close

Вход

Забыли?

вход по аккаунту

?

JP2014060524

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2014060524
Abstract: To provide a technology for reducing wind noise while leaving a significant audio
signal. A wind noise reduction circuit removes wind noise from an L channel audio signal and an
R channel audio signal acquired by an L channel microphone and an R channel microphone,
respectively. The first subtractor 150 generates a difference component between the L channel
audio signal and the R channel audio signal. The first adder 152 generates a sum component of
the L channel audio signal and the R channel audio signal. The first high pass filter 154 removes
low frequency components of the difference component. The second high pass filter 156
removes low frequency components of the sum component. The second adder 158 generates the
sum of the output signals of the first high pass filter 154 and the second high pass filter 156.
The second subtractor 160 generates the difference between the output signals of the first high
pass filter 154 and the second high pass filter 156. [Selected figure] Figure 8
Wind noise reduction circuit, wind noise reduction method, audio signal processing circuit using
the same, electronic apparatus
[0001]
The present invention relates to audio signal processing.
[0002]
A recording function is incorporated in various electronic devices such as digital video cameras,
digital cameras, mobile phone terminals, personal computers and the like.
10-05-2019
1
When an electronic device with a recording function is used in a windy environment, noise called
wind noise or wind noise is recorded. Wind noise can be reduced to some extent by attaching a
windshield to a microphone, but as a different approach, a technique for reducing wind noise by
signal processing has been proposed (Patent Document 1).
[0003]
The frequency spectrum of wind noise is concentrated below 1 kHz. Therefore, conventionally,
the presence or absence of wind noise is detected based on the frequency spectrum of the audio
signal acquired by the microphone, and when the wind noise is detected, the audio signals of L
channel and R channel are each passed through a high pass filter. The spectral component of
wind noise below the cutoff frequency of the high pass filter was reduced.
[0004]
JP 10-126878 A
[0005]
The conventional wind noise reduction method has a problem that significant audio signals to be
recorded are removed if they have spectral components below the cutoff frequency of the high
pass filter.
[0006]
The present invention has been made in such a situation, and one of the exemplary objects of an
aspect of the present invention is to provide a technique for reducing wind noise while leaving a
significant audio signal.
[0007]
One embodiment of the present invention relates to a wind noise reduction circuit that receives a
first channel audio signal and a second channel audio signal acquired by a microphone of a first
channel and a microphone of a second channel, respectively.
10-05-2019
2
The wind noise reduction circuit comprises a first subtractor for generating a difference
component between the first channel audio signal and the second channel audio signal, and a
first adder for generating a sum component of the first channel audio signal and the second
channel audio signal A first high pass filter for removing low frequency components of the
difference component generated by the first subtractor; a second high pass filter for removing
low frequency components of the sum component generated by the first adder; A second adder
for generating a sum of output signals of the high pass filter and the second high pass filter, and
a second subtractor for generating a difference between output signals of the first high pass filter
and the second high pass filter.
[0008]
According to the inventors of the present invention, wind noise components are often included in
the difference between the audio signals of the first and second channels, and significant audio
signals to be recorded are the audio of the first and second channels. It was recognized that
much is included in the signal sum.
Therefore, by converting the audio signals of the first channel and the second channel into a
difference component and a sum component and removing the respective low frequency
components, wind noise can be reduced while leaving a significant audio signal.
[0009]
The cutoff frequency of the first high pass filter may be set higher than the cutoff frequency of
the second high pass filter.
In this case, the attenuation factor of the difference component containing a large amount of
wind noise components is relatively large, and the attenuation factor of the sum component
containing a large amount of significant audio signal components becomes relatively small. The
wind noise can be reduced while leaving the
[0010]
The wind noise reduction circuit according to an aspect further includes a control unit that
controls the cutoff frequency of each of the first high pass filter and the second high pass filter
10-05-2019
3
based on at least one of the first channel audio signal and the second channel audio signal. Good.
[0011]
The control unit may set the cutoff frequency of the first high pass filter higher than the cutoff
frequency of the second high pass filter.
[0012]
The control unit generates data indicating wind noise intensity based on at least one of the first
channel audio signal and the second channel audio signal, and the larger the value of the data,
the cutoff frequency fc1 of the first high pass filter, The cutoff frequency fc2 of the two high pass
filters may be increased.
[0013]
The control unit sets the cutoff frequency fc1 of the first high pass filter and the cutoff frequency
fc2 of the second high pass filter to equal minimum frequencies fMIN when the value of the data
is lower than a predetermined minimum value, and the value of the data is predetermined. When
it is lower than the maximum value, the cutoff frequency fc1 of the first high pass filter is set to
the first maximum frequency fMAX1, and the cutoff frequency fc2 of the second high pass filter
is set to the second maximum frequency fMAX2 lower than the first maximum frequency fMAX1.
When the data value is larger than the minimum value and smaller than the maximum value, the
cutoff frequency fc1 of the first high pass filter is changed in the range from the minimum
frequency fMIN to the first maximum frequency fMAX1, and the cutoff frequency of the second
high pass filter The range of the minimum frequency fMIN to the second maximum frequency
fMAX2 is In may be changed.
[0014]
The minimum value and the maximum value may be set externally.
[0015]
The minimum frequency fMIN, the first maximum frequency fMAX1, and the second maximum
frequency fMAX2 may be externally settable.
[0016]
10-05-2019
4
The control unit includes a detection subtractor that generates a difference component between
the first channel audio signal and the second channel audio signal, and controls the cutoff
frequencies of the first high pass filter and the second high pass filter based on the difference
component. Good.
According to the findings of the present inventors, when the wind noise becomes large, the
difference between the audio signals of the first channel and the second channel becomes large,
so that the difference component of the audio signals of the first channel and the second channel
is generated. Thus, the presence or absence of wind noise can be accurately detected, and the
cutoff frequency can be set appropriately.
[0017]
The control unit may further include a detection adder that generates a sum component of the
first channel audio signal and the second channel audio signal.
The control unit may control the cutoff frequencies of the first high pass filter and the second
high pass filter based on the ratio of the difference component to the sum component.
According to the findings of the present inventors, since the sum of the audio signals of the first
channel and the second channel changes in accordance with the volume of the significant audio
signal to be recorded, the ratio of the difference component to the sum component is taken.
Thus, the relative amount of wind noise relative to the significant audio signal can be estimated
and reflected in the cutoff frequency of the high pass filter.
[0018]
The control unit may set the cutoff frequency of the first high pass filter and the second high
pass filter to be higher as the amplitude of the difference component is larger.
[0019]
In one aspect, the wind noise reduction circuit may be integrated on one semiconductor
substrate.
10-05-2019
5
"Integrated integration" includes the case where all of the circuit components are formed on a
semiconductor substrate, and the case where the main components of the circuit are integrally
integrated. A resistor, a capacitor or the like may be provided outside the semiconductor
substrate.
[0020]
Another aspect of the present invention relates to an audio signal processing circuit.
The audio signal processing circuit includes a first amplifier for amplifying an output signal of
the microphone of the first channel, a second amplifier for amplifying an output signal of the
microphone of the second channel, and a digital first channel for the output signal of the first
amplifier. Receiving a first A / D converter for converting an audio signal, a second A / D
converter for converting an output signal of a second amplifier to a digital second channel audio
signal, a first channel audio signal and a second channel audio signal, Even with any of the wind
noise reduction circuits described above that reduce wind noise, and a digital signal processing
unit that performs predetermined signal processing on the first channel audio signal and the
second channel audio signal that have passed through the wind noise reduction circuit. Good.
[0021]
Another aspect of the present invention relates to an electronic device.
The electronic device may include the above-described audio signal processing circuit.
[0022]
It is to be noted that any combination of the above-described constituent elements, or one in
which the constituent elements and expressions of the present invention are mutually replaced
among methods, apparatuses, systems, etc. is also effective as an aspect of the present invention.
[0023]
10-05-2019
6
According to the wind noise reduction circuit of the present invention, wind noise can be reduced
while leaving a significant audio signal.
[0024]
It is a block diagram which shows the structure of the electronic device provided with the wind
noise reduction circuit which concerns on 1st Embodiment.
It is a figure which shows the frequency characteristic of a high pass filter.
FIGS. 3A and 3B are diagrams showing the relationship between the amplitude of the difference
component and the cutoff frequency of the high pass filter.
It is a block diagram showing an example of composition of a high pass filter. It is a block
diagram showing an example of composition of a control part. It is a wave form diagram of a sum
component and a difference component. It is a perspective view which shows the electronic
device carrying an audio signal processing circuit. It is a block diagram which shows the
structure of the wind noise reduction circuit which concerns on 2nd Embodiment. It is a figure
which shows the relationship between the intensity | strength of the wind sound which the
control part detected, and the cut-off frequency of a 1st high pass filter and a 2nd high pass
filter. 10 (a) and 10 (b) are diagrams showing the relationship between wind noise intensity and
cutoff frequency according to the third embodiment.
[0025]
Hereinafter, the present invention will be described based on preferred embodiments with
reference to the drawings. The same or equivalent components, members, and processes shown
in the drawings are denoted by the same reference numerals, and duplicating descriptions will be
omitted as appropriate. In addition, the embodiments do not limit the invention and are merely
examples, and all the features and combinations thereof described in the embodiments are not
necessarily essential to the invention.
[0026]
10-05-2019
7
First Embodiment FIG. 1 is a block diagram showing a configuration of an electronic device 1
provided with a wind noise reduction circuit 100 according to a first embodiment. The electronic
device 1 includes an Lch microphone 2L, an Rch microphone 2R, and an audio signal processing
circuit 10.
[0027]
The first microphone (Lch microphone) 2L and the second microphone (Rch microphone) 2R
respectively convert acoustic signals into analog electrical signals (audio signals) S1L and S1R.
The audio signal processing circuit 10 receives the audio signals S1L and S1R, removes wind
noise from them, performs predetermined signal processing, and supplies the processed signal to
a circuit (shown) in the subsequent stage.
[0028]
The audio signal processing circuit 10 includes a wind noise reduction circuit 100, a first
amplifier (Lch amplifier) 200L, a second amplifier (Rch amplifier) 200R, an auto level controller
202, a first A / D converter (Lch A / D converter) 204L, A 2A / D converter (Rch A / D converter)
204R and a digital signal processing unit 206 are provided.
[0029]
The Lch amplifier 200L amplifies the Lch audio signal S1L.
The Rch amplifier 200R amplifies the Rch audio signal S1R. The auto level controller 202
controls the gains of the Lch amplifier 200L and the Rch amplifier 200R such that the volume is
constant.
[0030]
The Lch A / D converter 204L performs analog / digital conversion on the output S2L of the Lch
amplifier 200L to generate an L channel digital audio signal S3L. Similarly, the Rch A / D
10-05-2019
8
converter 204R performs analog / digital conversion on the output S2R of the Rch amplifier
200R to generate an R channel digital audio signal S3R.
[0031]
The wind noise reduction circuit 100 receives the Lch audio signal S3L and the Rch audio signal
S3R, and removes the wind noise component from them. The digital signal processing unit 206
performs predetermined signal processing on the audio signals S4L and S4R from which the
wind noise has been removed to generate audio signals S5L and S5R.
[0032]
The above is the entire configuration of the electronic device 1. Subsequently, the configuration
of the wind noise reduction circuit 100 will be described.
[0033]
The wind noise reduction circuit 100 includes a high pass filter 110 and a control unit 130. The
high pass filter 110 removes low frequency components of the L channel audio signal S3L and
the R channel audio signal S4R. The cutoff frequency fc of the high pass filter 110 is configured
to be variable.
[0034]
The control unit 130 generates a difference component (S3L-S3R) of the L channel audio signal
S3L and the R channel audio signal S3R, and controls the cutoff frequency fc of the high pass
filter 110 based on the difference component (S3L-S3R). Of course, (S3R-S3L) may be used as the
difference. Specifically, the cutoff frequency fc of the high pass filter 110 is set higher as the
level (amplitude) of the difference component (S3L to S3R) is larger.
[0035]
10-05-2019
9
The above is the basic configuration of the wind noise reduction circuit 100. Subsequently, the
operation will be described. Now, consider a virtual sound source. Assuming that the distance
between the sound source and the Lch microphone 2L is DL and the distance between the sound
source and the Rch microphone 2R is DR, the audio signals S1L and S1R acquired by the
microphones 2L and 2R are the difference between the distances DL and DR (DL-DR), It depends
on the wavelength of the audio signal.
[0036]
Empirically, the frequency of wind noise is mainly 100 Hz to 400 Hz, while the frequency of a
significant audio signal to be recorded by the microphone is often dominated by frequency
components higher than wind noise. As the frequency of the audio signal is higher, the phase
difference between the L channel and R channel audio signals can be ignored, and thus the inphase component becomes larger. That is, when the difference (S3L-S3R) between the L channel
audio signal S3L and the R channel audio signal S3R is calculated, a significant audio signal
having many in-phase components is canceled out, and wind sound having many differential
components is dominantly included. become.
[0037]
FIG. 2 is a diagram showing the frequency characteristic of the high pass filter 110. As shown in
FIG. As the amplitude of the difference component (S3L-S3R) increases, the cutoff frequency fc
increases in the directions of fc1, fc2, and fc3. The frequency spectrum of wind noise is in the
range of fwind.
[0038]
As the amplitude of the difference component (S3L-S3R) increases, that is, as the wind noise
increases, the cutoff frequency fc of the high pass filter 110 increases, and as a result, the
passing gain of the wind noise frequency band fwind decreases.
[0039]
FIGS. 3A and 3B show the relationship between the amplitude of the difference component and
10-05-2019
10
the cut-off frequency fc of the high pass filter 110. FIG.
The horizontal axis indicates the amplitude of the difference component (S3L-S3R), that is, the
intensity of wind noise, and the vertical axis indicates the cutoff frequency fc. In FIG. 3A, in the
range where the amplitude of the difference component is smaller than the predetermined
minimum value MIN, the cutoff frequency fc is the predetermined minimum value fMIN, and the
cutoff is in the range where the amplitude is larger than the predetermined maximum value
MAX. The frequency fc is a predetermined maximum value fMAX, and the cutoff frequency fc
changes continuously and linearly when its amplitude is included in the range between the
minimum value MIN and the maximum value MAX.
[0040]
In FIG. 3B, when the amplitude of the difference component is included in the range between the
minimum value MIN and the maximum value MAX, the cutoff frequency fc changes stepwise.
[0041]
According to the wind noise reduction circuit 100 of FIG. 1, the presence or absence or strength
of the wind noise is suitably detected based on the difference component between the L channel
and the R channel, and the cutoff frequency fc of the high pass filter 110 according to the result.
Can be properly controlled.
[0042]
The specific configurations of the high pass filter 110 and the control unit 130 are not
particularly limited, but in the following, configuration examples thereof will be described.
[0043]
FIG. 4 is a block diagram showing a configuration example of the high pass filter 110. As shown
in FIG.
The high pass filter 110 of FIG. 4 includes an Lch high pass filter 110L that removes low
frequency components of the L channel audio signal S3L, and an Rch high pass filter 110R that
removes low frequency components of the R channel audio signal S3R.
10-05-2019
11
The cutoff frequency fc of each of the high pass filters 110L and 110R is set by the control unit
130 to an equal value.
[0044]
FIG. 5 is a block diagram showing a configuration example of the control unit 130. As shown in
FIG.
The control unit 130 includes, in addition to the detection subtractor 132, a detection adder 134,
a first low pass filter 136, a second low pass filter 138, a first smoothing circuit 140, a second
smoothing circuit 142, and a detection unit 144. A cutoff frequency setting unit 146 is provided.
[0045]
The detection subtractor 132 generates a difference component S10 = (S3L−S3R) between the L
channel audio signal S3L and the R channel audio signal S3R. As described above, the control
unit 130 detects the presence or absence of wind noise and strength, and controls the cutoff
frequency fc of the high-pass filter 110 based on the difference component S10.
[0046]
In order to detect wind noise with higher accuracy, the control unit 130 of FIG. 5 considers not
only difference components but also sum components. The detection adder 134 generates a sum
component S11 (S3L + S3R) of the L channel audio signal S3L and the R channel audio signal
S3R. The control unit 130 controls the cutoff frequency fc of the high-pass filter 110 based on
the ratio S10 / S11 = (S3L−S3R) / (S3L + S3R) between the difference component S10 and the
sum component S11.
[0047]
The difference component S10 is input to the detection unit 144 via the first low pass filter 136
10-05-2019
12
and the first smoothing circuit 140, and the sum component S11 is detected via the second low
pass filter 138 and the second smoothing circuit 142. It is input to the part 144. The detection
unit 144 calculates data S16 = S14 / S15 indicating the ratio of the difference components S14
and S15. The cutoff frequency setting unit 146 sets the cutoff frequency fc of the high pass filter
110 according to the data S16.
[0048]
The cutoff frequency setting unit 146 may have a table indicating the relationship between the
data S16 and the cutoff frequency fc. Alternatively, the cutoff frequency fc may be calculated by
inputting the data S16 into a predetermined arithmetic expression.
[0049]
Wind noise has a large differential component between L channel and R channel and a small inphase component. On the other hand, in the significant audio signal in which the high frequency
component is dominant, the differential component of the L channel and the R channel is small
and the in-phase component is large. Therefore, when the sum component of the L channel and
the R channel is calculated, the wind sound component is canceled and the significant audio
signal is dominantly included. That is, it can be understood that the sum component S11
corresponds to the volume of the significant audio signal.
[0050]
FIG. 6 is a waveform diagram of the sum component S11 and the difference component S10.
Wind noise occurs in period (i) and wind noise is zero in period (ii). While the amplitude of the
sum component S11 is irrelevant to the presence or absence of wind noise and strength, the
amplitude of the difference component S10 increases in period (i) and decreases in period (ii).
That is, the amplitude of the difference component S10 has a correlation with the intensity of
wind noise.
[0051]
According to the control unit 130 of FIG. 5, the relative amount of wind noise to a significant
audio signal can be estimated by calculating the ratio S10 / S11 of the difference component S10
10-05-2019
13
and the sum component S11, and the cutoff of the high pass filter 110 It can be reflected on the
frequency fc.
[0052]
The cutoff frequencies of the first low pass filter 136 and the second low pass filter 138 are set
to about 400 Hz.
Pass the frequency component of wind noise. These are provided to increase the detection
accuracy of wind noise detection. By providing the low pass filters 136 and 138, wind noise can
be detected more accurately. The second low pass filter 138 may be omitted, and the first low
pass filter 136 may be omitted.
[0053]
The first smoothing circuit 140 takes a moving average of the difference component S12. If the
moving average time is long, that is, the number of times of averaging is increased, the response
speed of the difference component S14 input to the detection unit 144 becomes slow. By
providing the first smoothing circuit 140, the sensitivity to sudden or weak wind can be set
according to the moving average time. The moving average time is preferably settable from an
external microcomputer. Thereby, the sensitivity can be optimized according to the situation
where the wind noise reduction circuit 100 is used.
[0054]
The second smoothing circuit 142 is provided to balance the difference components S14 and
S15. Instead of the second smoothing circuit 142, a delay circuit for timing adjustment may be
provided.
[0055]
The control unit 130 may reflect the offset value DOFS in addition to the difference component
10-05-2019
14
S14 and the sum component S15 when calculating the data S16. The offset value DOFS is
preferably enabled from an external microcomputer according to the set value DEXT.
[0056]
The distance between the microphones 2L and 2R of the L channel and the R channel differs
depending on the electronic device 1 on which the wind noise reduction circuit 100 is mounted.
When the distance between the microphones 2L and 2R is different, the amplitude of the
difference component S14 with respect to the wind sound of the same volume is also different.
Therefore, the wind noise reduction circuit can be used on various platforms with different
distances between the microphones by introducing the offset value DOFS and changing the offset
value DOFS according to the distance between the microphones.
[0057]
The offset value DOFS may be set according to the gain g of the Lch amplifier 200L and the Rch
amplifier 200R in addition to the externally set value DEXT. Thereby, even in a situation where
the gain g of the Lch amplifier 200L and the Rch amplifier 200R changes, the strength of wind
noise can be detected based on the data S16 '.
[0058]
When the gain of the amplifier changes, the ratio of the difference component and the sum
component to the external set value changes, but by changing the offset value according to the
gain, the influence of the gain in wind noise detection can be reduced.
[0059]
Subsequently, applications of the audio signal processing circuit 10 will be described.
FIG. 7 is a perspective view showing an electronic device on which the audio signal processing
circuit 10 is mounted. FIG. 7 is a digital camera which is an example of the electronic device.
10-05-2019
15
[0060]
The digital camera 800 includes a housing 802, a lens 804, an imaging device (not shown), an
image processing processor, and a recording medium. In addition to that, the digital camera 800
includes an Lch microphone 2L, an Rch microphone 2R, and an audio signal processing circuit
10.
[0061]
In addition, the electronic device may be a digital video camera, a voice recorder, a mobile phone
terminal, a personal handy-phone system (PHS), a personal digital assistant (PDA), a tablet
personal computer (PC), an audio player, or the like.
[0062]
Second Embodiment FIG. 8 is a block diagram showing a configuration of a wind noise reduction
circuit 100a according to a second embodiment.
The wind noise reduction circuit 100a includes a high pass filter 110a and a control unit 130a.
[0063]
The high pass filter 110a includes a first subtractor 150, a first adder 152, a first high pass filter
154, a second high pass filter 156, a second adder 158, a second subtractor 160, a first
coefficient circuit 162, a first coefficient Circuit 164 is included.
[0064]
The first subtractor 150 generates a difference component S21 between the L channel audio
signal S3L and the R channel audio signal S3R.
The first adder 152 generates a sum component S22 of the L channel audio signal S3L and the R
channel audio signal S3R. The first high pass filter 154 removes low frequency components of
the difference component S21 generated by the first subtractor 150. The second high pass filter
10-05-2019
16
156 removes low frequency components of the sum component S22 generated by the first adder
152. The first high pass filter 154 and the second high pass filter 156 are configured to be able
to set the cutoff frequencies fc1 and fc2 independently.
[0065]
The second adder 158 generates a sum S25 of the output S23 of the first high pass filter 154
and the output S24 of the second high pass filter 156. The second subtractor 160 generates a
difference S26 between the output S23 of the first high pass filter 154 and the output S24 of the
second high pass filter 156. The first coefficient circuit 162 and the first coefficient circuit 164
multiply the outputs S 25 and S 26 of the second adder 158 and the second subtracter 160 by a
coefficient 1⁄2.
[0066]
The control unit 130a detects wind noise based on at least one of the audio signals S3L and S3R,
and controls the cutoff frequencies fc1 and fc2 of the first high pass filter 154 and the second
high pass filter 156 based on the detection result. .
[0067]
The method of detecting wind noise by the control unit 130a is not particularly limited, and may
be based on the difference S3L-S3R as in the first embodiment.
In this case, the detection subtractor 132 and the detection adder 134 in FIG. 5 may be used as
the first subtractor 150 and the first adder 152 in FIG.
[0068]
Alternatively, as in the prior art, the control unit 130a monitors at least one of the audio signals
S3L and S3R, and detects wind noise based on components below a predetermined frequency
(for example, 400 Hz) including the wind noise spectrum. It is also good.
[0069]
10-05-2019
17
FIG. 9 is a diagram showing the relationship between the intensity of wind noise detected by the
control unit 130 and the cutoff frequencies fc1 and fc2 of the first high pass filter 154 and the
second high pass filter 156.
As in the first embodiment, as the wind noise intensity increases, the cutoff frequencies fc1 and
fc2 rise while maintaining the relationship fc1> fc2. Specifically, in the region where the wind
noise intensity is lower than the minimum value MIN, the cutoff frequencies fc1 and fc2 are set
equal to fMIN. When the intensity becomes larger than the minimum value MIN, the cutoff
frequencies fc1 and fc2 rise while maintaining the relationship of fc1> fc2. When the wind noise
intensity becomes larger than the maximum value MAX, the cutoff frequencies fc1 and fc2 are
fixed to the respective maximum values fMAX1 and fMAX2. Naturally, the cutoff frequency may
change stepwise as shown in FIG. 3 (b).
[0070]
The above is the configuration of the wind noise reduction circuit 100a. Subsequently, the
operation will be described. The advantage of the wind noise reduction circuit 100a is clarified
by the contrast with the high pass filter 110 of FIG.
[0071]
The wind noise spectrum is equally included in each of the audio signals S3L and S3R, and the
significant audio signal spectrum is also equally included in each of the audio signals S3L and
S3R. Therefore, when the high-pass filter 110 of FIG. 4 is used, when wind noise is attenuated,
significant audio signals having the same spectrum as the wind noise are also attenuated. That is,
there arises a problem that significant low frequency components of the audio signal are
attenuated unnecessarily.
[0072]
As described in the first embodiment, a large amount of wind noise is included in the difference
component of audio signals S3L and S3R, and conversely, a significant audio signal is included in
the sum component of audio signals S3L and S3R. . According to the second embodiment, the
10-05-2019
18
high pass filters 154 and 156 are provided for each of the difference component and the sum
component, and the cut-off frequencies fc1 and fc2 thereof are individually set, whereby the
significant components included in the sum component are obtained. It is possible to suitably
attenuate the wind noise spectrum included in the difference component without unnecessarily
attenuating the spectrum of the audio signal.
[0073]
Third Embodiment In the first and second embodiments, the cutoff frequency fc (fc1, fc2) linearly
changes with respect to the wind intensity. As a result of examining the control, the present
inventors came to recognize the following problems. In the control of FIG. 3 (a), (b) or FIG. 9, the
slope changes discontinuously when the cutoff frequency fc becomes larger than the minimum
value fMIN. Therefore, when the wind noise intensity changes across the minimum value, the cutoff frequency fc changes rapidly, which causes the audio signal passed through the wind noise
reduction circuit to be uncomfortable.
[0074]
The third embodiment described below is a technology that can be combined with the first and
second embodiments.
[0075]
FIGS. 10 (a) and 10 (b) are diagrams showing the relationship between wind noise intensity x and
cutoff frequency y according to the third embodiment.
In FIGS. 10 (a) and 10 (b), the cutoff frequency is set to gradually increase in the vicinity of the
minimum value of wind intensity x. From another point of view, the slope of the cutoff frequency
dy / dx is determined to change continuously around the minimum value MIN.
[0076]
More specifically, the cutoff frequency y may be increased according to a quadratic function with
respect to the wind noise intensity x as shown in FIG. 10 (a). That is, the cutoff frequency may be
10-05-2019
19
determined as if y = a · x <2> + b. a、bはパラメータである。
[0077]
Alternatively, the cutoff frequency y may increase in accordance with an exponential function
with respect to wind noise intensity x. That is, the cutoff frequency may be determined as if y = a
· exp <x> + b. a、bはパラメータである。
[0078]
In FIG. 10A, y is defined such that both the cutoff frequency y and its slope dy / dx monotonously
increase. On the other hand, in FIG. 10 (b), although the cutoff frequency y monotonously
increases, its slope dy / dx does not monotonously increase, and then gradually increases and
then gradually decreases.
[0079]
Further, in FIG. 10A, the slope dy / dx of the cutoff frequency y is discontinuous near the
maximum value MAX, but in FIG. 10B, the slope of the cutoff frequency y also near the maximum
value MAX. dy / dx is continuous.
[0080]
The cutoff frequency characteristic of FIG. 10B may be determined using, for example, a
trigonometric function.
[0081]
According to the third embodiment, since the cutoff frequency fc gradually changes when the
wind noise intensity changes across the minimum value MIN, it is possible to reduce the sense of
discomfort in hearing.
[0082]
While the present invention has been described using specific terms based on the embodiments,
the embodiments merely show the principles and applications of the present invention, and the
embodiments are defined in the claims. Many variations and modifications of the arrangement
10-05-2019
20
can be made without departing from the concept of the present invention.
[0083]
DESCRIPTION OF SYMBOLS 1 ... Electronic apparatus, 2L ... Lch microphone, 2R ... Rch
microphone, 10 ... Audio signal processing circuit, 100 ... Wind-sound reduction circuit, 110 ...
High pass filter, 130 ... Control part, 200L ... Lch amplifier, 200R ... Rch amplifier, 202 ... auto
level controller, 204 L ... Lch A / D converter, 204 R ... Rch A / D converter, 206 ... digital signal
processing unit, 132 ... detection subtractor, 134 ... detection adder, 136 ... first low pass filter,
138 ... first 2 low-pass filter, 140: first smoothing circuit 142: second smoothing circuit 144:
detection unit 146: cutoff frequency setting unit
10-05-2019
21
Документ
Категория
Без категории
Просмотров
0
Размер файла
32 Кб
Теги
jp2014060524
1/--страниц
Пожаловаться на содержимое документа