close

Вход

Забыли?

вход по аккаунту

?

JP2009253525

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 JP2009253525
An object of the present invention is to change the sensitivity in the other direction while keeping
the sensitivity in the target direction high by changing the position of the microphone with time.
According to the present invention, there is provided a microphone array unit comprising a
plurality of microphones and amplifiers arranged in a space, and a data conversion unit for
converting multi-channel analog electrical signals input from the microphone array unit into
digital signals; The signal processing unit controls the delay time of each channel in accordance
with the target direction, and the power unit for moving the microphone or the microphone array
unit. The position of the microphone or microphone array changes with time. [Selected figure]
Figure 10
Microphone signal processing apparatus and processing method
[0001]
The present invention relates to information communication equipment and signal processing
technology, and to a microphone signal processing apparatus and processing method for
realizing a microphone array having excellent spatial sensitivity characteristics in a wide
frequency range.
[0002]
A microphone array that arranges a plurality of microphones in space and selectively resynthesizes a signal coming from a specific direction or position by controlling the delay time of
the signal of each channel is suitable for noise suppression and sound source separation. It is
04-05-2019
1
utilized as a method (patent documents 1 and 2).
Moreover, utilization is anticipated for sound source search (patent documents 3, 4).
[0003]
Conventional microphone arrays have the disadvantage that they are effective for signals near a
specific frequency but have less effect in other frequency bands, or the effects are uneven. If
stable effects can be obtained in a wide frequency range, application to wider fields can be
expected.
[0004]
The principle of the microphone array conventionally used will be described with reference to
FIG. 1, FIG. 2, and FIG. FIG. 1 is a diagram showing that the relative phase affects the amplitude
of the combined wave when combining two sound waves of the same frequency. In the figure, A,
B, C, and D are waveforms of the same frequency, and the maximum amplitude is 1, respectively.
Although A and B have the same phase (a phase difference of 0), A and C have a phase difference
of 2π / 3, and A and D have a phase difference of π. The maximum amplitude of the waveform
A + B that combines (adds) A and B with a phase difference of 0 is 2, but when the phase
difference λ is not 0, ie, 0 <λ <2π, the maximum amplitude of the combined wave is It is 0 or
more and less than 2. In particular, as shown by A + D, when the phase difference is π, the
amplitude of the combined wave is zero. As described above, when a plurality of identical signals
are added, the amplification effect by the addition is maximized if there is no phase difference. A
microphone array is a technology that provides maximum sensitivity in any direction or position
by exploiting such properties of sound waves.
[0005]
FIG. 2 is for explaining the operating principle of a microphone array composed of two
microphones. Assuming that the distance between the two microphones is r (m) and the line
connecting the microphones is the axis of the microphone array, the path from the direction of
the angle α to the axis until each microphone reaches is shown in the diagram. As shown in,
there is a distance difference of r cos α. The time difference generated by this difference in
04-05-2019
2
distance is the difference in distance r cos α divided by the speed of sound c (m / s). That is, the
signal arriving from the α direction reaches the microphone 2 with a delay of r cos α / c (s)
after reaching the microphone 1 (left figure). Therefore, in order to correct this time difference, a
delay of r cos α / c (s) is inserted in the path of channel 1 as shown on the right. At this time, the
time difference dα of the sound coming from the α direction in the outputs of both channels is:
dα = (r cos α / c) − (r cos α / c) = 0 (s). Since the time difference is 0 (s), signals coming from
the α direction are in phase between the channels, and when the outputs of channel 1 and
channel 2 are combined, they are efficiently amplified as shown by A + B in FIG.
[0006]
FIG. 3 is for explaining the effect of the microphone array for sounds coming from different
directions. The left figure is the same as the right figure of FIG. In this figure, since the delay of r
cos α / c (s) is inserted in the path of channel 1, the phases of the signals arriving from the α
direction are aligned between the channels (phase difference 0). At this time, assuming that the
inter-channel time difference relating to the sound coming from the β direction is dβ, dβ is
inserted into the channel 1 as the time difference r cosβ / c (s) physically generated by the
distance difference r cosβ from the sound source to both microphones. The delay r cos α / c (s)
is the difference, that is, dβ = r cos (α-β) / c (s). Due to this time difference, inter-channel phase
differences occur in sounds coming from the β direction. Therefore, even if the outputs of both
channels are combined, they are not amplified efficiently as in A + B of FIG. As a result, the sound
coming from the α direction is relatively emphasized, and directivity occurs. Note that the
direction in which sound is to be enhanced using a microphone array is referred to as a target. It
is possible to target any direction by changing the delay time inserted in channel 1. In an actual
multi-channel microphone array, delay times are set independently for all channels.
[0007]
By the above-described method, the sound coming from the β direction is relatively suppressed
as compared to the sound coming from the α direction. This is because the phase difference
caused by the inter-channel time difference of the sound coming from the β direction is not
zero. However, when the sound coming from the β direction is a periodic signal and the period
is 1 / n (n is an integer) of dβ, the phase difference caused by the inter-channel time difference
is 2nπ. Since a phase difference that is an integral multiple of 2π (360 °) is equivalent to a
phase difference of 0, the sound coming from the β direction is also emphasized like the sound
in the α direction. That is, directions other than the target direction are emphasized. Conversely,
when the time difference between the channels causes a phase difference of (2n + 1) π, the
04-05-2019
3
phases become opposite between the channels, and when the outputs of both channels are
combined, the amplitude becomes 0 as shown by A + D in FIG. As described above, in the space
characteristic of the conventional microphone array, a plurality of peaks or valleys may be
generated in sensitivity other than the target direction. Japanese Patent Application Publication
No. 2004-334218 Japanese Patent Application Publication No. 09-251299 Japanese Patent
Application Publication 2007-180942 Japanese Patent Application Publication 2007-180953
[0008]
The problems with the sensitivity characteristics of the conventional microphone array will be
described with reference to FIGS. 4 and 5. FIG. 4 shows sensitivity characteristics of signals of
various frequencies in a microphone array consisting of two microphones. The microphone array
has microphones 1 and 2 arranged at an interval of 34 cm as shown in the upper center view,
and 100 Hz, 200 Hz, 400 when targeting in the direction of the arrow (0 ° direction) in the
figure. The lower left figure shows the result of the sensitivity characteristic for the Hz signal,
and the lower right figure the result of the sensitivity characteristic for the 800 Hz, 1 kHz, and
1.6 kHz signals. Assuming that the sensitivity in the target direction is 0 dB, it can be seen that
the sensitivity decreases by about 2 dB at 100 Hz and by about 10 dB at 200 Hz in the 180 °
direction. When the frequency of the signal is 400 Hz, valleys of sensitivity occur near 115 °
and 255 °. At higher frequencies, there are peaks and valleys of multiple sensitivities, and it can
be seen that the sensitivities change rapidly depending on the direction. In this example,
excellent directivity is realized in the band around 200 Hz, but stable effects can not be expected
in other bands, particularly in the high band.
[0009]
FIG. 5 shows the results of the sensitivity characteristics of a microphone array using 17
microphones. As shown in the left figure, 17 microphones were arranged in a line, and the sound
from the arrow direction (0 ° direction) was set to be emphasized. At this time, it is the right
figure which showed the sensitivity characteristic to the signal of 100 Hz, 400 Hz, and 1 kHz. In
the case of 1 kHz, the sensitivity is maximum (0 dB) in the directions of 90 °, 180 °, and 270
° in addition to the direction of 0 °. As described above, in the conventional microphone array,
the sensitivity largely fluctuates depending on the frequency of the signal in directions other
than the target direction, and there is a problem that the sensitivity changes sharply in a high
frequency region with a short period, and stable directivity is difficult to obtain. .
04-05-2019
4
[0010]
When the conventional microphone array targets a specific direction, for example, the α
direction, inter-channel time differences occur in sounds in other directions, for example, the β
direction. Since this time difference is constant as long as the characteristics of the microphone
array are not changed, the inter-channel time difference becomes an integral multiple of the
signal period at a specific frequency. In the present invention, it is an object of the present
invention to change the sensitivity in the other direction while keeping the sensitivity in the
target direction high by changing the position of the microphone with time in order to avoid this
disadvantage.
[0011]
A microphone signal processing apparatus according to the present invention includes a
microphone array unit including a plurality of microphones and amplifiers disposed in space, and
a data conversion unit for converting multi-channel analog electrical signals input from the
microphone array unit into digital signals. A signal processing unit for controlling a delay time of
each channel in accordance with a target direction, and a power unit for moving the microphone
or the microphone array unit, the position of the microphone or the microphone array unit
Changes with time. In addition, the power unit rotates the microphone or the microphone array
at a constant angular velocity, and controls the delay time of each channel based on the rotation
angle information of the rotation. With this constant angular velocity rotational motion, the angle
θt at an arbitrary time t when making one rotation at T (s) is θt = (2π × t) / T, and the speed
of sound is c (m / s), the target Assuming that the direction angle is α, a time-varying delay of r
cos (θt−α) / c (s) is inserted into each channel as a reference channel.
[0012]
In the microphone signal processing method of the present invention, the inter-channel time
difference of the signal arriving from the target direction or position is always zero with respect
to the microphone or microphone array whose position changes with time. By changing the delay
time, the sensitivity in the relevant direction is kept constant. At this time, since the inter-channel
time difference of the sound coming from other than the direction changes with time, the
sensitivity is also time-varying, and averaging the peaks and valleys for a long time averages the
peaks and valleys of the sensitivity. Smooth sensitivity characteristics are realized.
04-05-2019
5
[0013]
According to the present invention, by changing the sensitivity in the other direction while
keeping the sensitivity in the target direction high, the inter-channel time difference becomes
time-varying in all directions other than the target direction, and the sensitivity characteristics
are steady peaks. , Can prevent the generation of valleys. Since a plurality of peaks and valleys
are less likely to occur in the sensitivity of the microphone array, it is possible to realize a stable
effect (directivity) in a wide frequency range.
[0014]
Hereinafter, the present invention will be described based on examples. Here, an example in
which the microphone array rotates with time will be described. The rotating microphone array
is referred to as a rotating microphone array, and the conventional microphone array is a fixed
microphone array.
[0015]
FIG. 6 shows a state in which sound comes from the α direction to a rotary microphone array
consisting of two microphones. In this microphone array, it is assumed that the microphone 2
rotates counterclockwise around the microphone 1. The left figure shows the state at time 0, and
the right figure shows the state after time Δt (s). The time difference for the sound arriving from
the α direction to reach both microphones at time 0 is r cos (θ 0 −α), and at time Δt, this is r
cos (θ t −α). In order to target the α direction, it is necessary to always set the inter-channel
time difference to 0 for the sound from the α direction. Therefore, the delay inserted into
channel 1 must change with time. Assuming that the angular velocity ω of the rotary
microphone array is constant and makes one rotation at T (s), the angle θt at an arbitrary time t
is θt = (2π × t) / T. Therefore, if a time-varying delay of r cos (θt−α) / c (s) is inserted in the
channel 1, it is possible to make the inter-channel time difference 0 for the sound in the α
direction at all times. At this time, the sensitivity in the α direction is kept high.
[0016]
04-05-2019
6
FIG. 7 shows that the sound comes from the β direction to the same rotary microphone array. A
delay of r cos (θt−α) / c (s) is added to channel 1 to target the α direction. The time difference
dβ between the sound in the β direction at the output of both channels is the time difference r
cos (θt−β) / c (s) due to the distance difference from the sound source to both microphones
and the delay r cos (θt−α) added to channel 1 ) / C (s), that is, dβ = r {cos (θt−α) −cos
(θt−β)} / c. This value is always 0 in the case of α = β, but is a time-varying value which
fluctuates with time in other cases. Therefore, if α ≠ β, the amount of amplification obtained by
combining the outputs of both channels is also time-varying, and even if the maximum value or
the minimum value may be instantaneously obtained, the target direction when averaged over
time T Smaller than the signal of In this way, peaks and valleys of sensitivity can be made less
likely to occur in directions other than the target.
[0017]
FIG. 8 shows an example of sensitivity characteristics of the rotary microphone array. As shown
in the left figure, the microphone array, in which 17 microphones were arranged in a line at 34
cm intervals, was rotated counterclockwise around the center microphone and targeted at the
arrow direction (0 ° direction) in the figure. The sensitivity characteristics obtained in the case
are shown in the right figure. Since the distance between adjacent microphones is 34 cm, the
radius r of the microphone array is 34 × 8 = 272 cm, and the delay amount inserted in the x-th
channel is (34 × x−272) × cos (θt− α) / c, but since the target direction (α) is 0 °, (34 ×
x−272) × cos (θt) / c.
[0018]
100
The sensitivity characteristic of Hz and 400 Hz is a smooth ellipse. Assuming that the sensitivity
in the target direction (0 ° direction) is 0 dB, the sensitivity in the opposite direction (180 °
direction) is about -10 dB at 100 Hz. At 400 Hz, it is about -16 dB. At 1 kHz, there is a peak of
sensitivity in the directions of 60 ° and 300 °, but it can be seen that the fluctuation is smaller
when compared with the sensitivity characteristic of the fixed microphone array (FIG. 5).
[0019]
FIG. 9 shows an example of the microphone arrangement of the rotary microphone array. In the
04-05-2019
7
rotary microphone array, the sensitivity in the target direction is always kept at the maximum,
and the sensitivity characteristics in the directions other than the target change with time, and
the stable sensitivity characteristics with less peaks and valleys as described above You can get
In the case of the microphone array shown in FIGS. 6 and 7, one rotation of the microphone
array provides an average of sensitivity characteristics. However, depending on the arrangement
of the microphones, it is not necessary to make one rotation. If the microphones are arranged as
shown in the left view of FIG. 9, an effect can be obtained by a 90.degree. Rotation, and if
arranged as shown in the right view, the microphones may be rotated by 60.degree.
[0020]
FIG. 10 is a diagram illustrating the configuration of the present invention. The analog signal
recorded and amplified by the microphone array unit 1 including a plurality of microphones and
amplifiers is converted into a digital signal by the data conversion unit 2, and is sent to the signal
processing unit 3. The signal processing unit 3 controls the delay time of each channel based on
the rotation angle information so that the inter-channel time difference of sounds coming from
any target direction is always 0, and the signal of all or some channels is It is synthesized. Power
for changing the position of each microphone in the microphone array unit 1 or the microphone
array unit 1 is supplied from a power unit 4 using a motor or the like.
[0021]
In the above description, a rotary microphone array is proposed as one example, but it is
important that the position of the microphone array or the microphone changes with time, and it
is not limited to the rotational movement. Various embodiments are possible, such as those in
which the microphones move linearly and those in which the pendulum moves.
[0022]
The basic embodiments of the present invention have been described above, but the present
invention is not limited to these embodiments. Appropriate modifications can be made without
departing from the scope of the present invention.
[0023]
04-05-2019
8
Explanatory diagram of amplification effect by synthesis of sine wave signal Diagram explaining
the principle of microphone array Diagram explaining the principle that directivity is caused by
the microphone array Figure 2 Calculation result of sensitivity characteristic in channel fixed
microphone array Figure 17 Channel fixed system A figure showing the calculation results of
sensitivity characteristics of the microphone array A figure showing the sound coming from the
target direction to the rotary microphone array A figure showing the sound from the non-target
direction coming to the rotary microphone array The figure which shows the sensitivity
characteristic calculation result of the array The example of microphone arrangement in the
rotary type microphone array The figure which illustrates the constitution of this invention
Explanation of sign
[0024]
1 microphone array unit 2 data conversion unit 3 signal processing unit 4 power unit
04-05-2019
9
Документ
Категория
Без категории
Просмотров
0
Размер файла
19 Кб
Теги
jp2009253525
1/--страниц
Пожаловаться на содержимое документа