which will be described later, the effective component is moved into a telephone band determined between transmission and reception. For example, let us consider a case in which a divided section having a bandwidth of 32 samples in the number of frequency components of the DFT operation is compressed to a frequency component of 8 samples. First, the spectrum power is obtained for every four samples, the frequency component of one sample having the maximum power is made valid, and the spectrum component is moved to a frequency position within a predetermined telephone band. Repeat this eight times,
By extracting the effective components of eight samples, the band is compressed by 1/4. In a divided section where 2/3 compression is performed, power is obtained for every three samples, and compression is performed using two samples of components that are not the minimum power as effective components.

After the signal whose band has been compressed by thinning out such frequency components is returned to a time-series signal by IDFT, section data in which adjacent frames overlap by an overlap process are added to obtain a continuous compressed signal. .

(C) Band Expansion Processing FIG. 6 is a flowchart showing the band expansion processing of the analog transmission system. In FIG. 6, similarly to the band compression process, the band expansion process performs a DFT operation after performing a flowchart process for extracting a finite number of data and a windowing process for multiplying by a time window function. In the band expansion, the frequency component moved in the band compression process is moved to a frequency position closest to an integer multiple of the frequency component based on the pitch detection result in the frame. In the movement of the frequency component, it is necessary to perform the phase correction described later, as in the compression. Pitch detection is obtained by an autocorrelation method (Basics of speech information processing issued by Ohmsha, 1981, Shozo Saito, Kazuo Nakada) via a low-pass filter (LPF) for frame data. Further, similarly to the band compression processing, a continuous expanded signal can be obtained by performing an IDFT operation for returning to a time-series signal and an overlap processing.

(D) Movement of Frequency Components and Phase Correction in Overlap Processing FIG. 7 is a diagram showing a phase shift due to movement of frequency components in the analog transmission system.

In the case of adding the frame overlap processing to the movement of the frequency components required for the band compression and expansion processing, a phase correction processing for each frequency component is required. In the worst case where the phase correction is not performed, when the phase difference between the frequency components shifted in the addition frame in the overlap section becomes π, the frequency components in the overlap section are reduced to 0 by adding those frames. It will be. FIG. 7 shows a state of a phase shift in the case of 1/2 frame overlap. J-th frame and (J + 1)
FIG. 7A shows a signal waveform when a frequency component is moved in the フ レ ー ム th frame, and FIG.
FIG. 7B shows an example of the movement of a frequency component that does not cause a phase difference. F (2), F _J (3), F _{J + 1} (4) and the like in FIG. 7 are frequency components in the DFT operation, and the frequency position is 2,
Indicated by 3 and 4 indices. FIG. 7A shows a waveform reproduced when the frequency component F (2) at the frequency position 2 is directly moved to the frequency position 3 in each frame. A phase shift π is shown in the overlap section between the waveforms F _J (3) and F _{J + 1} (3) reproduced in the processing in the J-th and (J + 1) -th frames. FIG. 7 (b)
Then, the movement is similarly to F (4), but there is no displacement in the overlap section.

The phase difference is related to the overlap width and the moving distance on the frequency axis, and can be obtained as follows.

First, N pieces of data f (n), (n = 0,
.., N−1) and N frequency components F (κ), (κ = 0,
, N-1) are defined as:

[0040]

(Equation 1) It is.

Now, the input signal g (i) in the time domain i, (i = 0,..., ∞) before being cut out as a frame.
Is assumed to have only the frequency component of the complex number G (u) at the position of the integer u indicating the frequency position, from the definition of IDFT,

[0042]

(Equation 2) Becomes This corresponds to one spectrum.

Assuming that a signal obtained by cutting out N samples of the input signal from 0 in the time domain i is the first frame and F ₁ (n) in the time domain n in the frame, f ₁ (n) = g (i), (N = i: 0 ≦ n <N) (5) When this is subjected to DFT, a component of G (u) is present at a frequency position u as shown in Expression (6), and components at other frequency positions become zero.

[0044]

(Equation 3) Here, assuming that a signal having a component G (u) at a position γ is F ₁ ′ (κ) instead of the position u,

[0045]

(Equation 4) Becomes IDFT is performed on F ₁ ′ (κ), and the signal is represented by f ₁ ′ (n).

[0046]

(Equation 5) Becomes This is returned to the original time domain i by using the conditional expression n = i in Expression (5), and the reproduced signal g ₁ ′ in the first frame is obtained.
(i) is obtained as in equation (9).

[0047]

(Equation 6) Next, M samples (0 <M <N) for the first frame
The following equation is obtained when the shifted point is defined as the second frame, and a signal obtained by cutting out N samples from the input signal g (i) is represented as f ₂ (n).

F ₂ (n), (n = i−M: 0 ≦ n ≦ N) (10) By performing DFT on f ₂ (n) and applying equations (4) and (10),

[0049]

(Equation 7) Becomes Components other than the frequency u are 0, and only f ₂ (u) has a frequency component. Here, similarly to the equation (7), F ₂ ′ (κ) is represented by the following equation.

[0050]

(Equation 8) For F ₂ ′ (κ), the signal for performing IDFT is given by F ₂ ′
Expressed by (n),

[0051]

(Equation 9) Becomes This is returned to the original time domain i by using the conditional expression n = i−M in Expression (10), and a reproduced signal g ₂ ′ (i) in the second frame is obtained.

[0052]

(Equation 10) The phase difference θ between the reproduced signal g ₁ ′ (i) in the first frame and the reproduced signal g ₂ ′ (i) in the second frame is given by Expression (9) and Expression (14).

[0053]

[Equation 11] Becomes Here, arg is a function that extracts a phase angle from a complex number. It can be seen that the phase difference θ is due to the overlap width depending on M and the moving distance (γ-u) of the frequency component. For example, the phase difference θ _1/4 of １ ／ frame overlap where M = N3 / 4 is θ _1/4 = −3π (γ−u) / 2 (16). Moving distances of frequency components 0, 1, 2, 3, 4,
For..., Θ _1/4 is 0, π / 2, π, 3π /
2, 0, π / 2,... That is, in order to realize an ideal phase difference of 0, it is necessary to perform a phase correction that is processed with a remainder of 4 with respect to the moving distance. The phase difference θ _1/2 at 1/2 frame overlap where M = N / 2 is θ _1/2 = −π (γ−u) (17), and the moving distance of the frequency component is 0, 1 , 2, 3,..., Θ _1/2 = 0, π, 0, π,. This can be interpreted as that a phase difference of π occurs only when the value of the index (u and γ) indicating the frequency position in the DFT operation result corresponds to the movement between the odd value and the even value.

These phase corrections basically need to be performed for each frame. Now, it is assumed that the phase difference between adjacent frames is θ, and the phase correction for the addition of the (J−1) th frame and the Jth frame is performed in the Jth frame. Here, J is a natural number satisfying J> 1. Here, in the (J + 1) -th frame, the phase difference θ with respect to the J-th frame is corrected, but if the J-th frame itself is already corrected for the phase difference θ with respect to the (J−1) -th frame, (J + 1) The phase correction in the ()) th frame is 2θ. That is, (J-
The phase correction in the (J + 1) -th frame with respect to the 1) -th frame is as follows: 1 = 0, 1, 2, 3, 4,..., L, θ, 2θ, 3θ, 4θ, 5θ,. Will be cumulative.

For example, if the moving distance on the frequency is 1, 1 /
In the case of 4-frame overlap, since θ _1/4 = π / 2, the phase correction amounts accumulated as θ, 2θ, 3θ, 4θ, 5θ,... Are π / 2, π, 3π / 2,. 0, π / 2, 3
π / 2, ... Similarly, in the case of _1/2 frame overlap, since θ _1/2 = π, θ, 2θ, 3
.., 0,...
π, 0, π,... This means that phase correction should be performed every other frame in the case of 1/2 frame overlap, and the phase correction can be performed with simpler processing than in the case of 1/4 frame overlap. .

(E) Simulation Simulation Setting Conditions FIG. 8 shows simulation setting conditions in which band compression and band expansion are performed between a 12-kHz wideband voice and a telephone-band (3.4 kHz) voice in the analog transmission system. FIG.

FIG. 9 is a diagram showing the relationship between the divided bands used in the simulation of the analog transmission system and the thinning rate. The moving position of the frequency component described in the section (B) is determined by FIG. For example, in the section of the の 間 thinning rate, the frequency positions 96, 97, 98, 99
Frequency positions 69 and 10 for the compressed voice for the four
This indicates that 70 is defined for 0, 101, 102, and 103.

(F) Subjective Evaluation Results FIG. 10 is a diagram showing the subjective evaluation results of the expanded voice of the analog transmission system.

The voice which has been subjected to the band compression / band expansion by the simulation is evaluated with a MOS value (Mean Opini
on Score). As an input signal, a single voice and, voice conference was speech and background noise or that assumes was using the added signal of the BGM (Back GroundMusic). Background noise includes general office ambient noise, BGM
Uses string quartet music. The evaluator watched a signal in a 12 kHz band as a reference signal of evaluation point 5. The evaluation was based on the double blind method (Swedish Br
oadcasting Corporation (SR) Research and Development
Department, 1991.5, The SR Report on The MPEG / Audi
o Subjective Listening Test, Sten Bergman, Christe
r Grewin, Thomas Ryden) using CCIR Rec. 5
62 (International Radio Consultative Committee
(CCIR Recommendation No.562,1990). MOS based on subjective evaluation results by 14 general raters who are not experts in voice evaluation
The values are shown in FIG.

In the expanded voice according to the analog transmission system, higher evaluation is obtained in each sample than in the existing telephone band voice. In addition, the evaluation result of the male voice sample in which the sample itself has few high frequency components is slightly lower than that of the female voice sample, and the difference from the telephone band voice is reduced. However, in the sample of male voice + BGM to which a song with a wide band is added, the MOS value of this analog transmission system is 3.0, which is higher than the telephone band voice 2.2, and the superiority of this system with an expanded band is recognized. Can be

As described above, in the present analog transmission system, a wideband voice is band-compressed into a telephone band and transmitted analog, and the wideband voice is reproduced by expanding the band on the receiving side. In band compression, broadband speech is compressed into the telephone band by thinning out middle and high frequency components. In the band expansion, pitch detection is performed from the low-frequency component of the compressed voice, and the mid-high frequency component is rearranged and expanded at a frequency position close to an integral multiple of the pitch.

Using a subjective evaluation result of a five-stage MOS value based on a computer simulation using a 12 kHz band voice, 3.1 which is higher in evaluation than a telephone band voice evaluation 2.4.
Was obtained. With this analog transmission system, a wideband audio conference terminal capable of making a call closer to the real voice can be realized using an existing analog telephone line.

[0063]

As described above, the present invention has an excellent effect of removing noise added to a voice signal and preventing voice quality deterioration.

[Brief description of the drawings]

FIG. 1 is a block diagram of an audio signal transmission apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a spectrum of a wideband audio signal of a transmission device of the audio signal transmission device of the present invention.

FIG. 3 is a block diagram of an analog transmission system.

FIG. 4 is a flowchart showing band compression processing of the analog transmission system.

FIG. 5 is a diagram showing a division section and a thinning rate in a band compression process of an analog transmission system.

FIG. 6 is a flowchart showing band extension processing of the analog transmission system.

FIG. 7 is a diagram showing a phase shift due to frequency component movement in the analog transmission system.

FIG. 8 is a diagram showing setting conditions of a simulation in which band compression and band expansion are performed between a 12-kHz band wide band voice and a telephone band (3.4 kHz band) voice of the analog transmission system.

FIG. 9 is a diagram illustrating a relationship between a divided band and a thinning rate used in a simulation of the analog transmission scheme.

FIG. 10 is a diagram showing a subjective evaluation result of an expanded voice in the analog transmission system.

[Explanation of symbols]

 REFERENCE SIGNS LIST 10 noise suppression device 11, 51 analog-to-digital converter 12, 53 FFT operation circuit 13 compression means 14, 56 inverse FFT operation circuit 15, 57 digital-analog converter 20 input means 30 transmission means 40 transmission apparatus 50 reception apparatus 52 EQL circuit 54 Pitch detection circuit 55 Decompression means 60 Two-wire to four-wire converter

Claims (2)

(57) [Claims] 1. Broadband voice to be transmitted is transmitted to a line.
A transmitting device, and receiving a transmission signal of the transmitting device from this line.
An audio signal transmission device comprising: a reception device that transmits the wideband audio in a time-series digital format.
Analog-to-digital converter that converts
The output digital signal of the log digital converter is divided by frequency
Conversion means for converting the signal into divided signals, the output of the conversion means
A compression method that thins out the high-frequency components of a signal and compresses the band
Stage, the output signal of this compression means is converted into a time-series digital signal.
Conversion means for converting to a signal, and the output of the inverse conversion means
Digital-to-analog conversion that converts signals to analog signals
Equipped with a noise suppression device, The conversion means is configured to output the analog-to-digital converter
Extraction of finite number of data from digital signal
Processing and windowing processing, and within this processed frame
Fast Fourier transform of the frequency domain
Including a fast Fourier transform operation circuit performed by The inverse transform means outputs the fast Fourier transform operation circuit.
Time-series digitization of force signal by inverse fast Fourier transform
Inverse fast Fourier to convert to overlapped signal
Including a conversion operation circuit, The compression means sets a thinning rate in a section where the high frequency is thinned.
High and leave the frequency component with the highest power in that section
Includes means to select as active ingredient and compress bandwidth That
Characteristic audio signal transmission device. 2. An analog-to-digital converter for converting a wideband voice into a time-series digital signal, a conversion unit for converting an output digital signal of the analog-to-digital converter into a frequency-divided signal, and an output signal of the conversion unit. Compression means for thinning out the high-frequency components and compressing the band, inverse conversion means for converting the output signal of the compression means into a time-series digital signal, and digital analog for converting the output signal of the inverse conversion means to an analog signal In a noise suppression device including a converter, the conversion means takes out a finite number of data from the digital signal from the analog-to-digital converter, performs frame processing and windowing processing, and performs frequency processing on the data in the processed frame. A fast Fourier transform operation circuit that performs the The inverse transform unit includes an inverse fast Fourier transform operation circuit that converts an output signal of the fast Fourier transform operation circuit into a time-series digital signal by inverse fast Fourier transform and performs an overlap process, and the compression unit includes: A noise suppression device comprising: means for increasing a thinning rate in a section where high frequency is thinned, selecting a frequency component having a maximum power in the section as an effective component to be left, and compressing a band.