close

Вход

Забыли?

вход по аккаунту

?

JP2009005274

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 JP2009005274
The present invention provides an acoustic device and a howling suppression method applied to
an electroacoustic transducing / reproducing apparatus in which signal oscillation hardly occurs
even if the amplification amount g (dB) of the signal is larger than the attenuation amount L (dB)
of a feedback loop. . SOLUTION: An acoustic input unit 1 which converts an acoustic signal into
an electric signal, an AD conversion unit 2 which converts an input signal into a digital signal, a
signal processing unit 3 which processes the digital signal, and converts the digital signal into an
analog signal , A signal amplification unit 5 for amplifying an analog signal, and an acoustic
output unit 6 for outputting the amplified signal as an acoustic signal. The signal processing unit
3 includes a comb filter unit 7 and frequency conversion. A sound device characterized in that it
includes a unit 8; [Selected figure] Figure 1
Acoustic device and method for suppressing howling
[0001]
The present invention relates to an acoustic device and a howling suppression method, and more
particularly to an acoustic device applied to an electroacoustic transducing / reproducing device
that amplifies an acoustic input signal such as a hearing aid, a loudspeaker, or a karaoke device
and outputs the signal from a speaker or earphone. It relates to a howling suppression method.
[0002]
In recent years, with the progress of miniaturization and high output of hearing aids, a
microphone unit for inputting sound and an earphone unit for outputting sound are arranged at
08-05-2019
1
a distance of several tens of millimeters.
In the case of a hearing aid for severe hearing loss people, the output level from the earphone
section may exceed 140 dB SPL (sound pressure level). Also in loudspeakers, high-power
speakers may be placed close to the microphone.
[0003]
However, in these devices in which the speaker or the earphone is in proximity to the
microphone, an acoustic feedback path (acoustic coupling) is provided, so that a part of the
output signal from the speaker or the earphone returns to the microphone. At this time, the
acoustic output signal is attenuated before reaching the microphone, but if the amplification
amount g (dB) of the device is larger than the attenuation amount L (dB), the feedback is
repeated and the signal level is infinite. Is amplified. That is, this phenomenon called howling not
only causes unpleasant sounds, but also causes problems such as overuse of speakers and
earphones and unnecessary power consumption.
[0004]
In order to suppress this howling, it is automatically estimated whether the howling is occurring
and the frequency at which the howling is occurring, and the component of the frequency is
canceled by the signal of the opposite phase, or near that frequency Methods for selectively
reducing the amplification factor (Patent Document 1, Patent Document 2), methods for
modulating the phase of the signal, methods for changing the delay time, and the like have been
proposed (Patent Document 3, Patent Document 4, Patents Document 5, Patent Document 6,
Patent Document 7). Also, a method of reducing acoustic coupling by providing a frequency
conversion unit between an input unit and an output unit (Patent Document 8). Furthermore, a
technique has been proposed in which oscillation in all bands is suppressed by alternately
applying two comb filters (Patent Document 9). With regard to hearing aids, various materials
and molding techniques for ear molds have been proposed to prevent so-called sound leakage, in
which the output from the earphone unit leaks out of the ear canal. It has also been proposed
that the power of the hearing aid can be turned on only when the earmold is properly attached to
the outer ear. Patent Document 1: Japanese Patent Application Laid-Open No. 6-22397 Patent
Document 2: European Patent Publication No. 0 467 499 A2 Japanese Patent Application LaidOpen No. 2000-341707 Japanese Patent Application Laid-Open No. 3-263999 Japanese Patent
Application Laid-Open No. 9-168195 US Patent No. 5621802 -43512 gazette
08-05-2019
2
[0005]
However, it is extremely difficult to reliably estimate the presence or absence of howling and the
howling occurrence frequency automatically, and the method of selectively reducing the
amplification factor of a specific frequency is effective when oscillation occurs in a limited
frequency range. Although there is a problem, there is a problem that it is difficult to obtain an
effect when oscillation occurs in a wide range. Although it is possible to reduce the acoustic
coupling by subjecting the signal to frequency conversion, it is difficult to obtain the effect when
the acoustic coupling occurs in a wide frequency range. Although the method of alternately
applying two comb filters has the effect of suppressing oscillation in a wide range, the frequency
characteristics of the filter become time-varying, so sideband waves occur in the stop range and
it is difficult to obtain a great effect.
[0006]
In addition, there are limits to the prevention of the sound leakage of the hearing aid, and it is
difficult to obtain the expected effect particularly in sweating, chewing and exercising. In order to
eliminate sound leakage, it is necessary to completely occlude the ear canal, which causes
psychological stress such as a sense of obstruction or pressure. Furthermore, in the case of a
growing child, there is a problem that the cost increases because the ear mold is replaced many
times as the child grows.
[0007]
An object of the present invention is, in view of the above problems, to provide an acoustic device
and a howling suppression method applied to an electroacoustic transducing and reproducing
apparatus in which signal oscillation is less likely to occur even if the amplification amount of the
signal is larger than the attenuation amount of the feedback loop. It is to do.
[0008]
The present invention adopts the following means in order to solve the above problems.
08-05-2019
3
The first means converts an audio signal into an electrical signal, an AD converter which converts
an input signal into a digital signal, a signal processor which processes the digital signal, and
converts the digital signal into an analog signal. It comprises: a DA conversion unit, a signal
amplification unit for amplifying an analog signal, and an audio output unit for outputting the
amplified signal as an acoustic signal, wherein the signal processing unit includes a comb filter
unit and a frequency conversion unit. It is an acoustic device to be characterized. A second means
is the acoustic device according to the first means, wherein the frequency converter moves the
frequency of the signal passed through the comb filter by an arbitrary amount. A third means is
an acoustic device according to the first means, wherein a signal whose frequency is moved by
the frequency conversion unit by an arbitrary amount is allowed to pass through the comb filter
unit. According to a fourth means, in any one of the first means to the third means, signal
processing is performed in the comb filter unit and the frequency conversion unit only for the
frequency band where howling tends to occur. It is an acoustic device to be characterized. A fifth
means is any one of the first means to the fourth means, wherein the comb filter has a pass band
and a stop band, and frequency components in the stop band are attenuated or silenced. A
frequency component in the pass band is amplified and output. A sixth means is characterized in
that in any one of the first means to the fifth means, the signal processing in the comb filter
section and the frequency conversion section is performed constantly or only when necessary. It
is an acoustic device. A seventh means according to any one of the first means to the sixth means
is a width of the pass band and the stop band of the comb filter section according to the presence
or absence or strength of howling, and / or According to another aspect of the present invention,
there is provided an acoustic device comprising: an amount of movement of a frequency of a
frequency conversion unit. The eighth means inputs the acoustic signal to the acoustic input unit,
processes and amplifies the signal output from the acoustic input unit, and outputs the acoustic
signal from the acoustic output unit. A step of passing the signal output from the input portion
through a comb filter portion having a pass band and a stop band, a step of causing the
frequency converter to shift the signal having passed the pass band, and the frequency shifted
signal Outputting the sound signal from the sound output unit as a sound signal, and when a part
of the sound signal output from the sound output unit is fed back to the sound input unit, the
signal returned from the sound output unit is blocked by the comb filter unit. And a step of
inputting the information.
A ninth means inputs an acoustic signal to an acoustic input unit, processes and amplifies the
signal output from the acoustic input unit, and outputs the acoustic signal from the acoustic
output unit. A step of causing the frequency converter to shift the signal output from the input
unit in the frequency converter; a step of passing the signal having the frequency shifted through
the comb filter portion having the passband and the stopband; A step of outputting as a sound
signal from the sound output unit, and a step of causing the frequency converter to shift the
frequency of the returned signal when a part of the sound signal output from the sound output
portion is returned to the sound input unit And the step of inputting the feedback and frequency-
08-05-2019
4
shifted signal into the stop band of the comb filter section.
[0009]
According to the present invention, only the frequency component passing through the pass
band of the comb filter unit is amplified in the signal processing unit of the acoustic device, and
this frequency component is subjected to the frequency shifting process in the frequency
conversion unit. Even if the output from is fed back to the input section, the feedback signal
enters the stop band of the comb filter, so that the feedback signal is not amplified again, the
continuation of oscillation is suppressed, and howling is suppressed. it can. Further, according to
the present invention, the comb filter process and the frequency conversion process are applied
to the external signal arriving at the input unit of the audio device, so that the sound quality
changes, but the width of the stop band of the comb filter unit If the frequency conversion
amount of the frequency conversion unit is sufficiently small and the frequency conversion
amount is approximately several percent, howling suppression can be effectively performed
without causing serious sound quality deterioration. Further, according to the present invention,
since it is not necessary to estimate the oscillation frequency or obtain the transfer characteristic
of the feedback path, howling suppression can be realized with a relatively simple circuit
configuration.
[0010]
Embodiments of the present invention will be described with reference to FIGS. 1 to 3. FIG. 1 is a
view showing the configuration of an acoustic device according to the invention of the present
embodiment. As shown in the figure, an acoustic signal is converted into an analog electrical
signal at an acoustic input unit 1. The analog electrical signal is converted to a digital signal by
the AD converter 2 and sent to the signal processing unit 3. The signal processing unit 3 includes
an alternating comb filter unit 7 and a frequency conversion unit 8. When a signal is input to the
signal processing unit 3, only frequency components passing through the passband of the comb
filter 7 are amplified, and frequency components passing through the stopband are attenuated.
The frequency converter 8 performs an arbitrary amount of frequency shift on the signal output
from the comb filter unit 7. The digital signal output from the signal processing unit 3 is
converted to an analog signal by the DA converter 4, amplified to a desired amplification amount
by the signal amplification unit 5 which is an analog amplifier, and the amplified analog signal is
used as a speaker or earphone Is converted into an acoustic signal by the acoustic output unit 6
composed of an electroacoustic transducer and output. When using a digital amplifier that
amplifies the digital signal output from the signal processing unit 3, the order of the DA
08-05-2019
5
converter 4 and the signal amplification unit 5 is switched.
[0011]
In this acoustic device, a desired spectrum envelope can be obtained by independently
controlling the amplification amount in the pass band of the comb filter unit 7 for each
frequency. In order to ensure stable operation, a comb filter by an FIR filter is desired, but other
means may be used if a comb characteristic can be realized. Normally, oscillation occurs when
the amplified acoustic signal is fed back to the acoustic input unit 1. However, oscillation can be
suppressed by moving the frequency of the feedback signal so that it falls within the stop band of
the comb filter unit 7. . That is, the frequency converter 8 performs frequency shift processing on
the signal output from the comb filter 7. An arbitrary amount of frequency shift processing can
be realized by using a Hilbert transformer, but other means may be used. In addition, although
the order of the comb filter unit 7 and the frequency conversion unit 8 is as shown in FIG. 1, it is
not essential to configure in this way, and the frequency is moved before passing through the
comb filter Also good.
[0012]
FIG. 2 is a diagram showing the output characteristics of the comb filter unit 7 and the frequency
conversion unit 8 in the signal processing unit 3. FIG. A is an output characteristic of the comb
filter unit 7 and FIG. B is a frequency shift by the frequency conversion unit 8. It shows the
output characteristics after. When the input signal passes through the comb filter unit 7, the
component passing through the passband is amplified by an amplification amount g (dB), and the
component passing through the stopband is attenuated by an attenuation amount l (dB).
Although the amplification amount g (dB) in the passband is constant in this figure, it may be set
independently for each frequency.
[0013]
As shown in FIG. 2, the relationship between the passband width fp and the stopband width fe in
the comb filter unit 7 and the frequency shift Δf in the frequency converter 8 will be described.
In the case of 1 / n of the range fp, that is, fp = n × Δf, the feedback signal will necessarily enter
the stop before n times of feedback. For example, if n is 1, the feedback signal does not pass
through the passband twice in a row. However, it is necessary that fe Δ Δf, that is, the frequency
08-05-2019
6
shift amount Δf be narrower than the stop band width fe. If the amount of frequency shift Δf is
larger than the stop band width fe, the effect is reduced because the feedback signal exceeds the
stop band and enters the next pass band. Further, in order to reduce the change in sound quality
as much as possible, it is desirable to reduce both the stop band width fe and the frequency shift
amount Δf.
[0014]
The principle of howling suppression in the present invention will be described with reference to
FIG. In the figure, I is the spectrum (solid line) of the initial input signal (A to E) input to the comb
filter unit 7 and II is the comb filter characteristic (grey is the passband) of the comb filter unit 7
and the comb Spectrum of the signal amplified after passing through the filter section 7 (solid
line), III is the spectrum of the signal after the frequency-shifted section of the amplified signal
after passing through the comb filter section 7 (solid line And IV are the spectrum of the signal
obtained by combining the feedback signal (solid line) input to the comb filter unit 7 and the
input signal (dashed line) newly input, and V is the new feedback signal (solid line) The spectrum
of the signal obtained by combining the input signal (broken line) after passing through the comb
filter unit 7, VI indicates the spectrum of the signal after the spectrum V has been frequencyshifted by the frequency conversion unit 8.
[0015]
That is, the initial input signal includes frequency components A to E (I). Among these, the
frequency components A, C and E are amplified through the pass band of the comb filter unit 7,
but the frequency components B and D are muted in the stop band (II). The frequency
components A ', C' and E 'thus amplified are subjected to frequency shift processing by the
frequency converter 8 and output (III). The feedback signals A ′ ′, C ′ ′ and E ′ ′ of the
output signal attenuated through the acoustic feedback path are input to the comb filter unit 7
together with the input signal newly arriving from the outside (IV). It can be seen that the
component of the signal input to the comb filter unit 7 that enters the stop band is silenced, so
that the feedback signal (solid line) is removed.
[0016]
According to the above principle, it is possible to prevent the feedback signal from being
08-05-2019
7
repeatedly amplified and to suppress oscillation. In FIG. 3, although the stop band width, the pass
band width, and the frequency shift amount are equal, they do not have to be the same. The filter
characteristics of the comb filter unit 7 and the frequency shift amount of the frequency
conversion unit 8 may be constant or may be changed according to the situation. In particular, if
the amount of frequency shift is large, the impression of voice etc. will be changed, so a method
of keeping the value small usually and increasing the amount of frequency shift so that the effect
appears early only when the level of the output signal becomes large It may be adopted.
[0017]
Although patent document 2 describes applying the notch filter which reduces the output of a
specific frequency range, changing the center frequency of a notch (stop region), the frequency
range which can be processed at once is one with a notch filter. It is limited. When oscillation
occurs in a wide band, a wide notch is required, but if a wide notch is used, desired frequency
characteristics can not be realized. On the other hand, in the present invention, since all
components can be made to pass through the stop band during feedback several times, an effect
is expected also for oscillation of a wide frequency band.
[0018]
Further, as shown in Patent Document 8, by performing frequency conversion on a signal,
acoustic coupling can be reduced. However, if the oscillation frequency is limited to a narrow
band, the oscillation can be suppressed with a slight frequency conversion, but if oscillation
occurs over a wide band, the oscillation can be suppressed unless the frequency is largely
converted. Can not. Substantial frequency conversion will significantly change the sound quality.
On the other hand, in the present invention, it is possible to pass all components in the stop band
by converting the frequency of the signal slightly by using both the frequency conversion
process and the comb filter, and it is possible to oscillate in a wide band. Is also valid.
[0019]
Further, Patent Document 9 proposes a howling suppression technique in which two comb filters
having complementary characteristics are alternately applied. However, when switching between
two comb filters, the frequency characteristics become time-variant, and it is necessary to apply
window processing such as a Hanning window to the filter. As a result, sideband waves are
08-05-2019
8
generated by the window processing in the stop area, and the effect is lost. On the other hand, in
the present invention, since the frequency characteristic of the comb filter may be fixed, the
influence of sideband waves can be reduced.
[0020]
Note that the acoustic device shown in FIG. 1 is merely an example, and the comb filter unit 7
and the frequency conversion unit 8 need not be in the order shown in the figure, and various
arrangements are possible. Further, in the present invention, since the frequency components are
removed by the comb filter and the remaining components also change in frequency, the change
in sound quality is inevitable. Therefore, it may be designed to operate only when the level of the
input or output signal is increased, or to operate only when oscillation is detected. For that
purpose, a level detector or an oscillation detector not shown may be added.
[0021]
Usually, howling does not occur at low frequencies such as 100 Hz and 200 Hz in hearing aids.
As described above, when the frequency band in which howling occurs is known in advance,
unnecessary sound quality change may be reduced by performing signal processing on howling
prevention only on the frequency where howling tends to occur.
[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]
It is a figure showing composition of an acoustic device concerning the invention of one
embodiment. FIG. 7 is a diagram showing output characteristics of a comb filter unit 7 and a
frequency conversion unit 8 in the signal processing unit 3 shown in FIG. 1. It is a figure for
demonstrating the principle of howling suppression in this invention.
08-05-2019
9
Explanation of sign
[0024]
DESCRIPTION OF SYMBOLS 1 sound input part 2 AD converter 3 signal processing part 4 DA
converter 5 signal amplification part 6 sound output part 7 comb-type filter part 8 frequency
conversion part
08-05-2019
10
Документ
Категория
Без категории
Просмотров
0
Размер файла
20 Кб
Теги
jp2009005274
1/--страниц
Пожаловаться на содержимое документа