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 JP2006237952 PROBLEM TO BE SOLVED: To provide a microphone device having sharp directivity and a variable directional direction. SOLUTION: An array microphone 10 is configured by arranging nine microphones M0 to M8 in three rows and three columns in the same plane. A directivity function processing circuit 13 is provided which processes the output signal of the array microphone 10, converts it into a signal of single directivity, and outputs it. The directivity function processing circuit 13 expands a directivity function having an incident angle of sound wave as a variable into a Fourier series, and at least the third order term. The variables in this expansion formula are configured from the output signals of the microphones M0 to M8 that constitute the array microphone 10. [Selected figure] Figure 15 Microphone device [0001] The present invention relates to a microphone device. [0002] For example, in a video conference, since the voice of the speaker is generally collected by the microphone placed on the desk, ambient noise may also be collected together, and the voice signal output from the microphone may be unclear. is there. Therefore, when the voice of the speaker is collected by the microphone, a clear voice signal can 03-05-2019 1 be obtained by the following method. [0003] That is, the first method is a method of giving directivity to a microphone, emphasizing voice at a stage where voice is inputted to the microphone, and suppressing noise, and a second method is to output from the microphone Noise component by adaptively processing the voice signal. According to these methods, since the level of the noise component contained in the audio signal is relatively small, a clear audio signal can be obtained. [0004] Then, six microphones are arranged around the reference microphone (microphone unit) as a microphone device adopting the above first method, and the microphone device is synthesized by combining the outputs of the respective microphones using Fourier transform. There is a thing which obtains a single directivity as a whole. [0005] In addition, there exist the following as a prior art document which shows such a microphone apparatus. Japanese Patent Application Laid-Open No. 2002-271885 [0006] However, in the above microphone device, when combining the outputs of the microphones, it is regarded that there is only one sound source, and the value of the first order approximation term in the Fourier transform is determined, and further, the value of this first order approximation term Since the value of the third order approximation term is obtained from the above, the directivity width (angle range in which the gain can be obtained) of a single directivity can be obtained as wide as about ± 60 ° with respect to the main axis. [0007] However, when the directivity is wide like this, in an environment such as a plurality of sound 03-05-2019 2 sources or noise sources, a sufficient effect as a directional microphone can not be obtained. [0008] SUMMARY OF THE INVENTION In view of such points, the present invention is intended to narrow the directivity width of a unidirectional microphone device and make it possible to electrically change the directivity direction. [0009] In the present invention, a microphone device that processes and outputs an output signal of an array microphone constituted by at least nine microphones, converts the output signal of the array microphone into a unidirectional signal and outputs the signal. A directivity function processing circuit is provided, and the directivity function processing circuit expands the directivity function using the incident angle of the sound wave as a variable into a Fourier series, and at least the third order term, and The microphone apparatus is configured such that the variable is configured from an output signal of a microphone that configures the array microphone. [0010] According to the present invention, it is possible to obtain sharp unidirectionality and to change the pointing direction. [0011] [1] Directionality Function A microphone is a converter that converts sound waves output from a sound source into audio signals (audio signals), and its conversion characteristics are predetermined characteristics with respect to the direction, frequency, etc. of the input sound waves. Is considered to indicate. [0012] Therefore, the characteristics of the microphone can be expressed by equation (1) in FIG. Here, the conversion characteristic D (θ, ω) is a function that changes according to the direction θ and the angular frequency ω of the input sound wave, and indicates the directivity of the microphone. 03-05-2019 3 For this reason, the conversion characteristic D (θ, ω) is generally called a “directivity function”. That is, the directivity function indicates the directivity of the microphone. [0013] For example, in the case of an omnidirectional (omnidirectional) microphone, its directivity pattern is as shown in FIG. 2A, and its directivity function is shown by D (θ, ω) = 1. Also, in the case of a bi-directional (bi-directional) microphone, its directivity pattern is as shown in FIG. 2B, and its directivity function is shown by D (θ, ω) = cos θ. [0014] And although Formula (1) is a case where there is one sound source, since Formula (1) is satisfied about each sound source when there are a plurality of sound sources, it can be expressed by Formula (2) of FIG. it can. [0015] [2] Analysis of Unidirectionality FIG. 3 shows the directivity function (directionality) that a unidirectional microphone is ideal. Where θ is the direction (angle) of the sound source viewed from the microphone θc: directivity direction (direction of directivity) θw: directivity width (angular range in which a predetermined gain can be obtained). [0016] Since this characteristic is considered to be a directivity function with respect to the variable θ, it can be represented by the equation (3) in FIG. 4 in the Fourier series, and this equation (3) is 03-05-2019 4 further expanded to an approximate equation up to n = 3 If it represents, it will become (4) Formula of FIG. [0017] In the equation (4), for example, when θw = 60 ° and the directivity direction θc is changed, directivity characteristics as shown in FIGS. 5A to 5C are obtained. That is, as shown in FIGS. 5A to 5C, if the directivity function satisfies the equation (4), relatively sharp directivity can be obtained, and the directivity direction θc can be arbitrarily changed. [0018] [3] Preparation of directivity function Now, as shown in FIG. 6, the array microphone 10 is configured by arranging nine microphones (microphone units) M0 to M8 in the same plane in three rows and three columns. However, the microphones M0 to M8 are nondirectional. In addition, the spacing in the row direction and the spacing in the column direction of each of the microphones M0 to M8 are all equal to the value d. Further, the central microphone M4 is used as a reference microphone. As an example, the microphones M0 to M8 are pressure type electret condenser microphones, and d = 21 mm. [0019] Then, there is a sound source (not shown) in the plane including the array microphone 10, and R: distance between the sound source and the reference microphone M4 θ: incident angle of the sound wave to the microphones M0 to M8 or directivity direction. The distance R is assumed to be sufficiently larger than the distance d of the microphones. Although the incident angle θ is arbitrary, in FIG. 6, the row direction of the arrangement of the microphones M0 to M8 is θ = 0. 03-05-2019 5 [0020] Further, the sound wave output from the sound source is shown by the equation (5) in FIG. 7 and x Mi (t): It is assumed that the output signal of the microphone Mi (i = 0 to 8). [0021] In such an array microphone 10, equation (1) is applied to the reference microphone M4, and equation (3) is substituted for deformation to obtain equation (6) in FIG. However, in Formula (6), it is approximation to n = 3 like (4) Formula. [0022] According to the equation (6), in the case of the array microphone 10, if cos θ to cos 3 θ and sin θ to sin 3 θ can be obtained, for example, directivity as shown in FIGS. 5A to 5C can be obtained. If the Fourier coefficients a0 to a3 and b1 to b3 are changed corresponding to θw, the pointing direction can be changed as shown in FIGS. 5A to 5C. [0023] [4] cos θ to cos 3 θ, sin θ to sin 3 θ In the equation (6), the values of cos θ to cos 3 θ and sin θ to sin 3 θ are required. These are microphones M0 to M3, as described in detail below. Obtained from the output signals of M5 to M8. [0024] [4-1-1] cos θ As shown in FIG. 8, when the sound wave output from the sound source is input to the microphones M3, M4, and M5 in the center row of the array microphone 10, the sound source is output Assuming that the sound wave is represented by equation (5) in FIG. 7, the difference in path length as shown in FIG. 8 is generated between the sound source and the microphones M3 to M5. The signal can be represented by equation (7) in FIG. In equation (7), the path length difference is based on the distance between the sound source and the reference microphone M4. 03-05-2019 6 [0025] Therefore, equation (8) in FIG. 8 can be obtained by obtaining the difference between the output signal of the microphone M3 and the output signal of the microphone M5. Then, applying the relation of the approximate expression sin α = α to the equation (8), the equation (8) becomes as shown in the equation (9) of FIG. 8 and the equation (10) can be modified by modifying the equation (9) You can get it. According to the equation (10), cos θ can be obtained by arithmetic processing of the output signals of the microphones M3 and M5. [0026] Further, when the microphone M4 is assumed to be at the center position between the microphones M3 and M5, it can be understood from the output signals of the microphones M3 and M5 that the output signal of the microphone M4 can be generated according to equation (10). Further, equation (10) also indicates that bi-directionality shown in FIG. 2B can be obtained by arithmetic processing of the output signals of the microphones M3 and M5. [0027] [4-1-2] When sin θ As shown in FIG. 9, when the sound wave output from the sound source is input to the microphones M1, M4 and M7 in the center row of the array microphone 10, the sound source and the microphones M1 and M4 are used. And M7, as shown in FIG. 9, the output signals of the microphones M1, M4 and M7 can be expressed by the equation (11) of FIG. In the equation (11), the path length difference is based on the distance between the sound source and the reference microphone M4. [0028] Then, when the difference between the output signal of the microphone M1 and the output signal of the microphone M7 is obtained, the equation (12) in FIG. 9 can be obtained. Then, when 03-05-2019 7 the relation of the approximate expression sin α = α is applied to the equation (12), the equation (12) becomes as shown in the equation (13) of FIG. 9 and the equation (14) is modified by modifying the equation (13) You can get it. [0029] Then, according to equation (14), sin θ can be obtained by arithmetic processing of the output signals of the microphones M1 and M7. Further, this equation (14) indicates that bi-directionality obtained by rotating the bi-directionality shown in FIG. 2B by 90 ° can be obtained by arithmetic processing of the output signals of the microphones M1 and M7. [0030] [4-2-1] cos 2θ Equation (10) also indicates that the output signal of the microphone M4 at the center thereof can be obtained from the output signal of the microphone M3 and the output signal of the microphone M5. [0031] Therefore, as shown in FIG. 10, a virtual microphone V3 is assumed at the center of the microphone M3 and the microphone M4, and a virtual microphone V5 is assumed at the center of the microphone M4 and the microphone M5. [0032] Then, the output signals of these virtual microphones V3 and V5 can be processed in the same manner as when formula (10) is derived, and can be expressed by formulas (15) and (16) in FIG. Then, when the difference between the equation (15) and the equation (16) is obtained, the equation (17) of FIG. 10 can be obtained, and in the same manner as deriving the equation (8) to the equation (10), The equation (18) in FIG. 10 can be obtained from the equation (17). [0033] Then, when the equation (18) is substituted into the equation (17) and organized, it becomes the equation (19). If the double angle formula shown as the equation (20) in FIG. 10 is applied to the 03-05-2019 8 equation (19), Equation (21) can be obtained, and equation (22) in FIG. 10 can be obtained by modifying equation (21). [0034] According to the equation (22), cos2θ can be obtained by arithmetic processing of the output signals of the microphones M3 to M5. [0035] [4-2-2] In the case of sin 2θ In this case as well, sin 2θ can be obtained in the same manner as when cos 2θ is obtained. That is, as shown in FIG. 11, a virtual microphone V3 is assumed at the center of the microphone M0 and the microphone M6, and a virtual microphone V5 is assumed at the center of the microphone M2 and the microphone M8. [0036] Then, the output signals of these virtual microphones V3 and V5 can be expressed by the equations (23) and (24) in FIG. 11 in the same manner as when the equation (14) is derived. Then, when the difference between the equation (23) and the equation (24) is obtained, the equation (25) of FIG. 11 can be obtained, and in the same manner as deriving the equation (8) to the equation (10), Formula (26) of FIG. 11 can be obtained from Formula (25). [0037] Then, when the equation (26) is substituted into the equation (25) and organized, it becomes the equation (27). If the double angle formula shown as the equation (28) in FIG. 11 is applied to the equation (27), Equation (29) can be obtained. [0038] According to the equation (29), cos2θ can be obtained by arithmetic processing of the output 03-05-2019 9 signals of the microphones M0, M2, M6 and M8. [0039] [5-2-1] cos 3θ As shown in FIG. 12, a virtual microphone V0 is assumed at the center of the microphone M0 and the microphone M3, and a virtual microphone V6 is assumed at the center of the microphone M3 and the microphone M6, A virtual microphone V3 is assumed at the position of the microphone M3. Further, a virtual microphone V5 is assumed at the center of the microphone M2 and the microphone M5, a virtual microphone V8 is assumed at the center of the microphone M5 and the microphone M8, and a virtual microphone V5 is assumed at the position of the microphone M5. [0040] Then, the output signals of these virtual microphones V0 and V6 can be processed in the same manner as when equation (14) is derived, and can be expressed by equations (30) and (31) in FIG. Then, when the difference between the equation (30) and the equation (31) is obtained, the equation (32) of FIG. 12 can be obtained, and in the same manner as deriving the equation (8) to the equation (10), The equation (33) in FIG. 12 can be obtained from the equation (32). Then, when equation (33) is substituted into equation (32) and organized, equation (34) is obtained. Similarly, the equation (35) is obtained for the virtual microphones V2, V8 and V5. [0041] Then, assuming the virtual microphone V4 at the position of the microphone M4, and obtaining the output signal of the virtual microphone V4 from the equations (34) and (35), the equation (36) in FIG. 12 is obtained. Then, by substituting the equations (36) and (10) into the triple angle formula shown as the equation (37) in FIG. 12, the equation (38) in FIG. 12 can be obtained. 03-05-2019 10 [0042] According to the equation (38), cos3θ can be obtained by arithmetic processing of the output signals of the microphones M0, M2, M3, M5, M6, and M8. [0043] [5-2-2] When sin 3θ As shown in FIG. 13, the virtual microphone V3 is assumed at the position of the microphone M3, the virtual microphone V4 is assumed at the position of the microphone M4, and the virtual microphone V5 is located at the position of the microphone M5. Assume. [0044] Then, the output signals of these virtual microphones V3, V4, V5 can be processed in the same manner as when formula (10) is derived, and can be expressed by formulas (39), (40), (41) in FIG. . [0045] Then, assuming a virtual microphone Va at the center of the virtual microphone V3 and the virtual microphone V4 and a virtual microphone Vb at the center of the virtual microphone V4 and the virtual microphone V5, the output signals of these virtual microphones Va and Vb are the same. Equations (42) and (43) in FIG. Then, when the output signal of the virtual microphone V4 is obtained from the signals represented by the equations (42) and (43), the equation (44) in FIG. 13 is obtained. [0046] Therefore, equation (46) in FIG. 13 can be obtained by substituting equations (44) and (14) into the triple angle formula shown as equation (45) in FIG. [0047] According to the equation (46), sin3θ can be obtained by arithmetic processing of the output signals of the microphones M0 to M3 and M5 to M8. 03-05-2019 11 [0048] [5] Synthesis of Microphone Outputs If the equations (10), (14), (22), (29), (38) and (46) are substituted into cos θ to cos 3 θ and sin θ to sin 3 θ in equation (6), FIG. Equation (47) of can be obtained. According to equation (47), by combining the output signals of the other microphones M0 to M3 and M5 to M8 with the output signal of the reference microphone M4, as shown in FIG. It can be seen that the directivity function) is obtained, and the pointing direction θc can be arbitrarily changed. [0049] In the equation (47), a part of the terms is multiplied by 1 / (jω), but in this operation, the corresponding signal may be subjected to Fourier transform and processed in the frequency domain. That is, multiplication of 1 / j means that the phase of the audio signal component of each frequency is advanced by 90 °, and in the actual arithmetic processing, the value of the imaginary part of the audio signal component of each band after Fourier transform is a real part The value of the real part may be sign-inverted and replaced with the value of the imaginary part. [0050] In addition, since the amplitude (level) of the signal component changes corresponding to the frequency (ω / 2π) by being multiplied by 1 / ω, this amplitude is also corrected. [0051] [6] Embodiment FIG. 15 shows an example of a microphone device according to the present invention. 03-05-2019 12 In this microphone device, the directivity width θw is narrow and the directivity direction θc can be changed in accordance with the above-mentioned idea. [0052] That is, while the array microphone 10 is configured as described with reference to FIG. 6, the output signals of the microphones M0 to M8 are supplied to the 9-channel A / D converter circuit 12 through the 9-channel microphone amplifier 11 to be digital signals. The digital signal is A / D converted, the digital signal is supplied to the directivity function processing circuit 13, the processing shown in the equation (47) is executed (the specific processing method will be described later), and the signal y (t) is extracted. [0053] Then, this output signal y (t) is supplied to the D / A converter circuit 14 to be D / A converted into an analog signal, and the analog signal is taken out to the output terminal 15 as a microphone output. [0054] At this time, the directivity function processing circuit 13 is constituted of, for example, a microcomputer, and the operation key 13C is connected. Then, when the pointing direction θc and the pointing width θw are designated by the operation key 13C, Fourier coefficients a0 to a3 and b1 to b3 corresponding to the pointing direction θc and the pointing width θw are generated and set in the equation (47). Thus, in the processing circuit 13, the output signals of the microphones M0 to M8 have characteristics corresponding to the directivity direction θc and the directivity width θw, and are synthesized into the signal represented by the equation (47). [0055] Therefore, according to the apparatus of FIG. 15, it is possible to obtain a microphone apparatus 03-05-2019 13 in which the directivity width θw is narrow and the directivity direction θc is variable. Moreover, in that case, according to equation (47), the parameters necessary for the calculation are only the output signals of the microphones M0 to M8 and the values for defining the directivity characteristics (values indicating the directivity direction θc and the directivity width θw) The directivity can be set even if it is unknown from which direction the sound wave is coming. [0056] 16A and 17A show simulation results of the directivity of the microphone device according to the present invention, and FIGS. 16B and 17B show simulation results of the directivity of the microphone device described in the prior art. Note that, as is apparent from FIG. 16, in the main frequency band, the frequency characteristics are almost flat, so in FIG. 17, the case where the frequency of the sound wave is 1.5 kHz is represented. [0057] And, according to these figures, the microphone device according to the invention (characteristics of FIGS. 16A and 17A) is unidirectional compared to the microphone device (characteristics of FIGS. 16B and 17B) described in the prior art. It can be seen that the directivity as a sex microphone is improved. In particular, in the range of θ <−60 ° or θ> 60 °, the sound waves from those directions are considerably suppressed. [0058] [7] Processing Contents and Processing Procedure of Directional Function Processing Circuit 13 The directional function processing circuit 13 can realize the processing of the equation (47) by executing, for example, the routine 100 shown in FIG. In this example, the 2048 sample period of the audio signal is one frame period. 03-05-2019 14 [0059] The processing of the routine 100 starts from step 101, and then, at step 102, the output signals of the microphones M0 to M8, that is, the audio data output from the A / D converter circuit 12 are acquired by 1 sample × 9 channels. Be Subsequently, in step 103, the sums and differences within {} in equation (47) are calculated. For example, in the case of the term (the term corresponding to the equation (10)) on the third line in the equation (47), {xM3 (t) −xM5 (t)} is calculated. [0060] Then, in step 104, it is checked whether or not the processing of steps 102 and 103 has been performed for one frame period, and if not, the processing returns to step 102. [0061] Thus, when the processes of steps 102 and 103 are performed for one frame period, the process proceeds from step 104 to step 111. In step 111, the calculation result of step 103 is converted to frequency domain data by FFT processing. In the following step 112, the part which is a coefficient with respect to {} in the equation (47) is phase-transformed. For example, in the case of the term (the term corresponding to the equation (10)) in the third line of the equation (47), c / (2jωd) is obtained with respect to {xM3 (t) −xM5 (t)}. Since it is a coefficient, its value c / (2ωd) is calculated and converted to the value of the real part. [0062] Subsequently, in step 114, the Fourier coefficients a0 to a3 and b1 to b3 corresponding to the target directivity are multiplied by the values calculated in steps 103 and 113, and the sum of Fourier series is calculated ( 47) The value of the equation is calculated, and the calculation result is IFFT-processed in step 114 to be converted into data in the time domain. [0063] Then, in step 121, data of the conversion result of step 115 is supplied to the D / A converter circuit 14 every sample period for every one sample period, and in step 122, the process of step 03-05-2019 15 121 is performed for one frame period. If it is not checked, the process returns to step 121. [0064] Thus, when the process of step 121 is performed for one frame period, the process proceeds from step 122 to step 123. In this step 123, the process of one frame period is ended. [0065] Thus, according to the routine 100, the processing according to the equation (47) can be realized. Then, in this routine 100, the value in {} is calculated for each sample in step 103 before performing the FFT process in step 111, so that the process can be performed appropriately and smoothly. . [0066] [8] Another Calculation Method of cos 2θ FIGS. 19 and 20 show another calculation method of cos 2θ. That is, although cos 2θ can be modified as shown in equation (48) in FIG. 19, at this time, assuming that the angles θ and φ are in the relationship of equation (49) in FIG. Equation (50) in FIG. 19 is obtained. [0067] Then, as shown in FIGS. 20A and 20B, the microphones M0, M2, M6, and M8 are rotated by 45 ° (= φ-θ) in the direction in which the incident angle θ decreases, with the reference microphone M4 as the center. Assuming the virtual microphones V0, V2, V6, and V8, the incident angle of the sound wave to the virtual microphones V0, V2, V6, and V8 is the angle φ from the relationship of the equation (49). [0068] 03-05-2019 16 Therefore, the relationship between the sound wave of incident angle φ and the output signals of virtual microphones V0, V2, V6, V8 is equal to the relationship between the sound wave of incident angle θ and the output signals of microphones M0, M2, M6, M8. By processing the output signals of the microphones V0, V2, V6, and V8 in the same manner as the equation (29) (represented in FIG. 19), the signal represented by the equation (51) in FIG. 19 can be obtained. [0069] Then, as shown in FIG. 20C, the positions of the virtual microphones V0, V2, V6, and V8 are changed to the direction of the reference microphone M4, and are assumed to be the positions of the microphones M3, M1, M7, and M5. Then, the output signals of the virtual microphones V0, V2, V6, V8 are equal to the output signals of the microphones M3, M1, M7, M5. The interval between the virtual microphones V0, V2, V6, and V8 has a value 2d in FIG. 20B, but has a value √2 · d in FIG. 20C. Therefore, in the case of FIG. 20C, equation (51) becomes equation (52) in FIG. [0070] Then, equation (53) in FIG. 19 is obtained by substituting equation (50) into equation (52). Therefore, equation (47) can also be calculated using this equation (53). [0071] [9] Others For example, in {(10)} indicates that the difference signal between the output signal of the microphone M3 and the output signal of the microphone M5 is obtained, but the interval d between the microphones M0 to M8 is narrow. If the frequency of the sound wave to be input is low, the difference between the sound wave to be input to the microphone M3 and the sound wave to be input to the microphone M5 becomes small, and the level of the above difference signal becomes small. 03-05-2019 17 [0072] On the other hand, when the interval d is wide, when the frequency of the sound wave to be input is high, the path difference between the sound wave to be input to the microphone M3 and the sound wave to be input to the microphone M5 becomes one wavelength or more. Processing becomes inappropriate. [0073] And since the above thing is the same also about the difference signal or sum signal of the output signal of all the microphones M0-M8, as a result, the precision of the arithmetic processing in (47) Formula falls and the directivity which is aimed at It may be difficult to obtain. [0074] Therefore, in such a case, if two sets of array microphones 10 are prepared, and one of the array microphones and the other array microphones have different microphone spacings d and a common central microphone as a reference Good. Then, the low frequency component of the audio signal is extracted from the wide-spaced microphone of the microphone, and the high frequency component of the audio signal is extracted from the narrow-spaced microphone, and processing of equation (47) is performed on the signal obtained by adding both components. If implemented, good directivity can be obtained over a wide band. [0075] Further, in the above-described microphone device, when the noise comes from the same direction as the intended sound wave, the noise can not be suppressed, but in such a case, for example, the output signal of the directivity function processing circuit 13 Adaptive processing can suppress the noise signal. By doing so, even if noise is mixed in the voice of the speaker in a video conference or the like, the noise can be suppressed and a clear voice signal can be obtained. 03-05-2019 18 [0076] Furthermore, first, the direction of the sound source is detected, and then the directivity direction θc and the directivity width θw can be set again according to the detection result to emphasize a target signal or suppress unnecessary signals. That is, the directivity function can be set so as to capture sound in a specific direction or not capture sound in a specific direction. Also, by arranging a plurality of sets of array microphones 10 in the same plane and orienting the respective pointing directions to a specific point, it is also possible to emphasize the sound of the sound source at that point. [0077] Also, the target sound can be collected more clearly by setting the directivity direction to the direction of the target sound and the direction of the noise sound and subtracting the signal in the direction of the noise sound from the signal of the direction of the target sound. it can. Also, it is possible to estimate and remove the sound wave input regardless of the pointing direction, such as noise from the vertical direction. [0078] Furthermore, when providing a function such as an echo canceller, for example, by separately learning the information of the impulse response of the echo canceller with respect to the output of the array having directivity in directions of 5.degree. It is possible to instantly remove the echo of the voice in the direction of pointing. Alternatively, for example, the information on the impulse response of the echo canceller is separately learned in eight directions, and the impulse response in the direction in which the direction is to be directed is the whole by using the impulse response in the near direction among eight directions as the initial value. It is possible to reduce the amount of calculation of and to reduce the echo remainder more than when calculating from a completely initial value. [0079] [List of abbreviations] A / D: Analog to Digital D / A: Digital to Analog FFT: Fast Fourier 03-05-2019 19 Transform IFFT: Inverse Fast Fourier Transform [0080] It is a figure which shows the directivity function of a microphone. It is a characteristic view showing directivity of a microphone. It is a characteristic view for analyzing the directivity of a unidirectional microphone. It is a figure which shows the analysis result of the directivity of a unidirectional microphone. It is a characteristic view showing directivity of a unidirectional microphone. It is a layout for demonstrating the array microphone in this invention. It is a figure which shows the directivity function of a unidirectional microphone by an approximation formula. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. It is a figure which shows the directivity function to which this invention is applied. It is a systematic diagram showing an example of this invention. It is a figure which shows the characteristic of the example of application of this invention. It is a figure which shows the characteristic of the example of application of this invention. It is a flowchart which shows an example of the routine which implement | achieves the directivity function of FIG. It is a figure for demonstrating a part of directivity function. It is a figure for demonstrating a part of directivity function. Explanation of sign [0081] DESCRIPTION OF SYMBOLS 10 ... Array microphone, 12 ... A / D converter circuit, 13 ... Directionality function processing circuit, 14 ... D / A converter circuit, M0-M8 ... Microphone (microphone unit) 03-05-2019 20

1/--страниц