JPH0646489A - Noise reduction device - Google Patents

Abstract

PURPOSE:To further reduce a noise in point of audible sense by inserting dither to an input side in addition to noise reduction processing by noise shaping. CONSTITUTION:A quantize noise in a quantizer 11 is fetched by an adder 12, and it is fed back via a noise filter 13, and is added on an input audio signal from a terminal 1. The coefficient of the filter 13 is calculated and set by a filter coefficient calculation circuit 14 based on the information of loudness-level characteristic from a circuit 15. A dither signal from a dither generation circuit 21 is added on the input audio signal at the terminal 1, and the frequency characteristic and level of the dither signal generated at the circuit 21 are controlled by an input signal level from a level detector 20.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio signal noise reducing apparatus for reducing a quantization error (noise) generated when an audio signal is quantized.

[0002]

2. Description of the Related Art At present, as a digital audio device for handling a digital audio signal, there is a so-called compact disc (CD) player, a so-called digital audio tape recorder (DAT) and the like. In these digital audio devices, various unified standards are defined. For example, the bit length of a digital audio signal handled by these devices is defined as 16 bits by the unified standard. As a digital audio signal in these digital audio devices, a digital audio signal obtained by encoding an analog audio signal (sound waveform signal) by simple linear quantization such as so-called PCM which is linear pulse coding. I am using.

By the way, in recent years, in the above digital audio equipment, it is desired to obtain a reproduced sound of higher quality in terms of hearing than the reproduced sound actually obtained from the unified standard. In order to obtain such a reproduced sound that is better in hearing, it is considered effective to reduce the noise component contained in the digital audio signal handled by these digital audio devices, for example. In this way, the reproduced sound obtained from the audio signal in which the noise component has been reduced becomes better in terms of hearing with less noise.

As a noise component reducing process of the digital audio signal, for example, a quantization error reducing process by so-called error feedback which is performed at the time of quantizing the audio signal is known. This is to feed back a quantization error (quantization noise, quantization distortion) generated when the audio signal is quantized by the quantizer to the input side of the quantizer via a noise filter.

Regarding the noise reduction device using such error feedback, the applicant of the present application discloses the specifications of Japanese Patent Application No. 2-185552, Japanese Patent Application No. 2-185555, and Japanese Patent Application No. 2-185556. It is proposed in the drawings.

[0006]

SUMMARY OF THE INVENTION Here, the above-mentioned PC
Quantization noise in linear quantization such as M coding has a flat frequency characteristic over the entire frequency band of an audio signal. However, the human ear has different auditory sensitivities depending on the frequency of the sound, and the quantization error reduction process by error feedback is not always effective in terms of hearing. Further, depending on the level of the input signal, the quantization noise may cause a hearing problem.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a noise reduction device that can effectively reduce a quantization error (quantization noise) from the viewpoint of hearing. It is intended.

[0008]

According to the noise reducing apparatus of the present invention, an audio signal in which the quantization error generated in the quantizer is fed back to the input side of the quantizer via a noise filter. In the noise reduction device, the level detection means for detecting the level of the input signal, the dither generation means for generating a dither signal according to the output from the level detection means, and the output from the dither generation means to the input signal. By having an adding means for adding,
The above problems are solved.

Here, the level detecting means may supply an output signal to the dither generating means when the input signal is at a level that is audible. The dither generation means may change the dither level according to the level of the input signal.

The noise reducing apparatus according to the present invention further comprises a quantizing means for quantizing an input audio signal, and a quantizing error by subtracting the input signal to the quantizing means from the output signal of the quantizing means. Subtraction means for extracting (quantization noise), filter means (noise filter) to which the output from the subtraction means is supplied, and combining means such as an adder for combining the output from the filter means with the input signal A dither generating means for generating a random or pseudo-random dither signal, a dither control means for controlling the frequency characteristic and level of the dither generating means, and a means for controlling the characteristic of the filter means, the filter coefficient calculating means And a data generating means for generating data relating to the equal loudness characteristic, the data being supplied to the filter coefficient calculating means. By comprising and a motor control means, to solve the problems described above.

Here, the dither control means further comprises:
It is preferable to include means for controlling the frequency characteristic and level of the dither signal based on the signal level of the input audio signal. Further, it is preferable that the filter control means further includes means for generating control data for correcting the low frequency band component of the filter characteristic. Further, it is preferable that the filter control means includes means for controlling the filter characteristic according to the masking effect of the input audio signal.

Further, the filter control means realizes noise shaping such that an allowable noise spectrum obtained in consideration of the masking effect on the frequency axis or the time axis is given according to the spectrum of the input audio signal. It is preferable to generate a control signal for obtaining a filter characteristic to be used. Furthermore, the filter control means takes the equal loudness curve into consideration when the signal level of the input signal is small, and the allowable noise spectrum in which the masking effect is taken into consideration as the signal level increases, the allowable noise spectrum of the input audio signal becomes It is preferable to generate a control signal that obtains a filter characteristic for realizing noise shaping that is given according to the signal level.

[0013]

In this noise reducing apparatus, the input signal is first adjusted by adding the dither signal, and then the quantization error generated by the quantizing means is input to the quantizing means via the filter means (so-called noise filter). Is returned to the side. The filter coefficient of the noise filter is set based on the information related to the equal loudness curve according to the human auditory characteristics.
According to the present invention, attention is paid to the fact that the equal loudness curve is in accordance with the auditory characteristics of the human ear, and the quantization error is set to a noise filter whose coefficient is set based on the information related to this equal loudness curve. Since it is fed back to the input of the quantizer via, it is possible to reduce the quantization error in the frequency band that is easily heard. Therefore, it is possible to reduce the noise on the hearing and improve the dynamic range on the hearing.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment to which a noise reduction device according to the present invention is applied will be described with reference to the drawings. In FIG. 1, the quantization error (noise) according to the first embodiment.
The schematic construction of the reduction device of FIG. 1 is shown in block diagram form.

In the embodiment of the quantizing error (noise) reducing apparatus shown in FIG. 1, the quantizing error generated by the quantizer 11 is fed back to the input side of the quantizer 11 via the noise filter 13. There is. The filter coefficient of the noise filter 13 is set based on information related to the so-called equal loudness curve RC shown in FIG. 2 according to the human auditory perception characteristics.

Here, a digital audio signal obtained by sampling at an arbitrary sampling frequency is sent to the input terminal 1 of FIG. This digital audio signal is added to the dither signal generated by the dither generation circuit 21, requantized by the quantizer 11, and output from the output terminal 2.

The quantization error reduction device of this embodiment (and the other embodiments described below) consists of two subsystems, a dither system and a noise shaping system. These subsystems act in a complementary manner and their properties can change independently.

The dither subsystem is a level detector 2.
0, dither generation circuit 21 and adder 10. The dither signal generated by the dither generation circuit 21 is added to the input signal by the adder 10. The purpose of the dither subsystem is to generate a so-called dither signal, defined as a random or pseudo-random signal, which is quantized so that the signal error between the quantization error and the input signal is reduced. It has been added to the previous input signal. Correlated errors are more audible than uncorrelated errors, and reducing the correlation between this signal and the error reduces the audible quantization noise. Further, this decorrelation can enhance the effect of the noise shaping subsystem (described later) on the input audio signal in which the correlation between the signal and the error is remarkable.

The dither generation circuit 21 generates dither. The frequency characteristics of this dither signal depend on the specific settings of this embodiment and are determined by the characteristics of the class of input signal to be processed. The spectrum of a typical dither has the so-called white dither (equal at all frequencies) that theoretically forms the optimum decorrelation, and the audible noise by having the energy concentrated in the non-audible regions of the signal spectrum. It includes more reduced so-called high frequency dither and all-zero dither, which is suitable when the input signal is already sufficiently random and need not be decorrelated.

The power spectrum of a non-dithered 1-bit amplitude sine wave is shown in FIG. 4, with the error correlation shown as a resonant spectral peak. The same signal, quantized with white dither, is shown in FIG. 5, with a single spectral peak, indicating that the quantization error is uncorrelated with the signal. FIG. 6 shows the same signal quantized with high frequency dither, but the dither power is large, but the dither is less audible because it is concentrated in an area that is not perceptible to the ear.

Due to the particular settings of this embodiment, the level detector 20 is used to control the type of dither generated by the dither generation circuit 21, or the level of the dither, or both. For example, it is inappropriate to add dither to the input signal consisting of all zeros, so this condition is detected by the level detector 20 and the dither generation circuit 21 is properly controlled. Moreover, low level input signals often require greater dither than high level signals, so the level of dither gradually decreases as the signal level increases.

By the way, in the quantization error reducing apparatus of this embodiment, the quantization error is reduced by the noise shaping subsystem. This noise shaping subsystem subtracts the input to the quantizer 11 from the output of the quantizer 11 to obtain a quantization error generated at the time of quantization in the quantizer 11, and an adder 12; A filter characteristic is set by a filter coefficient, which will be described later, and a noise filter 13 that performs filtering processing on the output from the adder 12 and outputs the result. 10 constitutes a so-called error feedback circuit. Quantization error reduction processing (so-called noise shaping processing) is performed by this error feedback circuit.

Further, the quantizing error reducing apparatus of the present embodiment includes an equal loudness curve generating circuit 15 for generating data of an equal loudness curve RC as shown in FIG. 2 corresponding to the human auditory characteristic, and this equal loudness curve. The filter coefficient calculation circuit 14 calculates the filter coefficient of the noise filter 13 based on the output from the generation circuit 15.

Here, the equal loudness curve RC of FIG.
It corresponds to human hearing characteristics. This curve RC
Is obtained, for example, by connecting a curve segment of the sound pressure for each frequency that sounds as loud as the sound pressure of a pure tone of 1 kHz (loudness). This is an isosensitivity curve of the loudness (loudness). It is also called. According to the equal loudness curve RC as shown in FIG. 2, human hearing is sensitive at around 4 kHz, and even if the sound pressure is 8 to 10 dB lower than the sound pressure of 1 kHz, it is substantially as loud as 1 kHz (loudness). You can hear the sound. On the other hand, for example, 10 kHz
In that case, it is difficult to hear the sound by about 20 dB as compared with around 4 kHz.

Information related to the equal-loudness curve RC (or its approximate curve) (acceptable noise spectrum information) is output from the equal-loudness curve generation circuit 15 and sent to the filter coefficient calculation circuit 14. Therefore, in the equal loudness curve generation circuit 15, this equal loudness curve RC
The filter coefficients are calculated based on the information associated with.
The calculated filter coefficient is further added to the noise filter 13
Sent to. In this way, noise shaping of the audio signal is performed by the error feedback circuit using the noise filter 13 having the filter coefficient based on the information related to the equal loudness curve RC,
The audible dynamic range can be improved.
That is, by performing noise shaping using an allowable noise spectrum (acceptable noise level) obtained by taking the equal loudness curve RC into consideration, noise shaping more effective in hearing is guided, whereby the reproduced sound is reproduced. It is possible to improve the audible dynamic range.

Further, in the present embodiment, a so-called masking effect is taken into consideration when determining the filter characteristic of the noise filter 13. Here, the masking effect is a phenomenon in which a certain signal is masked by another signal and becomes inaudible due to human hearing characteristics. The masking effect includes a masking effect for signals on the time axis (temporal masking) and a masking effect for signals on the frequency axis (simultaneous masking).
Even if there is noise in the part that is masked, the noise cannot be heard due to the masking effect. As a result, when the method by the quantization error reduction process considering the masking effect is executed, the audible dynamic range is improved. In order to determine the filter characteristic in which the masking effect is taken into consideration, the filter coefficient in consideration of the masking effect on the frequency axis is preset in the filter coefficient calculating circuit 14 of the present embodiment. For example,
In order to deal with normal audio speech containing many middle and low frequency components, fixed filter coefficients are set considering the masking effect in the low frequency range. Alternatively, techniques such as generating adaptive filter coefficients depending on the spectrum are used in order to obtain the ability to cope with masking effects corresponding to the spectrum of the input audio signal.

In this way, the filter coefficient calculation circuit 1
The filter coefficients from 4 are obtained with the equal loudness curve RC and the masking effect taken into account. Therefore, the filter characteristic of the noise filter 13 is set based on the fixed or adaptive filter coefficient in which the masking effect is taken into consideration and the filter coefficient associated with the equal loudness curve.

That is, the noise filter 13 at this time
Acts as a filter with the filter characteristics dictated by the curve MR as shown in FIG. 3 resulting from the masking effect and the equal loudness curve. Curve M in FIG.
The filter characteristic designated by R is set to the noise filter 13
, The quantization error spectrum sent to the noise filter 13 changes according to the curve MR. By adding the output from the noise filter 13 to the input audio signal, the quantization error in the quantizer 11 is reduced (based on noise shaping). Here, in the curve MR of FIG. 3, when the equal loudness curve RC of FIG. 2 is considered, it is possible to increase the response in the frequency band lower than 4 kHz (that is, the allowable noise can be increased). The filter characteristic in the frequency band lower than 4 kHz is flat as in the second embodiment described later. The reason why such a mechanism works will be described below.

At 4 kHz or less of the equal loudness curve RC, the change is abrupt even though the bandwidth is not wide. Therefore, when the noise filter 13 adapted to the equal loudness curve RC of 4 kHz or less is created, the dimension of the filter is changed. This is because it becomes expensive. When the dimension of the filter is increased in this way, the configuration becomes complicated and the scale becomes large. However, at this time, since an effect commensurate with the scale of the filter cannot be obtained, the filter characteristic of 4 kHz or less is flat as described above in the second embodiment described below.

That is, FIG. 7 shows a second embodiment of the present invention, in which blocks corresponding to the respective portions in FIG. In FIG. 7, a low frequency band correction control signal generation circuit 16 is further provided, and the low frequency band correction signal generated by this circuit is sent to the filter coefficient calculation circuit 14. Therefore, a flat filter characteristic is realized in the low frequency band as shown by the curve MR in FIG. Further, the low frequency band correction signal is formed in consideration of the masking effect. Generally, the middle and high frequency bands are often used in normal audio sound, and the above-described masking effect in the middle and high frequency bands of the audio sound is effective. Therefore, in the noise filter 13 of this embodiment, the response of the curve MR of the filter coefficient of FIG. 3 is not reduced to such an extent that the response of the equal loudness curve RC of FIG. 2 is reduced (the curve MR is equal loudness. It is more gradual compared to the curve). That is, in order to realize this, as described above, the filter coefficient in which the masking effect is taken into consideration is set. The filter characteristic of the noise filter 13 is set as shown in FIG. 3, and the frequency characteristic of the quantization noise after performing the quantization noise processing using the actual audio voice is shown in FIG.

A third embodiment using the masking effect will be described with reference to FIG. Also in FIG. 9, blocks corresponding to those of FIG. 1 are designated by the same reference numerals.

The quantization noise reducing device of FIG. 9 is configured to feed back the quantization error generated by the quantizer 11 to the input side of the quantizer 11 via the noise filter 11. More specifically, this quantization noise reduction device includes a level detector 19 for detecting the level of an input audio signal, a frequency analysis circuit 17 for frequency-analyzing the input audio signal for each critical band (critical band),
An equal loudness curve generating circuit 15 for generating information based on a so-called equal loudness curve RC as shown in FIG. 2 according to human auditory characteristics, an output from the frequency analyzing circuit 17 and an output from the equal loudness curve generating circuit 15. And a permissible noise spectrum calculating circuit 18 for calculating a permissible noise spectrum on the basis of the obtained composite information. In this quantization noise reduction device, the filter characteristic of the noise filter 13 is the allowable noise spectrum calculation circuit 18 described above.
It is set based on the output information of.

That is, in the quantization error reducing apparatus of the present embodiment, by subtracting the input to the quantizer 11 from the output of the quantizer 11, the quantum generated in the quantizer 11 is quantized. Adder 12 which obtains the digitization error and this adder 12
A so-called error feedback circuit is configured by the noise filter 13 that filters and outputs the output of 1 and the adder 10 that adds the output of the noise filter 13 to the input side of the quantizer 11. Here, the filter characteristic of the noise filter 13 is determined as follows. Actually, the filter coefficient calculation circuit 14 calculates the filter coefficient based on the information of the tolerable noise spectrum of the tolerable noise spectrum calculation circuit 18, which will be described later, and sends this filter coefficient information to the noise filter 13.

Therefore, in the above-mentioned error feedback circuit, a quantization error reduction process (so-called noise shaping process) based on an allowable noise spectrum described later is executed. The signal thus processed is output from the output terminal 2.

By the way, when the quantization error reduction process (noise shaping process) of the audio signal is performed by the above-mentioned error feedback circuit, the process considering the so-called masking of the input signal spectrum is performed, and the dynamic range in the auditory sense is obtained. Can be expanded. By performing processing that considers this masking effect,
For example, noise shaping according to a spectrum of an input audio signal in which a pattern of a signal spectrum is fixed to some extent, that is, noise shaping using an allowable noise spectrum obtained by considering a so-called masking effect of the input audio signal spectrum, which will be described later, is given. be able to.

Alternatively, when the spectrum of the input audio signal changes, there is noise shaping using a permissible noise spectrum adaptive to the spectrum change obtained by considering masking of the spectrum. Here, the masking effect is due to the characteristics of human hearing,
This is a phenomenon in which one signal is masked by another signal and cannot be heard. The masking effect includes a masking effect for signals on the time axis (temporal masking) and a masking effect for signals on the frequency axis (simultaneous masking). Due to this masking effect, even if there is noise in the masked part,
I can't hear this noise. For example, regarding the same-time masking effect, as shown in FIG. 10, when the frequency response of the signal S of a certain frequency is 0 dB, the signal S
This causes the masking effect to act on the curve M (about -15 dB or less).

Furthermore, human auditory characteristics are utilized to divide into so-called critical bandwidths (critical bands), division is performed within this critical bandwidth, and frequency analysis is performed for each of these critical bands. As the division in the critical bandwidth at this time, for example, the input audio signal is converted into a component on the frequency axis by, for example, a fast Fourier transform (FFT), and then the amplitude term Am (m = 0 to 0) of this FFT coefficient. 1024) is the above-mentioned critical bandwidth in which the bandwidth becomes wider in a higher frequency range in consideration of human auditory characteristics, and is divided into, for example, 25-band group Gn (n represents the number of each band, n = 0 to 24) ( Band division).

Further, regarding the frequency analysis for each critical band, for example, the sum of the amplitude terms Am (peak, average, or energy sum of the amplitude terms Am) for each band is calculated by the following equation (1). The so-called Bark spectrum (sum spectrum) Bn thus obtained can be analyzed. Bn = 10 log ₁₀ Cn (Pn) ² (1)

Here, n is 0 to 24, Cn is the number of elements in the nth band (band), that is, the amplitude term (the number of points), and Pn is the peak value of each band. The Bark spectrum Bn of each band is shown in FIG. 11, for example. In the example of FIG. 11, in order to simplify the figure,
The total number of bands is expressed by 12 bands (B _{1 to} B ₁₂ ).
In the frequency analysis circuit 17, the division in the above-mentioned critical bandwidth and the frequency analysis for each band are performed, and the output information is sent to the allowable noise calculation circuit 18.

The equal loudness curve generation circuit 15 generates and outputs information on the equal loudness curve RC. That is, by performing noise shaping using the allowable noise spectrum obtained in consideration of the equal loudness curve, more effective noise shaping can be performed on the sense of hearing. Therefore, the audible dynamic range of the reproduced sound can be improved. Information on the equal loudness curve RC (or its approximate curve) is the same loudness curve generating circuit 15 described above.
Output from the above-mentioned allowable noise spectrum calculation circuit 1
Sent to 8.

Therefore, in the allowable noise spectrum calculation circuit 18, the equal loudness curve generation circuit 15 is used.
The permissible noise spectrum is calculated on the basis of the output information from the frequency analysis circuit 17 and the output information from the frequency analysis circuit 17. At this time, from the Bark spectrum Bn in the frequency analysis circuit 17, convolution is performed (convolution of a predetermined weighting function) in consideration of the influence between bands using the following equation (2). The convolution-processed Bark spectrum Sn for each band is calculated.

Sn = Hn * Bn (2) Here, Hn is a coefficient of convolution. By this convolution, the sum total of the parts shown by the dotted lines in FIG. 11 is obtained. Furthermore, this convolved Bark spectrum Sn and the required S /
Using On (N = 0 to 24) which is the N value, the following (3)
The masking threshold value Tn subjected to the convolution processing is calculated according to the equation (4).

On = N−K × n (3) Tn = Sn−On (4) For example, when N is 38, K can be 1. At this time, the sound quality is less deteriorated. That is, as shown in FIG. 12, the sound having the intensity equal to or lower than the level indicated by each level of the masking threshold value Tn subjected to the convolution processing is masked. Then, the permissible noise level (permissible noise spectrum) TFn is calculated by deconvolving the convolution-processed masking threshold value Tn using the following equation (5). In practice, for example, the direct current (DC) gain Dn of convolution by the coefficient Hn is subtracted.

TFn = Tn−Dn (5)

In the allowable noise spectrum calculation circuit 18, the output information from the frequency analysis circuit obtained as described above is combined with the output information from the equal loudness curve generation circuit 15 described above. The allowable noise spectrum is obtained based on the obtained combined information.

Here, the allowable noise level in the allowable noise spectrum based on the equal loudness curve RC may be lower than the allowable noise level at which the masking effect acts depending on the level of the input audio signal. That is, for example, when the level of the input audio signal is high, the level of the allowable noise spectrum based on the equal loudness curve may be simultaneously masked by the allowable noise level on which the masking effect of the input audio signal acts. is there.

From the above viewpoint, in the present embodiment, the level of the input audio signal is detected by the level detector 19, and the output information from the equal loudness curve generating circuit 15 and the frequency are detected based on the level detection output. Analysis circuit 17
The composition ratio with the output information from is changed. Here, the synthesis of the output information from the equal loudness curve generation circuit 15 and the frequency analysis circuit 17 is performed for each frequency band, for example. In this case, the level detection by the level detector 19 is performed for each band. Therefore,
The synthesis ratio can be changed for each band based on the level detection output for each band. That is, regarding the synthesis information for obtaining the noise spectrum in the allowable level spectrum calculation circuit 18, for example, when the level of the low frequency band of the input audio signal is high and the masking effect in the low frequency band is large, the low frequency band is high. The synthesis information is prepared with a synthesis ratio such that an allowable noise spectrum having a high level and a low level in the level is obtained. On the other hand, for example, when the high frequency level is high and the masking effect in the high frequency range is large, the synthetic information is obtained at a synthetic ratio that gives an allowable noise level such that the high frequency is high and the low frequency is low. Made The information on the allowable noise spectrum obtained in this way is sent to the filter coefficient calculation circuit 14. Then, the filter coefficient calculation circuit 14 outputs a filter coefficient corresponding to the allowable noise spectrum and sends it to the noise filter 13.

From the above, the filter characteristic of the noise filter 13 depends on the filter coefficient based on the allowable noise spectrum obtained by varying the synthesis ratio for each band according to the level of the input audio signal. It becomes a thing. Here, for example, when the level of the input audio signal is flat, it is assumed that the filter characteristics of the noise filter 13 are as shown by the curves MR in FIGS. 13 to 16. At this time, the input audio signal is
For example, as shown in FIG. 13, when the signal S ₁ having a slightly high level in the low frequency range is obtained, the filter ratio is changed by changing the composition ratio as described above.
The low range is slightly raised and the high range is lowered slightly (Fig. 13).
The characteristics can be changed as shown by the medium curve MR ₁ .

Also, for example, the input audio signal is as shown in FIG.
When the signal S ₂ has a low level and a large level as shown in 4, the filter characteristic of the noise filter 13 is
The characteristic can be changed to that shown by a curve MR ₂ in FIG. 14 in which the low range of the curve MR is greatly increased and the high range thereof is greatly decreased. On the contrary, when the input audio signal is the signal S ₃ having a slightly high level in the high range as shown in FIG.
The filter characteristic is changed to the characteristic shown by a curve MR ₃ in FIG. 15 in which the low range of the curve MR is slightly lowered and the high range thereof is slightly raised, and, for example, as shown in FIG. When the signal S _{4 has} a high level and a high level, the characteristics are changed to those shown by a curve MR ₄ in FIG. 16 in which the low range of the curve MR is greatly lowered and the high range is greatly increased. As shown in FIG. 13 to FIG. 16 described above, by changing the filter characteristic, it becomes possible to perform noise shaping according to the human auditory characteristic.

That is, in other words, in other words,
In the apparatus of this embodiment, when the level of the input audio signal is low, the noise filter 13 is made to have a filter characteristic such as an equal loudness curve RC for noise shaping. Further, in order to make the quantization noise less noticeable depending on the level of the input audio signal as the signal level increases, the noise filter 1
The characteristic of 3 is made flat according to the signal level of the input audio signal. Further, when the signal level is low, the noise filter 13 causes the equal loudness curve R
A characteristic such as C is approximated to a flat according to the signal level, and is changed to a noise shaping characteristic (masking characteristic or the like) matched to the signal characteristic. That is,
The noise filter 13 has a characteristic such as an equal loudness curve RC when the signal level is low,
When the signal level is high, the filter characteristic is set in consideration of the masking effect.

In the above-mentioned curve MR showing the filter characteristic when the level of the input audio signal is flat in FIGS. 13 to 16, the equal loudness curve R in FIG.
Considering C, it is possible to raise the level below 4 kHz (that is, the allowable noise can be increased), but the equal loudness curve RC is 4 kHz.
In the following, the change is abrupt even though the bandwidth is not wide. Therefore, if the noise filter 13 is created according to the equal loudness curve RC of 4 kHz or less, the dimension of the filter becomes high. When the dimension of the filter is increased in this way, the configuration becomes complicated and the scale becomes large. However, at this time, since an effect commensurate with the scale of the filter cannot be obtained, in this embodiment, the filter characteristic of 4 kHz or less is flat as described above. Further, in the noise filter 13 of the present embodiment, the masking effect in the middle and low frequencies of the audio sound is effective as described above in the middle and low frequencies which are usually used frequently in the audio sound.
The curve MR of FIGS. 13 to 16 is replaced with the equal loudness curve R of FIG.
C is not lowered (curve MR is a gentler curve than equal loudness curve RC). That is,
To do this, the filter coefficient is set in consideration of the masking effect as described above.

FIG. 17 shows a concrete example of a system configuration in which the apparatus of this embodiment is used for a so-called compact disc (CD) encoder / decoder system. In FIG. 17, an analog audio signal is supplied to the input terminal 31. This analog audio signal is converted into a 20-bit digital signal by the A / D converter 32, and then sent to a 20-bit compatible encoder 33 incorporating the quantization error reducing device of this embodiment. The encoder 33 performs quantization error reduction processing and
The data encoded into bit data is the above CD
Recorded in. The data recorded on the CD is converted into an audio signal by the reproducing circuit 34 and the D / A converter 35 of the existing CD reproducing device, output from the output terminal 36, and reproduced. That is, the data recorded on the CD is
Since the quantization error is reduced by the quantization error reduction device of the present embodiment, the sound obtained by reproducing this CD has a high audible dynamic range.

Further, in FIG. 18, unlike the above CD, for example, 1
A specific example of a system configuration via a medium that records data with 0 bits will be shown. In this case, the analog signal input and supplied to the input terminal is, for example, a signal which is converted into 16-bit digital data by the A / D converter 42, and is a 10-bit encoder 43 in which the quantization error reducing device of this embodiment is built. Sent to. The encoder 43 performs the quantization error reduction process and the data encoded into 10-bit data is recorded on the medium. The data recorded on this medium is the reproduction circuit 4 of the existing reproduction device.
4 and the D / A converter 45 converts the analog signal and outputs the analog signal from the output terminal 46. In this case as well, the audible dynamic range of the obtained reproduced signal is similarly increased.

FIG. 19 shows a specific example in which the apparatus of this embodiment is used in a D / A conversion system which performs oversampling. In this case, the analog signal input and supplied to the input terminal 51 is converted into, for example, 20-bit digital data by the D / A converter 52 that performs oversampling, and the quantization error reduction device 53 of this embodiment is transmitted via the transmission line. Sent to. Quantization error reduction processing is performed by the quantization error reduction device 53, converted into an analog signal via the D / A converter 54, and output from the output terminal 55. As a result, oversampling can be performed, the resolution of D / A conversion can be reduced, and linearity can be increased accordingly.
It becomes easy to create the converter 54.

The present invention is not limited to the above embodiments, and various additions and changes can be made without departing from the scope defined and guaranteed by the claims of the present invention. Of course.

[0056]

As is apparent from the above description, according to the noise reducing apparatus of the present invention, the quantization error generated in the quantizer is fed back to the input side of the quantizer via the noise filter. In the noise reduction device for audio signals, the dither signal generated according to the level of the input signal is added to the above input signal, and when the quantization noise is an input signal that causes a hearing problem. , The noise becomes hard to hear, and the quantization noise does not become a problem in hearing.

Further, by setting the filter coefficient of the noise filter in consideration of the equal loudness curve and the masking effect, for example, the synthesis ratio of the information of the equal loudness curve and the frequency analysis information of the input audio signal becomes the level of the input audio signal. It is possible to reduce perceptual noise and improve the perceptual dynamic range by, for example, setting based on the information obtained by varying the perceptual noise.

Furthermore, the noise reduction device according to the present invention is
For example, by applying it to a digital audio device of an existing standard, it is possible to obtain a reproduced sound having a dynamic range which is audibly expanded as compared with the dynamic range actually obtained from the standard. That is, it is possible to improve the audible dynamic range of the reproduced audio signal by using the same reproducing apparatus as the conventional one without changing the unified standard and maintaining the unified standard.

[Brief description of drawings]

FIG. 1 is a block circuit diagram showing a schematic configuration of an audio signal quantization error reducing apparatus as a first embodiment of the present invention.

FIG. 2 is a characteristic diagram showing an equal loudness curve.

FIG. 3 is a characteristic diagram showing a filter characteristic.

FIG. 4 is a characteristic diagram showing a power spectrum of a 1-bit amplitude sine wave without dither.

FIG. 5 is a characteristic diagram showing a power spectrum of a 1-bit amplitude sine wave with white dither.

FIG. 6 is a characteristic diagram showing a power spectrum of a 1-bit amplitude sine wave with a high frequency dither.

FIG. 7 is a block diagram showing a circuit configuration of a second exemplary embodiment of the present invention.

FIG. 8 is a characteristic diagram showing noise frequency characteristics when an actual audio signal is passed through the device of the embodiment.

FIG. 9 is a block diagram showing a circuit configuration of a third exemplary embodiment of the present invention.

FIG. 10 is a frequency characteristic diagram for explaining a masking effect.

FIG. 11 is a characteristic diagram showing a Bark spectrum.

FIG. 12 is a frequency characteristic diagram showing a masking threshold (threshold).

FIG. 13 is a frequency characteristic diagram showing an example of filter characteristics of a noise filter.

FIG. 14 is a frequency characteristic diagram showing an example of filter characteristics of a noise filter.

FIG. 15 is a frequency characteristic diagram showing an example of filter characteristics of a noise filter.

FIG. 16 is a frequency characteristic diagram showing an example of filter characteristics of a noise filter.

FIG. 17 is a block diagram showing a specific example in which the apparatus of this embodiment is applied to a compact disc encoder / decoder system.

FIG. 18 is a block diagram showing a specific example in which the device of this embodiment is applied to a 10-bit system.

FIG. 19 is a block diagram showing a specific example in which the apparatus of this embodiment is applied to a D / A converter system that performs oversampling.

[Explanation of symbols]

 10, 12 ... Adder 11 ... Quantizer 13 ... Noise filter 14 ... Filter coefficient calculation circuit 15 ... Equal loudness curve generation circuit 16 ...・ ・ ・ Low frequency band correction signal generation circuit 17 ・ ・ ・ Critical bandwidth spectrum analysis circuit 18 ・ ・ ・ Allowable noise spectrum calculation circuit 19, 20 ・ ・ ・ Level detector 21 ・ ・ ・..Dither generation circuits

Claims (10)

[Claims] 1. A noise detection device for an audio signal, wherein a quantization error generated in a quantizer is fed back to an input side of the quantizer through a noise filter, and a level detection for detecting the level of the input signal. Means, a dither generating means for generating a dither signal according to the output from the level detecting means, and an adding means for adding the output from the dither generating means to the input signal. Noise reduction device. 2. The noise reducing apparatus according to claim 1, wherein the level detecting means supplies an output signal to the dither generating means when the input signal is at a level that is audible. 3. The noise reduction device according to claim 1, wherein the dither generation means varies the dither level according to the level of the input signal. 4. A quantizing means for quantizing an input audio signal, a subtracting means for subtracting an input signal to the quantizing means from an output signal of the quantizing means, and an output from the subtracting means are supplied. Filter means, combining means for combining the output from the filter means and the input signal, dither generating means for generating a random or pseudo-random dither signal, and controlling the frequency characteristic and level of the dither generating means. Filter control means for controlling the characteristics of the filter means, including dither control means, data generation means for generating data relating to the filter coefficient calculation means and equal loudness characteristics, and supplying the data to the filter coefficient calculation means. An audio signal noise reduction device comprising: 5. The noise reduction device according to claim 4, wherein the dither control means further includes means for controlling the frequency characteristic and level of the dither signal based on the signal level of the input audio signal. 6. The filter control means further includes means for generating control data for correcting a low frequency band component of the filter characteristic.
The described noise reduction device. 7. The noise reduction device according to claim 5, wherein the filter control means further includes means for controlling the filter characteristic according to a masking effect of the input audio signal. 8. The filter control means obtains a filter characteristic for realizing noise shaping such that an allowable noise spectrum obtained in consideration of a masking effect on the frequency axis is given according to the spectrum of the input audio signal. The noise reduction device according to claim 7, wherein the control signal is generated. 9. The filter control means obtains a filter characteristic for realizing noise shaping such that an allowable noise spectrum obtained in consideration of a masking effect on a time axis is given according to the spectrum of the input audio signal. The noise reduction device according to claim 7, wherein the control signal is generated. 10. The filter control means takes the equal loudness curve into consideration when the signal level of the input signal is small, and the allowable noise spectrum in which the masking effect is taken into consideration as the signal level increases, the allowable noise spectrum is the input audio signal. 8. The noise reduction device according to claim 7, wherein the control signal is generated to obtain a filter characteristic for realizing noise shaping that is given according to the signal level of the signal.