`
`(12) Japanese Unexamined Patent
`Application Publication (A)
`
`
`(11) Japanese Unexamined Patent
`Application Publication Number
`Japanese Unexamined Patent
`Application Publication
`H11-18186
`(43) Publication date: January 22, 1999
`
`
`(51) Int. Cl.6
`
`Identification codes
`
` FI
`
`
`
`
`Request for examination: Not yet requested Number of claims: 6 OL (Total of 10 pages)
`
`(21) Application number
`
`
`Japanese Patent Application
`No. H9-167812
`
`(71) Applicant
`
`(22) Application date
`
`June 25, 1997
`
`(72) Inventor
`
`
`
`
`
`
`
`
`
`(72) Inventor
`
`(74) Agent
`
`000136848
`Primo Company Ltd.
`6-25-1 Mure, Mitaka-shi, Tokyo
`Kimihiro IKEDA
`℅ Primo Company Ltd.
`6-25-1 Mure, Mitaka-shi, Tokyo
`Yasuo MAEKAWA
`℅ Primo Company Ltd.
`6-25-1 Mure, Mitaka-shi, Tokyo
`Patent Attorney Shizuyo TAMAMURA
`
`
`FIG. 1
`
`
`(54) (TITLE OF THE INVENTION) PRIMARY PRESSURE GRADIENT MICROPHONE
`
`(57) (ABSTRACT)
`(PROBLEM) To provide a primary pressure gradient
`microphone with which sound from a target sound source
`in a specific direction can be acquired with good S/N in an
`environment in which noise is generated from every
`direction.
`(SOLUTION) First and second omnidirectional microphone
`units (M1, M2) are disposed apart from one another,
`wherein a difference between a delay signal of the output of
`the second omnidirectional microphone unit and the output
`of the first omnidirectional microphone unit is determined
`by a first adding means (3), and a difference between the
`output of the second omnidirectional microphone unit and a
`delay signal of the output of the first omnidirectional
`microphone unit is determined by a second adding means
`(4). The outputs of the first and second adding means have
`directional characteristics in opposite directions so that the
`directional characteristics of the one facing a target sound
`source are used to pick up a target sound and noise while
`the directional characteristics of the other are used to pick
`up noise, and the difference between the two is taken so as
`to cancel the noise component.
`
`
`
`
`ADAPTIVE
`FILTER
`
`DIGITAL
`ADDER
`
`
`
`Page 1 of 12
`
`GOOGLE EXHIBIT 1013
`
`
`
`Japanese Unexamined Patent Application Publication No. H11-18186
`(2)
`
`(SCOPE OF THE PATENT CLAIMS)
`(CLAIM 1) A primary pressure gradient microphone
`comprising: first and second omnidirectional microphone
`units disposed a prescribed distance apart from one another;
`a first delaying means for delaying an output of the first
`omnidirectional microphone unit; a second delaying means
`for delaying an output of the second omnidirectional
`microphone unit; a first adding means for outputting a
`difference between an output of the first omnidirectional
`microphone unit and an output of the second delaying
`means; and a second adding means for outputting a
`difference between an output of the second omnidirectional
`microphone unit and an output of the first delaying means.
`(CLAIM 2) A primary pressure gradient microphone
`comprising: first and second omnidirectional microphone
`units disposed a prescribed distance apart from one another;
`a
`third microphone unit disposed near
`the second
`microphone unit; a first delaying means for delaying an
`output of the first omnidirectional microphone unit; a
`second delaying means for delaying an output of the second
`omnidirectional microphone unit; a first adding means for
`outputting a difference between an output of the first
`omnidirectional microphone unit and an output of the
`second delaying means; and a second adding means for
`outputting a difference between an output of the third
`omnidirectional microphone unit and an output of the first
`delaying means.
`(CLAIM 3) A primary pressure gradient microphone
`comprising: first and second omnidirectional microphone
`units disposed a prescribed distance apart from one another;
`third and fourth omnidirectional microphone unit disposed
`a prescribed distance apart from one another in a cross
`direction with respect to disposing direction of the first and
`second omnidirectional microphone units; a delaying
`means for delaying an output of the second omnidirectional
`microphone unit; a first adding means for adding a
`difference between an output of the first omnidirectional
`microphone unit and an output of the delaying means; and a
`second adding means for outputting a difference between
`an output of the third omnidirectional microphone unit and
`an output of the fourth omnidirectional microphone unit.
`(CLAIM 4) A primary pressure gradient microphone
`comprising: first and second omnidirectional microphone
`units disposed a prescribed distance apart from one another;
`third and fourth omnidirectional microphone units disposed
`adjacent to and in identical sequence with the first and
`second omnidirectional microphone units; a second
`delaying means for delaying an output of the second
`omnidirectional microphone unit; a first delaying means for
`delaying an output of the third omnidirectional microphone
`unit; a first adding means for outputting a difference
`between an output of the first omnidirectional microphone
`unit and an output of the second delaying means; and a
`second adding means for outputting a difference between
`an output of the fourth omnidirectional microphone unit
`and an output of the first delaying means.
`(CLAIM 5) The primary pressure gradient microphone
`according to any one of Claims 1 to 4, further comprising: a
`first A/D conversion means for converting an output of the
`first adding means to a digital signal; a second A/D
`conversion means for converting an output of the second
`
`
`
`
`adding means to a digital signal; and a digital signal
`processing means
`for cancelling noise components
`contained in both an output of the first A/D conversion
`means and an output of the second A/D conversion means.
`(CLAIM 6) The primary pressure gradient microphone
`according to Claim 5, wherein the digital signal processing
`means comprises: an adaptive filter means for receiving an
`output of the second A/D conversion means; and a digital
`adding means for calculating a difference between an
`output of the adaptive filter means and an output of the first
`A/D conversion means.
`(DETAILED DESCRIPTION OF THE INVENTION)
`(0001)
`(TECHNICAL FIELD OF THE INVENTION) The present invention
`relates
`to a primary pressure gradient microphone
`exhibiting two types of unidirectional characteristics using
`a plurality of omnidirectional microphone units, which
`exhibits excellent noise component reducing capacity and
`contributes to the miniaturization of the microphone, and
`may be effectively applied to a primary pressure gradient
`microphone for automobile use, for example.
`(0002)
`(CONVENTIONAL TECHNOLOGY) Previous applications filed
`by the present applicant (Japanese Unexamined Utility
`Model Application Publication No. S62-147476 and
`Japanese Unexamined Utility Model Application
`Publication No. S64-52393) disclose a primary pressure
`gradient microphone that achieves unidirectionality using
`two omnidirectional microphone units of essentially the
`same performance. In principle, this primary pressure
`gradient microphone
`is essentially
`the same as
`the
`configuration illustrated in FIG. 2. Two omnidirectional
`microphone units M1 and M2 are disposed at a distance d
`apart from one another, wherein the output of the
`microphone unit M2 is delayed by τ2 (τ2: amount of delay
`expressed in terms of distance) by a delay circuit 2, and the
`difference between the delayed output ed2 and the output
`e1 of the microphone unit M1 is extracted.
`(0003) In a microphone configured in this way, when a
`sound comes from a direction at an angle θ with respect to
`the axial direction of the two microphone units M1 and M2,
`the output Vm of the microphone is given by the following
`formula:
`
`
`Here, k=2π/λ (λ: wavelength of sound), and the sound
`arriving at the microphone units is a sinusoidal sound wave
`(sound wave with a sinusoidal wave shape). From the
`above formula, appropriately setting τ2 to be smaller than d
`yields directional characteristics such as those illustrated in
`FIG. 3.
`(0004) To facilitate understanding here, the principle of the
`aforementioned primary pressure gradient microphone will
`be described qualitatively with reference to FIG. 6. A case
`in which τ2=d and θ=0 will be used as an example. An
`observation point corresponding to the output e1(a) of the
`microphone unit M1 with respect to a sound arriving from
`the Sa direction is Pe1(a), and an observation point
`corresponding
`to
`the delayed output ed2(a) on
`the
`microphone unit M2 side is apparently Ped2(a). An
`
`Page 2 of 12
`
`
`
`Japanese Unexamined Patent Application Publication No. H11-18186
`(3)
`
`observation point corresponding to the output e1(b) of the
`microphone unit M1 with respect to a sound arriving from
`the opposite direction Sb is Pe1(b), and an observation
`point corresponding to the delayed output ed2(b) on the
`microphone unit M2 side is apparently Ped2(b). As is clear
`from the drawing, because the observation points Pe1(b)
`and Ped2(b) are
`the same positions, e1(b)=ed2(b).
`Therefore, the output Vm of the microphone becomes:
`
`
`It can be easily understood that unidirectionality is
`achieved as a result.
`(0005)
`(PROBLEM TO BE SOLVED BY THE INVENTION) However, the
`primary pressure gradient microphone described above
`only has unidirectional characteristics with respect to the
`front surface side. Accordingly, in an environment in which
`sound (noise) is generated from all directions (four
`directions) (an environment in which the noise source
`cannot be fixed to a specific direction), the input of noise
`cannot be arrested by attempting to input only sound from a
`target sound source in a specific direction.
`(0006) Therefore, although not publicly known, the present
`inventors investigated a microphone configured to have
`two
`types of directionality by making
`two of
`the
`aforementioned primary pressure gradient microphones 10-
`1 and 10-2 face opposite directions as one another, as
`illustrated in FIG. 13. The first microphone 10-1 is made to
`face a specific target sound source so as to pick up the
`sound of the target sound source and ambient noise using
`the unidirectional characteristics
`thereof. The second
`microphone 10-2, which is made to face the opposite
`direction as the first microphone 10-1, picks up only noise
`using
`the unidirectional characteristics
`thereof. By
`determining the difference between the output of the first
`microphone 10-1 and the output of the second microphone
`10-2, it is possible to reduce noise arriving from all
`directions and to obtain the target sound.
`(0007) As a result of further investigating the technology
`described above, the present inventors ascertained the
`following. First, the present inventors ascertained that it is
`physically impossible to dispose the two primary pressure
`gradient microphones described above at the same position,
`and that directional characteristics such as those illustrated
`in FIG. 14 introduce logical error into the removal of noise
`components, making it difficult to expect high-precision
`noise
`cancellation. Second,
`the present
`inventors
`ascertained
`that
`in order
`to meet
`the demand for
`miniaturization, it is necessary to realize a primary pressure
`gradient microphone which has two types of directional
`characteristics using fewer
`than four omnidirectional
`microphone units.
`(0008) The present invention was conceived in light of the
`circumstances described above, and an object thereof is to
`provide a primary pressure gradient microphone that
`exhibits an excellent noise cancelling effect and with which
`two types of directionality can be achieved using four or
`fewer omnidirectional microphone units.
`(0009) Another object of the present invention is to provide
`a primary pressure gradient microphone with which a
`
`sound from a target sound source in a specific direction can
`be acquired with good S/N in an environment in which
`noise is generated from every direction.
`(0010) The aforementioned and other objects and novel
`features of the present invention will be clear from the
`descriptions of this specification and the attached drawings.
`(0011)
`(MEANS FOR SOLVING THE PROBLEM) As illustrated in FIG.
`1, the best primary pressure gradient microphone according
`to the present invention achieves two types of directional
`characteristics using two omnidirectional microphone units,
`and
`comprises:
`first
`and
`second omnidirectional
`microphone units (M1, M2) disposed a prescribed distance
`apart from one another; a first delaying means (1) for
`delaying an output of the first omnidirectional microphone
`unit; a second delaying means (2) for delaying an output of
`the second omnidirectional microphone unit; a first adding
`means (3) for outputting a difference between an output of
`the first omnidirectional microphone unit and an output of
`the second delaying means; and a second adding means (4)
`for outputting a difference between an output of the second
`omnidirectional microphone unit and an output of the first
`delaying means.
`(0012) When the first omnidirectional microphone unit
`(M1) is defined as the front surface of the microphone, a
`signal system for obtaining the output of the first adding
`means (3) is made by the two microphone units (M1, M2)
`to function as a main microphone (see FIGS. 2 and 3)
`having directional characteristics on the front surface side,
`and a signal system for obtaining the output of the second
`adding means (4) is made by the two microphone units
`(M1, M2) to function as a reference microphone having
`directional characteristics on the back surface side. In an
`environment in which noise is generated from every
`direction so that a noise source cannot be specified, the
`main microphone facing the target sound source picks up
`the target sound and noise, and the reference microphone
`picks up only noise. By determining the difference between
`the output signal component of the main microphone and
`the output signal component of the reference microphone,
`the noise components can be cancelled. At this time, the
`main microphone and
`the reference microphone are
`configured using the same microphone units (M1, M2), so
`both of the microphones can be considered to be disposed
`at physically the same position. In other words, the noise
`components picked up by both of the microphones can be
`considered to be essentially the same. It is therefore
`possible to acquire only the target sound component with
`high S/N.
`(0013) Essentially the same functions as those described
`above can be
`realized using
`three omnidirectional
`microphone units. That is, as illustrated in FIG. 8, a third
`omnidirectional microphone unit (M3) is disposed near the
`second omnidirectional microphone unit,
`the main
`microphone is as described above, and the reference
`microphone is configured to obtain the difference between
`the output of the first delaying means (1) and the output of
`the third microphone unit with the second adding means.
`As a result, a primary pressure gradient microphone having
`two types of directional characteristics can be realized with
`three microphone units. Although the positions of the main
`
`
`
`Page 3 of 12
`
`
`
`Japanese Unexamined Patent Application Publication No. H11-18186
`(4)
`
`microphone and the reference microphone cannot be
`considered to be exactly the same, the second microphone
`unit and the third microphone unit can be disposed near one
`another without any distance therebetween, so a good noise
`cancelling effect can be achieved.
`(0014) Essentially the same functions as those described
`above can be
`realized using
`four omnidirectional
`microphone units. That is, as illustrated in FIG. 9, third and
`fourth omnidirectional microphone units (M3, M4) are
`disposed a prescribed distance apart from one another in a
`cross direction with respect to the disposing direction of the
`first and second omnidirectional microphone units; the
`difference between the output of the delaying means for
`delaying
`the output of
`the second omnidirectional
`microphone unit and the output of the first omnidirectional
`microphone unit is determined by the first adding means;
`and the difference between the output of the third
`omnidirectional microphone unit and the output of the
`fourth omnidirectional microphone unit is determined by
`the second adding means. Although the number of
`microphone units is four, the main microphone and the
`reference microphone can be logically considered to be
`positioned in the center of the four microphone units
`disposed radially, which makes it possible to achieve an
`excellent noise cancelling effect.
`(0015) Further, when four omnidirectional microphone
`units are used, as illustrated in FIG. 12, the main
`microphone and the reference microphone configured from
`two omnidirectional microphone units are configured from
`different omnidirectional microphone units (M1 to M4).
`That is, the main microphone is configured using the first
`and second omnidirectional microphone units (M1, M2),
`and the reference microphone is configured using the third
`and fourth omnidirectional microphone units (M3, M4).
`The main microphone has directional characteristics on the
`front surface side, and the reference microphone has
`directional characteristics on the back surface side. At this
`time, the sequence of the third and fourth omnidirectional
`microphone units is identical to the sequence of the first
`and second omnidirectional microphone units, and the first
`and second omnidirectional microphone units are disposed
`adjacent to the third and fourth omnidirectional microphone
`units. Therefore, the difference in sound fields between
`noise picked up by the main microphone and noise picked
`up by the reference microphone is minimized, which makes
`it possible to acquire the target sound with relatively good
`S/N.
`(0016) Although the processing for cancelling noise using
`the outputs of the first and second adding means can also
`be performed with an analog technique, the configuration
`further comprises the following when a digital technique is
`used: a first A/D conversion means (5) for converting the
`output of the first adding means to a digital signal; a second
`A/D conversion means (6) for converting the output of the
`second adding means to a digital signal; and a digital signal
`processing means for cancelling the noise components
`contained in both the output of the first A/D conversion
`means and the output of the second A/D conversion means.
`(0017) The digital signal processing means comprises: an
`adaptive filter means (7) for receiving the output of the
`second A/D conversion means (6); and a digital adding
`means (8) for calculating the difference between the output
`
`of the adaptive filter means and the output of the first A/D
`conversion means.
`(0018) An example of the primary pressure gradient
`microphone according to the present invention is illustrated
`in FIG. 1. In the microphone illustrated in the drawing, two
`omnidirectional microphone units M1 and M2 having
`essentially the same performance are disposed at a distance
`d apart from one another, wherein the output e1 of the
`microphone unit M1 is delayed by τ1 (τ1: amount of delay
`expressed in terms of distance) by a delay circuit 1, and the
`output e2 of the microphone unit M2 is delayed by τ2 (τ2:
`amount of delay expressed in terms of distance) by a delay
`circuit 2. Also provided are an inversion adding circuit 3
`for calculating the difference between the output e1 of the
`microphone unit M1 and the output ed2 of the delay circuit
`2, and an inversion adding circuit 4 for calculating the
`difference between the output e2 of the microphone unit
`M2 and the output ed1 of the delay circuit 1. The inversion
`adding circuits 3 and 4 may be formed, for example, from
`operational amplifiers, feedback resistors, and
`input
`resistors, where the non-inverted inputs (+) are defined as
`e1 and e2, and the inverted inputs (-) are defined as ed2 and
`ed1.
`(0019) The microphone is housed in a casing, not
`illustrated, so that the microphone unit M1 is disposed on
`the front surface. Ceramic or electrolyte-type microphone
`units or the like may be used for the microphone units M1
`and M2.
`(0020) At this time, the microphone can be understood as
`two types of primary pressure gradient microphones
`including the main microphone of FIG. 2 having a signal
`path for obtaining a main signal Vm from the inversion
`adding circuit 3 and the reference microphone of FIG. 4
`having a signal path for obtaining a reference signal Vr
`from the inversion adding circuit 4.
`(0021) The main microphone has directional characteristics
`on the front surface side, and therefore has directional
`characteristics for picking up a target sound and noise from
`the front surface side. The main signal Vm is given is
`follows:
`
`
`(0022) The main microphone, which can be understood as
`illustrated in FIG. 2, has directional characteristics with
`respect to the front surface side and has the directional
`characteristics illustrated in FIG. 3. The fact that such
`directional characteristics are achieved can be understood
`from
`the content described qualitatively above with
`reference to FIG. 6.
`(0023) The
`reference microphone has directional
`characteristics on the back surface side, and therefore has
`directional characteristics for picking up noise from the
`back surface side without picking up the target sound and
`noise from the front surface side. The reference signal Vr is
`given as follows:
`
`
`Here, in the two formulas above, k=2π/λ (λ: wavelength of
`sound), and the characteristics of the two omnidirectional
`microphone units M1 and M2 are essentially the same. The
`
`
`
`Page 4 of 12
`
`
`
`Japanese Unexamined Patent Application Publication No. H11-18186
`(5)
`
`arriving sound wave is assumed to be a sinusoidal sound
`wave.
`(0024) The reference microphone illustrated in FIG. 4 has
`directional characteristics with respect to the back surface
`side, as illustrated in FIG. 5. The fact that such directional
`characteristics are achieved will be described qualitatively
`with reference to FIG. 7. That is, in order to facilitate
`understanding in FIG. 7, a case in which τ1=d and θ=0 will
`be used as an example. An observation point corresponding
`to the delayed output ed1(a) on the microphone unit M1
`side with respect to a sound arriving from the Sa direction
`is apparently Ped1(a), and an observation point
`corresponding to the output e2(a) of the microphone unit
`M2 is Pe2(a). An observation point corresponding to the
`delayed output ed1(b) of the microphone unit M1 with
`respect to a sound arriving from the opposite direction Sb is
`apparently Ped1(b), and an observation point corresponding
`to the output e2(b) of the microphone unit M2 is Pe2(b). As
`is clear from the drawing, because the observation points
`Ped1(a) and Pe2(a) are the same positions, ed1(a)=e2(a).
`Therefore, the output Vr of the reference microphone
`becomes:
`
`
`This makes it possible to achieve unidirectionality on the
`back surface side.
`(0025) As described above, a main microphone and a
`reference microphone with
`respectively
`different
`directional characteristics are realized by commonly using
`two omnidirectional microphone units M1 and M2, so this
`is equivalent to a case in which the main microphone and
`the reference microphone having different directional
`characteristics are disposed at physically the same position.
`Accordingly, the microphone illustrated in FIG. 1 picks up
`sounds of different components of the same sound field. On
`the other hand, as illustrated in FIG. 13, when the main
`microphone and
`the reference microphone are each
`configured using two different omnidirectional microphone
`units, both microphones are disposed at physically different
`positions, as illustrated in FIG. 14. Therefore, both
`microphones pick up sounds different components of
`different sound fields separated by a prescribed distance. In
`the configuration
`illustrated
`in FIG. 13,
`there are
`differences that cannot be ignored between the components
`of sounds picked up by the microphones 10-1 and 10-2
`disposed apart from one another in an environment in
`which noise is generated from every direction. In the case
`of FIG. 1, the main microphone and the reference
`microphone are in exactly the same sound field, so the
`noise components picked up by both of the microphones
`can be considered to be essentially the same, even in an
`environment in which noise is generated from every
`direction.
`(0026) FIG. 10 illustrates an example of a polar pattern
`expressing the directional characteristics achieved by
`combining the respective directional characteristics of the
`main microphone and the reference microphone. In this
`example, τ2=d/3 and τ1=d. As is clear from this polar
`pattern,
`the primary pressure gradient microphone
`
`
`consisting of the main microphone and the reference
`microphone
`achieves
`two
`types
`of
`directional
`characteristics on the front surface side and the back
`surface side with the two microphone units M1 and M2.
`(0027) The output Vm of the inversion adding circuit 3 is
`converted to a digital signal Dm by an A/D converter 5, and
`the output Vr of the inversion adding circuit 4 is converted
`to a digital signal Dr by an A/D converter 6. Main signal
`data Dr outputted from the A/D converter 6 is supplied to
`an adaptive filter 7 such as a transversal filter, and the
`difference between the filter output of the adaptive filter 7
`and the main signal data Dm outputted from the A/D
`converter 5 is calculated by a digital adder 8.
`(0028) The adaptive filter 7 is a filter that adaptively varies
`its characteristics with respect to changes in the input signal
`over time. That is, the filter characteristics are varied in real
`time by performing an output operation for obtaining filter
`output by multiplying the input signal Dr by a tap
`coefficient and adding the value in multiple tap units, and a
`tap coefficient updating operation for updating the tap
`coefficient based on an error signal obtained as the
`difference between the filter output and a signal from the
`target system to return the filter output, and successively
`rewriting the tap coefficient using the updating operation.
`The error signal is defined as the output of the digital adder
`8 at a training sequence timing for determining the filter
`response characteristics. The training sequence timing is
`defined as a timing at which the microphone units M1 and
`M2 do not pick up the target sound. Therefore, the filter
`response characteristics of
`the adaptive filter 7 are
`determined so as to cancel the respective noise components
`contained in the main signal data Dm and the reference
`signal data Dr at the training sequence timing. The training
`sequence is successively performed at an appropriately
`timing so as to follow the state in which the noise
`generation state of the sound field is varied from moment to
`moment.
`(0029) As a result of the response characteristics of the
`adaptive filter 7 being determined successively in this way,
`the noise component contained in the main signal Vm is
`cancelled by the noise component contained in the
`reference signal Vr due to subtraction by digital adder 8.
`The output of the digital adder 8 is converted to an analog
`signal by a D/A converter 9. A signal primarily consisting
`of the target sound component can be acquired from the
`D/A converter 9. As a result of creating and testing a
`prototype of the circuit illustrated in FIG. 1 having the
`directional characteristics illustrated in FIG. 10, the S/N
`was improved by approximately 10 dB in comparison to a
`conventional primary pressure gradient microphone
`configured
`to obtain single directionality with
`two
`omnidirectional microphone units.
`(0030) Note that the adaptive filter 7, the digital adder 8,
`and the like can be configured a digital signal processor
`DSP. In addition, a low pass filter for removing high-
`frequency components may be disposed at the stage
`preceding the A/D converters 5 and 6.
`(0031) FIG. 8 illustrates an example of a case in which a
`primary pressure gradient microphone equivalent to that
`described above is configured using three omnidirectional
`microphone units. That
`is, a
`third omnidirectional
`microphone unit M3 is disposed near the omnidirectional
`
`
`
`Page 5 of 12
`
`
`
`Japanese Unexamined Patent Application Publication No. H11-18186
`(6)
`
`microphone unit M2. In FIG. 8, the reference signal Vr is
`generated by obtaining the difference between the output of
`the third microphone unit M3 and the output of the delay
`circuit 1 with the inversion adding circuit 4. The main
`signal Vm is generated in the same manner as in FIG. 1. In
`the case of this example, one additional omnidirectional
`microphone unit M3 is required in comparison to FIG. 1. In
`addition,
`the main microphone and
`the
`reference
`microphone are no longer physical disposed at exactly the
`same position. However, because the microphone units M2
`and M3 can be disposed near one another, there is a better
`noise cancelling effect than a configuration in which two
`sets of primary pressure gradient microphones are disposed
`apart
`from one another with
`reversed directional
`characteristics, as illustrated in FIG. 13, and this is
`achieved with one fewer microphone units.
`(0032) FIG. 9 illustrates an example of a case in which a
`microphone equivalent
`to
`that described above
`is
`configured using four omnidirectional microphone units.
`That is, four omnidirectional microphone units M1 to M4
`are disposed radially at 90° intervals on the same periphery.
`The main signal Vm is the same as in the configuration of
`the main microphone illustrated in FIG. 1. The reference
`microphone for forming
`the reference signal Vr
`is
`configured
`to acquire
`the difference between
`the
`microphone units M3 and M4 with the inversion adder 4.
`The axial line of the microphone units M3 and M4
`constituting the reference microphone is orthogonal to the
`axial line of the microphone units M1 and M2 constituting
`the main microphone. Accordingly,
`the directional
`characteristics thereof are as illustrated in FIG. 11, and the
`reference microphone does not have directional
`characteristics with respect to a target sound wave from the
`front surface. Therefore, a microphone having roughly the
`same two types of directional characteristics as in FIG. 1 is
`realized. In the case of this example, two additional
`omnidirectional microphone units M3 and M4 are required
`in comparison to FIG. 1. However, since the microphone
`units M1 and M4 are disposed radially at 90° intervals on
`the same periphery, the main microphone and the reference
`microphone can be considered to be physically disposed at
`the same position. In terms of the noise cancelling effect,
`this is superior to the configuration of FIG. 13 in which two
`sets of primary pressure gradient microphones are disposed
`apart from one another.
`(0033) Further, when four omnidirectional microphone
`units are used, as illustrated in FIG. 12, the main
`microphone and the reference microphone configured from
`two omnidirectional microphone units can be configured
`from different omnidirectional microphone units M1 to M4.
`The main microphone consists of the omnidirectional
`microphone units M1 and M2, the delay circuit 2, and the
`inversion adding circuit 3. The reference microphone
`consists of the omnidirectional microphone units M3 and
`M4, the delay circuit 1, and the inversion adding circuit 4.
`The main microphone has directional characteristics on the
`front surface side, as illustrated in FIG. 3, and the reference
`microphone has directional characteristics on the back
`surface side, as illustrated in FIG. 5. The primary pressure
`gradient microphone illustrated in FIG. 12 is housed in a
`single casing.
`
`(0034) The sequence of the omnidirectional microphone
`units M3 and M4 constituting the reference microphone is
`identical
`to
`the
`sequence of
`the omnidirectional
`microphone units M1 and M2 constituting the main
`microphone, and the microphone units M1 and M2 are
`disposed adjacent to the microphone units M3 and M4.
`That is, the distance Ls between the axial line of the
`microphone units M1 and M2 and the axial line of the
`microphone units M3 and M4 in FIG. 13 is much smaller
`than in the configuration illustrated in FIG. 13. In this
`regard, the