JP2010028591A - Digital acoustic signal processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a digital acoustic signal processing apparatus capable of forming an acoustic beam, which is narrow and elongated from near a speaker array to behind a focal position and includes high directivity, by setting a focal point of a convergent beam of directional array speakers to a long distance. <P>SOLUTION: The present invention relates to a digital acoustic signal processing apparatus in which a digital acoustic signal input from a sound source is played back by a speaker array comprised of a plurality of speakers arrayed along a direction of a straight line while controlling a directional property through a two-dimensional (2D) digital filter. When a spectrum of the acoustic signal, on which 2D Fourier transform is performed with respect to a time and speaker positions, is represented on a 2D frequency plane constituted of a temporal frequency axis and a spatial frequency axis, the 2D digital filter expands an amplitude property in low frequencies in parallel with the spatial frequency axis to a non-physical domain and sets the spectral as a single-peak pass area of which the cross section is spread from a physical domain to the non-physical domain. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a digital acoustic signal processing apparatus that generates an acoustic reproduction signal that can emit sound waves with high directivity in a specific direction by an array speaker in which a plurality of speakers are arranged.

  Applications for radiating sound in a narrow range in a specific direction are increasing, but such super-directional beam forming has not been commercialized yet other than by an ultrasonic method (see Non-Patent Document 1). However, a super-directional beam can be formed by designing a very narrow directional characteristic with a normal directional array speaker (Non-Patent Documents 2 and 3) composed of a speaker array and a two-dimensional digital filter. Here, in order to improve the low frequency characteristics as much as possible, a large size is required for the speaker array. However, if the low frequency is not so important, the speaker array can be made small.

  Similarly, there is a method of forming a focused acoustic beam at a specific position, that is, at the focal point using a speaker array, and the two-dimensional digital filter used has a two-dimensional phase characteristic of a linear phase that varies depending on the direction for focusing the sound. (See Non-Patent Documents 4 and 5). In this method, at a distance close to the speaker array, the phase characteristic works effectively, and an acoustic beam that is focused well at the focal position is formed.

Kiyoshi Nishikawa, “Two-dimensional domain analysis of beamforming”, IEICE Transactions, The Institute of Electronics, Information and Communication Engineers, September 1994, Vol. J77-A, No. 9, p. 1304-1306 Matsumoto Yasushi and Nishikawa Kiyoshi, “Design Method of Directional Array Speaker with Constant Sidelobe”, IEICE Technical Report, The Institute of Electronics, Information and Communication Engineers, October 2004, EA 2004-74, p. 13-18 Kiyoshi Nishikawa, 4 others, "Design method of 2D FIR fan filter for wideband beam forming by 2D Fourier series approximation", IEICE Transactions, IEICE, December 2000, J833-A Volume 12, No. 12, p. 1357-1367 Kiyoshi Nishikawa, Tetsuya Yokoyama, Mikiko Miyagishi, “Method of moving sound image using linear speaker array and 2D FIR filter”, IEICE Transactions, Institute of Electronics, Information and Communication Engineers, July 2000, J83- Volume A, No. 7, p. 839-849 Kiyoshi Nishikawa, Satoshi Shimura, Tetsuya Yokoyama, Mikiko Miyagishi, “Sound image movement and focused beam formation using a two-dimensional digital filter” AES Tokyo Convention '99 Proceedings, p. 166-169

In the case of narrow-angle beam formation in which the focal position is set far, the phase setting range in the low band is narrow, so that the effect of phase characteristics is difficult to be produced, and an acoustic beam that converges well at the focal position is not formed.

  An object of the present invention is to set a focal point of a focused beam of a directional array speaker at a long distance, thereby forming a narrow and narrow acoustic beam having high directivity from the vicinity of the speaker array to the rear of the focal position. An acoustic signal processing apparatus is provided.

The present invention is a digital audio signal processing apparatus that reproduces a digital audio signal input from a sound source by a speaker array composed of a plurality of speakers arranged along a straight line direction with a directional characteristic controlled by a two-dimensional digital filter. And
The two-dimensional digital filter has a low-frequency amplitude characteristic when a spectrum of an acoustic signal obtained by two-dimensional Fourier transform with respect to time and a speaker position is represented on a two-dimensional frequency plane composed of a time frequency axis and a spatial frequency axis. Is a digital acoustic signal processing device characterized in that a cross section parallel to the spatial frequency axis is set as a unimodal passband extending from the physical region to the non-physical region.

Further, the present invention is a digital acoustic signal processing apparatus for controlling directivity characteristics of a signal output from a microphone array composed of a plurality of microphones arranged along a straight line direction by a two-dimensional digital filter,
The two-dimensional digital filter has a low-frequency amplitude characteristic when the spectrum of an acoustic signal two-dimensionally Fourier transformed with respect to time and microphone position is represented on a two-dimensional frequency plane composed of a time frequency axis and a spatial frequency axis. Is a digital acoustic signal processing device characterized in that a cross section parallel to the spatial frequency axis is set as a unimodal passband extending from the physical region to the non-physical region.

  According to the present invention, the two-dimensional digital filter that controls the directivity characteristics of the digital sound signal input from the sound source has the time frequency axis, the spatial frequency axis, and the spectrum of the sound signal that has been two-dimensional Fourier transformed with respect to time and the position of the speaker. Since the amplitude characteristics in the low frequency range are expanded to the non-physical region when expressed on a two-dimensional frequency plane consisting of: An acoustic beam having a narrow and long high directivity can be formed to the rear.

  According to the present invention, the two-dimensional digital filter for controlling the directivity of the signal output from the microphone array includes a time frequency axis and a spatial frequency axis for a spectrum of an acoustic signal obtained by two-dimensional Fourier transform with respect to time and a microphone position. When it is expressed on a two-dimensional frequency plane consisting of the above, the amplitude characteristic in the low band is expanded to the non-physical area, so that a narrow and long beam having high directivity can be obtained.

  FIG. 1 is a block diagram showing the overall configuration of a digital acoustic signal processing apparatus including a two-dimensional digital filter 4 according to an embodiment of the present invention. The digital audio signal processing apparatus according to the present embodiment is connected in series to a speaker array 2 composed of a plurality (seven in this embodiment) of speakers 2a, 2b,..., 2f, 2g, and the speakers 2a to 2g of the speaker array 2. And a two-dimensional digital filter 4 composed of one-dimensional filters 3a, 3b,..., 3f, 3g. The two-dimensional digital filter 4 is realized by, for example, a two-dimensional FIR filter.

  In this embodiment, instead of using parameters of frequency and direction, a two-dimensional digital fan filter on a two-dimensional frequency plane represented by a time frequency and a spatial frequency is designed to approximate a desired frequency and directivity. The procedure will be described. In the case of narrow-angle beam forming where the focal position is set far, the phase setting range in the low range is narrow and the effect is difficult to produce, so the phase setting range in the low range is expanded and the amplitude is set Describes a digital acoustic signal processing apparatus including a two-dimensional digital filter capable of obtaining high directivity by setting phases in all regions, that is, not only in a region in which a spectrum is distributed but also in a non-physical region 5 in which a spectrum is not distributed. To do.

As shown in FIG. 1, the directional array speaker 1 includes a speaker array 2 (speaker interval D, number of speakers (N ₂ +1)) and a two-dimensional digital filter 4, and amplitude characteristics of the two-dimensional digital filter 4. By setting the phase characteristics, as shown in FIG. 1, the focal point of the acoustic beam can be formed as a virtual sound source point at a point P at an angle φ _{0 and a} distance r ₀ from the center point 0 of the speaker array 2. The outline of the design method of the two-dimensional digital filter 4 will be described below.

  FIG. 2 is a graph for explaining the relationship between the spectrum of the two-dimensional acoustic signal and the fan filter characteristic in the two-dimensional frequency plane, the horizontal axis is the time frequency f1, and the vertical axis is the spatial frequency f2.

  In the two-dimensional frequency plane, the non-physical region 5 is a region where the absolute value f2 / f1 of the ratio value of f2 and f1 is larger than ρ, with a straight line of φ = 90 ° and a straight line of φ = −90 ° as a boundary. That is, it refers to a region where the spectrum of the acoustic signal does not exist. A region excluding the non-physical region 5 on the two-dimensional frequency plane is referred to as a physical region 6. The amplitude characteristic of the fan filter in the low band is set as a single-peak pass band extending from the physical region 6 to the non-physical region 5 with a cross section parallel to the spatial frequency f2.

(Amplitude characteristics setting)
The spectrum of the acoustic wave radiated from the speaker array 2 at an angle φ is distributed on a straight line of the following expression on the two-dimensional frequency plane.

However, _{f 1} is the temporal frequency, _{f 2} is the spatial frequency, respectively 1 / T: is normalized with (T sampling interval) and 1 / D.

Directional sound with φ = φ ₀ as beam center, φ = φ _{P +} to φ _P− as beam shoulder width, φ = φ _{C +} to φ _{C− as} beam half width, and φ = φ _{S +} to φ _{S− as} beam spread. In order to form a beam, the fan filter characteristic of a wedge-shaped transient region is set and used as the target amplitude characteristic A _d (f ₁ , f ₂ ) of the two-dimensional digital filter 4 as shown in FIG. In FIG. 2, ψ is an angle formed by the center of the passband and the f ₁ axis, α ₁ , α ₂ and β ₁ , β ₂ are parameters that determine the opening width of the passband and the transition zone, respectively, and ψ is the beam Depending on the correspondence with the center angle φ ₀ , α ₁ , α ₂ and β ₁ , β ₂ are determined by the correspondence between the half-value cutoff angles φ _C− and φ _{C +} of the beam and the beam end angles φ _S− and φ _{S +} , respectively. . At this time, the following equation is used in order to avoid the influence of spatial aliasing.

The region of | f ₂ |> ρ | f ₁ | is called a non-physical region 5 because no spectrum exists.

When handling φ ₀ = 0 ° (ψ = 0 °), φ _{C +} = −φ _C− = φ _C , φ _{S +} = −φ _S− = φ _S , α ₁ = α ₂ = α, β ₁ = β Let ₂ = β. Further, in the low range for improved amplitude deterioration by finite order, and set a square characteristic of the width 2W on f ₂ axis.

  In FIG. 1, the propagation delay difference τ between both wavefronts in the φ direction from the origin is set by transforming the wavefront (solid line) centered at the origin 0 into a small circle wavefront (broken line) centered at the point P. (Φ) is derived from the wavefront position as ∞, and the phase derived from this

Is set as the phase characteristic of the two-dimensional digital filter 4. In order to ensure beam formation, the beam end angle φ _{S +} = φ ₀ + φ _{e +} and φ _S− = φ ₀ −φ _e− are set. However, φ _{e +} and φ _e− are angles determined by the length N ₂ · D of the speaker array 2 and the focal position (distance r ₀ , angle φ ₀ ) of FIG. 1, and are given by the following equations.

Note that the phase according to the equation (3) is effectively set as ρ sinφ _S− · f ₁ <f ₂ <ρ sinφ _{S +} · f ₁ .

The target characteristic H _d (f ₁ , f ₂ ) of the two-dimensional digital filter 4 is expressed as follows using the amplitude A _d (f ₁ , f ₂ ) and the phase θ _d (f ₁ , f ₂ ).

When the order is (N ₁ , N ₂ ), the transfer function H (z ₁ , z ₂ ) of the two-dimensional digital filter obtained by the two-dimensional discrete Fourier series approximation is

However, the values of M ₁ and M ₂ are selected to be values of powers of 2 that are about 10 times greater than N ₁ and N ₂ .
FIG. 3 is a diagram showing a setting region of amplitude and phase when φ ₀ = 0 °. FIG. 3A is a diagram illustrating an amplitude setting area, and FIG. 3B is a diagram illustrating a phase setting area. In this study, a method of forming a super-directivity (narrow directivity) beam with a small beam end angle φ _S is examined by setting the beam center φ ₀ = 0 ° and setting the focal position r ₀ far. When phi _S is small, as can be seen from FIG. 3 (b), a wide range of over the midrange low-frequency, not the set enough change in phase according to equation (3) appears, for which the focus formation As a result, the beam formation is insufficiently focused. Therefore, in order to enhance the focus formation, a method for setting the phase characteristic in the low frequency region including the non-physical region 5 is studied. Note that f ₁ = f _L shown in FIG. 3A is the lower end of the band that gives the same directivity, and almost coincides with the upper end of the low band where the square amplitude of width 2W is set, and is given by It is done.

FIG. 4 is a diagram showing enlargement (1) of the phase setting area. First, two methods are considered as a method of setting the phase beyond the phase setting range of FIG. 3B to the amplitude setting range of FIG. 3A wider than that. In FIG. 4A, the same value as the phase at the boundary φ = φ _S is set in the region of ρ sin φ _S | f ₁ | <| f ₂ | <W while changing in the f ₂ direction. b), the phase according to Equation (3) is applied to the region of ρsinφ _S | f ₁ | <| f ₂ | <ρ | f ₁ | (| f ₂ | <W), and ρ | f ₁ | <| f ₂ | < the W region of is for setting remained the same value as the phase at the boundary phi = 90 ° in the f ₂ direction, will be referred to respectively w-1-s, and w-1-90.

As a result of the two-dimensional discrete Fourier series approximation, the two-dimensional filter 4 set to have a square cross-sectional amplitude in the low band has a cross-sectional amplitude characteristic in the f ₂ direction in the low band similar to the sinc function, and the main amplitude range ( The main lobe area) is doubled to 1.89W. Therefore, in order to expand the phase setting range in the low frequency range, the amplitude characteristic resulting from the Fourier series approximation of the square amplitude characteristic in the f ₂ direction is used as the target amplitude characteristic in the low frequency range, and the phase setting range is | f ₂ | <1.89W.

  FIG. 5 is a diagram showing expansion (2) of the phase setting area. FIG. 5A is for w-2-s, and FIG. 5B is for w-2-90.

Speaker array length N ₂ D = 1.26 m, speaker spacing D = 0.07 m, filter order (N ₁ , N ₂ ) = (30, 18), focal length r ₀ = 5.0 m Design a super directional beam. At this time, the beam end angle φ _S = 7.18 °, the beam half-value angle φ _C = 5.71 °, and the sampling frequency f _S = 8635 Hz. However, the sampling frequency in the text is set to f _S = 5550 Hz for convenience in order to enlarge the figure.

6 to 9 show the frequency characteristics of the resulting amplitude and phase. FIG. 6 is a diagram showing the frequency characteristics (w-1-s) of the design result. FIG. 6A shows the frequency characteristic of the amplitude, and FIG. 6B shows the frequency characteristic of the phase. FIG. 7 is a diagram showing the frequency characteristics (w-1-90) of the design result. FIG. 7A shows the frequency characteristic of the amplitude, and FIG. 7B shows the frequency characteristic of the phase. FIG. 8 is a diagram illustrating the frequency characteristics (w-2-s) of the design result. FIG. 8A shows the frequency characteristic of the amplitude, and FIG. 8B shows the frequency characteristic of the phase. FIG. 9 is a diagram illustrating the frequency characteristics (w-2-90) of the design result. FIG. 9A shows the frequency characteristic of amplitude, and FIG. 9B shows the frequency characteristic of phase. In view of these amplitude characteristics, phi = the amplitude of at phi _S sees frequencies of -6dB amplitude of φ = 0 °, than _f L = 0.247 from either equation (9) (1370Hz) It is a little smaller, and the value of w-2-90 has the smallest value. As for the phase characteristics, w-2-s and w-2-90 show the difference in the phase transition depending on the direction φ from the lower range, and show the effect of expanding the phase setting range. Furthermore, w-1-90 and w-2-90 show a decrease in amplitude so as to be proportional to the frequency in the low frequency range, and a phase drop occurs correspondingly.

  FIG. 10 is a diagram illustrating the amplitude and phase characteristics (w-2-90) of the two-dimensional digital filter 4. FIG. 10A is a diagram showing amplitude characteristics, and FIG. 10B is a diagram showing phase characteristics. FIG. 11 is a diagram illustrating the directivity characteristics of w-2-90. FIG. 10 and FIG. 11 show the amplitude, phase characteristics, and amplitude directivity characteristics of the two-dimensional digital filter 4 for w-2-90, which has a large phase setting effect among the four types of design results.

Regarding the phase setting shown in FIG. 4 and FIG. 5, even when the phase is not placed in the non-physical region 5 | f ₂ |> ρ | f ₁ | The effect was slightly lower than that of the case where it was placed.

For the super-directional beam, the spatial response is obtained and the effect of the low-frequency phase expansion setting is compared. FIG. 12 is a diagram illustrating a spatial distribution of impulse train responses. FIG. 12A is a diagram obtained for w-2-s, and FIG. 12B is a diagram obtained for w-2-90. However, each response is corrected in magnitude so that the amplitude becomes 1 when x → ∞, and a portion having an amplitude of 0.5 or more is displayed. For reference, a speaker array arrangement (left side), φ = ± φ _S , ± φ _C radiation lines, and concentric circles centered on the focal point are shown superimposed. Both responses do not appear to converge at the focal position and are shaped like a wedge beam that grows with distance from the origin, but w-2-90 is slightly measured in the y direction at the focal position. Narrow width.

FIG. 13 is a diagram showing a sinusoidal sound pressure distribution obtained for w-2-90. FIG. 13A is a diagram illustrating f ₁ = 0.1 (555 Hz). FIG. 13B is a diagram illustrating f ₁ = 0.2 (1110 Hz). FIG. 13C is a diagram illustrating f ₁ = 0.3 (1665 Hz). Figure 13 (d) _is a diagram showing a f 1 = 0.45 (2498Hz). This also displays a portion having an amplitude of 0.5 or more. At a high frequency, a narrow beam corresponding to the growth of the impulse train response shown in FIG. 12, that is, a super-directional beam, is formed, and the beam width increases as the frequency decreases.

FIG. 14 is a diagram showing a sinusoidal sound pressure distribution obtained for w-2-s. FIG. 14A is a diagram illustrating f ₁ = 0.1 (555 Hz). FIG. 14B is a diagram illustrating f ₁ = 0.2 (1110 Hz). Compared to the sinusoidal sound pressure distribution shown in FIG. 14 obtained for w-2-s, there is almost no difference at medium and high frequencies, but the beam width of w-2-90 is narrower at low frequencies.

  Next, the quality of the super directional beam due to the difference in each phase setting is evaluated in connection with the degree of focus formation. Therefore, quantitative evaluation is performed using the result of the impulse train response. Focusing on the first, third and ninth (distances are 2 m, 5 m and 14 m, respectively) in the impulse train response shown in FIG. H5, H14), the width (W2, W5, W14) of the location of amplitude 0.5 measured in the y-axis direction and the sharpness H5 / W5 of the focal point. Actually, for comparison, the peak value, width, and focus for a total of six types including the case of setting according to FIG. 3 (this is set to w-0) and the case of no phase setting (this is set to phase-0). The sharpness was determined.

  They are shown in Tables 1 and 2. As can be seen from Tables 1 and 2, w-2-s and w-2-90 are particularly large for the peak value H5, and w-1-90 and w-2-90 are particularly small for the width W5. It has become. The resulting focus sharpness H5 / W5 is higher than not forming the focus, and the phase setting range can be expanded from the normal setting and set in a wide range as much as possible according to the amplitude setting. w-2-90 was the maximum.

FIG. 15 is a diagram showing the spatial distribution of the impulse train response of a super-directional beam whose beam center is tilted to φ ₀ = 60 ° using the focal position changing method based on the super-directional beam in the front direction of the array. FIG. 15A shows w-1-90, and FIG. 15B shows w-2-90. Although the beam width deteriorates to twice or more compared to the beam in the front direction, w-2-90 is formed as the narrowest beam among them.

  Attempt wideband design of super directional beam. For this purpose, instead of the amplitude characteristics of the two-dimensional fan filter described above, a one-dimensional filter of the blocking region ripple approximation is designed so as to have a large ripple amplitude only in the non-physical region 5 of the two-dimensional frequency plane. The amplitude characteristics synthesized by laying down the cross section are used. FIG. 16 is a diagram showing a method for setting the cross-sectional amplitude and phase for forming a broadband superdirective beam. In order to form a focused beam, the phase setting area is substantially the same as w-2-90 in FIG. 5B, but actually it is surrounded by a line A having an amplitude of 0 as shown in FIG. The main lobe area.

The superdirective beam was designed by adding the non-physical region amplitude to the specifications shown in the design of the superdirective beam. FIG. 17 is a diagram illustrating amplitude characteristics of the two-dimensional digital filter 4. FIG. 18 is a diagram showing the directivity characteristics of a superdirective beam. FIG. 19 is a diagram illustrating frequency characteristics of amplitude and phase. FIG. 19A shows the amplitude, and FIG. 19B shows the phase. In the present embodiment using the non-physical region 5, the lower band frequency f _L = 0.214 (1188 Hz), and from this, it can be seen by comparing the amplitude characteristics of FIG. 19 (a) and FIG. 9 (a). In addition, the band that gives the same directivity is wider in the lower range in this design method. It can also be seen from FIG. 18 and FIG. 19A that a substantially constant side lobe amount is given. Further, with regard to the phase characteristics, the present embodiment shows a greater change in phase due to the direction φ from a lower range.

  Therefore, the spatial distribution of the impulse train response was also obtained for this example, and the focal sharpness H5 / W5 calculated from the spatial distribution was 0.926. Therefore, the effect of wide band and wide phase setting is reflected in this numerical value.

  In addition to this, as a method for setting the phase, a case was also designed in which the phase is set over the entire physical region up to φ = 90 °. The resulting amplitude characteristics showed some reduction in side lobes, but the main lobe was slightly widened, and the phase characteristics were somewhat reduced in phase change at medium and high frequencies. As a result, the focal sharpness obtained from the impulse train response was reduced to 0.845.

  In the present embodiment, a method of forming a super-directional beam with improved characteristics using a focused beam forming method was proposed. In order to enhance the focus formation effect, the expansion of the phase setting range in the low frequency range is examined, and not only in the two-dimensional frequency range where the amplitude is set, that is, not only in the physical region 6 where the spectrum is distributed but also in the non-physical region 5. It was also found that setting the phase gave good results, and an example of application to a broadband design was also shown.

  We actually confirmed the formation of superdirective acoustic beams using an array speaker system. As can be seen from FIG. 1, the super-directional beam by the proposed method is rotationally symmetric with respect to the speaker array axis (y-axis). . In addition, there is a feature that the electronic change of the beam direction is easy. FIG. 20 is a diagram illustrating an application example of a superdirective acoustic beam to a pedestrian crossing. A super-directional beam having the width of the pedestrian crossing is irradiated to the entire pedestrian crossing by the speaker array 2 provided obliquely above the pedestrian crossing. Pedestrians trying to cross or cross a pedestrian crossing will hear the necessary information. FIG. 21 is a diagram illustrating an application example to an announcement for avoiding danger of a superdirective acoustic beam. A super-directional beam is emitted to alert a pedestrian who is about to cross the back of a car that is moving backward.

  FIG. 22 is a block diagram showing the configuration of the two-dimensional digital filter 4 in the directional array microphone 8. Although the directional array speaker 1 has been described in the present embodiment, the directional array microphone 8 which is another embodiment will be described. The output from the microphone array 7 constituting the directional array microphone 8 is input to the two-dimensional digital filter 4. The two-dimensional digital filter 4 represents an amplitude characteristic at a low frequency when a spectrum of an acoustic signal obtained by two-dimensional Fourier transform with respect to time and a microphone position is represented on a two-dimensional frequency plane including a time frequency axis and a spatial frequency axis. Is expanded to the non-physical region 5, so that a narrow and elongated beam having high directivity can be obtained. The signal output from the two-dimensional digital filter 4 can be output from a speaker or can be recorded on an audio recording medium.

It is a figure which shows the focus formation of the directional array speaker. It is a graph for demonstrating the relationship between the spectrum of a two-dimensional acoustic signal in a two-dimensional frequency plane, and a fan filter characteristic. It is a figure which shows the setting area | region of an amplitude and phase in the case of (phi) ₀ = 0 degree. It is a figure which shows expansion (1) of a phase setting area. It is a figure which shows expansion (2) of a phase setting area. It is a figure which shows the frequency characteristic (w-1-s) of a design result. It is a figure which shows the frequency characteristic (w-1-90) of a design result. It is a figure which shows the frequency characteristic (w-2-s) of a design result. It is a figure which shows the frequency characteristic (w-2-90) of a design result. It is a figure which shows the amplitude and phase characteristic (w-2-90) of the two-dimensional digital filter 4. FIG. It is a figure which shows the directional characteristic of w-2-90. It is a figure which shows the spatial distribution of an impulse train response. It is a figure which shows the sinusoidal sound pressure distribution calculated | required with respect to w-2-90. It is a figure which shows the sinusoidal sound pressure distribution calculated | required with respect to w-2-s. It is a figure which shows the spatial distribution of the impulse train response of the super-directional beam inclined to φ ₀ = 60 °. It is a figure which shows the setting method of the cross-sectional amplitude and phase for broadband superdirective beam formation. FIG. 4 is a diagram illustrating amplitude characteristics of the two-dimensional digital filter 4. It is a figure which shows the directivity characteristic of a super-directional beam. It is a figure which shows the frequency characteristic of an amplitude and a phase. It is a figure which shows the example of application to the pedestrian crossing of a super-directional acoustic beam.

It is a figure which shows the example of application to the announcement for avoiding the danger of a super-directional acoustic beam. 4 is a block diagram showing a configuration of a two-dimensional digital filter 4 in the directional array microphone 8. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Directional array speaker 2 Speaker array 3 One-dimensional digital filter 4 Two-dimensional digital filter 5 Non-physical region 6 Physical region 7 Microphone array 8 Directional array microphone

Claims (2)

A digital acoustic signal processing device for controlling a directivity characteristic by a two-dimensional digital filter and reproducing a digital acoustic signal input from a sound source by a speaker array including a plurality of speakers arranged along a linear direction,
The two-dimensional digital filter has a low-frequency amplitude characteristic when a spectrum of an acoustic signal obtained by two-dimensional Fourier transform with respect to time and a speaker position is represented on a two-dimensional frequency plane composed of a time frequency axis and a spatial frequency axis. A digital acoustic signal processing device characterized in that a cross section parallel to the spatial frequency axis is set as a single-peak passband extending from the physical region to the non-physical region. A digital acoustic signal processing apparatus for controlling directivity characteristics of a signal output from a microphone array composed of a plurality of microphones arranged along a straight line direction by a two-dimensional digital filter,
The two-dimensional digital filter has a low-frequency amplitude characteristic when the spectrum of an acoustic signal two-dimensionally Fourier transformed with respect to time and microphone position is represented on a two-dimensional frequency plane composed of a time frequency axis and a spatial frequency axis. A digital acoustic signal processing device characterized in that a cross section parallel to the spatial frequency axis is set as a single-peak passband extending from the physical region to the non-physical region.

Images