Speaker system

Abstract

A speaker system (100) comprises a plurality of speakers, and a processing unit (1000) that performs signal processing on a sound source signal inputted from a sound source (10). The plurality of speakers include a first speaker (101), and a second speaker (102) positioned behind the first speaker (101). The processing unit (1000) is provided with a first signal processing unit (1100) that generates a first reproduction signal by performing signal processing on the sound source signal, and a second signal processing unit (1200) that generates a second reproduction signal by performing signal processing on the sound source signal. The first signal processing unit (1100) generates a first reproduction signal and the second signal processing unit (1200) generates a second reproduction signal so that reproduction sounds reproduced from the plurality of speakers achieve a first target characteristic at a first control point and so that the reproduction sounds achieve a second target characteristic at a second control point.

Description

Speaker system

[0001] The present disclosure relates to a speaker system.

[0002] A directional speaker has been used for a speaker that reproduces sound in a specific direction.

[0003] Known examples of directional speakers include parabolic speakers installed on the ceiling of an art gallery or museum so that only people standing in front of an exhibit can hear guide sounds, and line array speakers that divide a large space such as a music hall or baseball stadium into multiple areas and play optimal sound for each area. With such directional speakers, a common method is to acoustically control the directionality by combining multiple speakers. Recently, speaker systems have also been put to practical use in which the directionality is controlled by individually processing signals for each speaker to emit sound from multiple speakers.

[0004] As a conventional technique for a speaker system that controls directivity using such signal processing, for example, Patent Document 1 discloses a local sound reproduction device that uses a speaker array in which multiple speakers are arranged in a line, and that drives each speaker with a drive signal based on a sound source and a spatial filter function. Also, Patent Document 2 discloses a loudspeaker system that includes a loudspeaker source that radiates loudspeaker sound, a control sound source that radiates control sound, and signal processing means that generates a control sound signal for forming an acoustic space with desired directivity and provides the control sound signal to the control sound source.

[0005] JP2015-231087A JP11-127494A

[0006] The present disclosure provides a speaker system having directionality that can improve sound quality while suppressing an increase in size.

[0007] A speaker system according to one aspect of the present disclosure includes a plurality of speakers arranged side by side so that their diaphragms are positioned coaxially about an axis extending in a front-to-rear direction, and a processing unit that performs signal processing on a sound source signal input from a sound source, wherein the plurality of speakers include a first speaker that emits sound forward and a second speaker that is positioned behind the first speaker, and the first speaker has a first horn that has an opening in front of it, and the processing unit includes a first signal processing unit that performs signal processing on the sound source signal to generate a first reproduction signal and input the first reproduction signal to the first speaker, and a second signal processing unit that performs signal processing on the sound source signal to generate a second reproduction signal and input the second reproduction signal to the second speaker, and the first speaker performs signal processing on the first reproduction signal to generate a second reproduction signal. the second speaker emits sound based on the second reproduction signal, the first signal processing unit generates the first reproduction signal and the second signal processing unit generates the second reproduction signal so that, when a first control point is a position that is a certain distance in front of and in front of a reference position on the axis based on the position of the first speaker and a second control point is a position that is away from the reference position in a direction different from the direction toward the first control point, the reproduced sound based on the sound source signals reproduced from the plurality of speakers realizes a first target characteristic in a predetermined frequency band at the first control point and the reproduced sound realizes a second target characteristic in the predetermined frequency band at the second control point, and the second target characteristic is a characteristic such that the level of the reproduced sound is lower than the first target characteristic.

[0008] According to the present disclosure, it is possible to provide a speaker system having directionality that can improve sound quality while suppressing an increase in size.

[0009] FIG. 1A is a block diagram showing the configuration of a speaker system according to Embodiment 1 in a service mode. FIG. 1B is a plan view showing the appearance of the speaker system according to Embodiment 1. FIG. 2A is a diagram showing frequency characteristics before and after control of the reproduced sound at a first control point of the speaker system according to Embodiment 1. FIG. 2B is a diagram showing frequency characteristics after control of the reproduced sound at the first control point and the second control point of the speaker system according to Embodiment 1. FIG. 3 is a diagram showing an example of the directional characteristics of the reproduced sound of the speaker system according to Embodiment 1. FIG. 4A is a block diagram showing the configuration of the speaker system according to Embodiment 1 in a coefficient design mode. FIG. 4B is a diagram for explaining the placement of microphones in the speaker system according to Embodiment 1 in the coefficient design mode. FIG. 5A is a diagram showing another example of the position of the second control point in the speaker system according to Embodiment 1. FIG. 5B is a diagram showing another example of the position of the second microphone in the speaker system according to Embodiment 1. FIG. 6 is a plan view showing the appearance of a speaker system according to Variation 1 of Embodiment 1. FIG. 7 is a plan view showing the appearance of a speaker system according to Variation 2 of Embodiment 1. FIG. 8 is a plan view showing the appearance of a speaker system according to Modification 3 of Embodiment 1. FIG. 9 is a block diagram showing the configuration of a speaker system according to Modification 4 of Embodiment 1 in coefficient design mode. FIG. 10A is a block diagram showing the configuration of a speaker system according to Embodiment 2 in service mode. FIG. 10B is a plan view showing the appearance of a speaker system according to Embodiment 2. FIG. 11A is a diagram showing frequency characteristics before and after control of reproduced sound at a first control point of the speaker system according to Embodiment 2. FIG. 11B is a diagram showing frequency characteristics after control of reproduced sound at a first control point and a second control point of the speaker system according to Embodiment 2. FIG. 11C is a diagram showing frequency characteristics before and after control of reproduced sound at a third control point of the speaker system according to Embodiment 2. FIG. 12 is a diagram showing an example of the directional characteristics of reproduced sound of the speaker system according to Embodiment 2. FIG. 13A is a block diagram showing the configuration of a speaker system according to Embodiment 2 in coefficient design mode. FIG. 13B is a diagram for explaining the arrangement of microphones in the speaker system according to Embodiment 2 in coefficient design mode.Fig. 14 is a block diagram showing a configuration of a speaker system according to Modification 1 of Embodiment 2 in a coefficient design mode. Fig. 15 is a block diagram showing a configuration of a speaker system according to Modification 2 of Embodiment 2 in a coefficient design mode.

[0010] (How One Aspect of the Present Disclosure Was Obtained) Before describing embodiments of the present disclosure in detail, how one aspect of the present disclosure was obtained will be described.

[0011] As a speaker system that controls directivity using signal processing, the local sound reproduction device described in Patent Document 1 uses a linear speaker array in which multiple speakers are arranged in a line. The linear speaker array is composed of multiple speakers (64 in the example of Patent Document 1) arranged at regular intervals in a straight line in the x-axis direction (see, for example, Figure 1 of Patent Document 1).

[0012] In addition, in the local sound reproduction device described in Patent Document 1, a drive device, for example, made up of a computer, creates a drive signal based on a sound source S(ω) provided from a sound source unit such as an appropriate storage medium (memory, HDD, optical disk) or a network. Each speaker is then driven by the drive signal created by the drive device.

[0013] More specifically, the driving device models the sound pressure at a listening position at a predetermined distance from the linear speaker array using a rectangular window in which the bright area where the sound source is heard is "1" and the dark area where the sound source is not heard is "0", and performs a spatial Fourier transform in the x-axis direction to generate a spatial filter function F ｉ Then, the driving device calculates the spatial filter function F ｉ By multiplying the signal S(ω) by the sound source S(ω), a drive signal for each speaker is generated to drive each speaker. In this way, the local sound reproduction device described in Patent Document 1 is capable of realizing local sound reproduction, in which only specific listeners can hear the sound source, and other listeners cannot hear it.

[0014] However, the technology described in Patent Document 1 requires a configuration in which a large number of speakers are arranged in a line in a direction perpendicular to the direction in which the sound is emitted, which makes the device large and expensive.

[0015] In contrast to this, the technology described in Patent Document 2 achieves directivity using two speakers: a loudspeaker source and a control source.

[0016] The public address system described in Patent Document 2 includes a public address source, a control sound source, an acoustic signal source, and signal processing means (see, for example, Figure 1 of Patent Document 2). The public address source converts a signal from the acoustic signal source into a public address sound and radiates it. On the other hand, the control sound source converts a signal from the signal processing means into a control sound and radiates it. The public address source and the control sound source are attached coaxially and in opposite directions. The signal processing means generates a control sound signal by performing signal processing on the amplitude, phase, etc. of the acoustic signal from the acoustic signal source. The signal processing means has an adaptive filter whose coefficients are set by performing a coefficient update calculation using an LMS (Least Mean Square) algorithm (least squares method) or the like so that an error input is always minimized (see, for example, Figure 7 of Patent Document 2).

[0017] The loudspeaker system described in Patent Document 2 uses signal processing with two speakers, a loudspeaker sound source and a control sound source, arranged coaxially and facing in opposite directions, thereby achieving a directional characteristic in which the loudspeaker level is high in the front direction of the loudspeaker sound source and low in the rear direction of the loudspeaker sound source, i.e., in the front direction of the control sound source.

[0018] However, in the technology described in Patent Document 2, signal processing is performed to control the reproduced sound at one control point (the position of the error detector, see FIG. 6 of Patent Document 2, etc.) using one speaker (control sound source), so it is unclear how the signal processing affects the acoustic characteristics of points other than the one control point. For example, the acoustic characteristics in the front direction of the loudspeaker sound source may differ from the acoustic characteristics before control, i.e., the acoustic characteristics of the original loudspeaker sound source. As a result, a problem may arise in which the sound quality in the front direction of the loudspeaker sound source deteriorates.

[0019] Furthermore, when considering typical speaker usage, listeners are often located in front of the speaker, not behind it, and directivity such as that described in Patent Document 1, which requires sound to be heard only in an area where listeners are located in front of the speaker, is often required. When limiting the listening area in this way, it is usually necessary to set a playback area in front of the speaker so that listeners in front of the speaker outside that area cannot hear the sound. To achieve this goal, the technology described in Patent Document 2 has difficulty reducing the playback sound level at the position of a single control point without affecting the sound quality in the front direction, even if the position of one control point is set in front of a speaker other than the speaker in front of the speaker. On the other hand, while the technology described in Patent Document 1 can achieve this goal, as described above, it requires a large number of speakers, and the signal processing calculation volume and memory capacity required for the signal processing calculation are large, resulting in a system that consumes a lot of power, is large, and is expensive.

[0020] The present disclosure is intended to solve these problems and provides a speaker system having directionality that can improve sound quality while suppressing an increase in size.

[0021] (Summary of the Present Disclosure) As an overview of the present disclosure, an example of a speaker system according to the present disclosure will be described below.

[0022] For example, a speaker system according to a first aspect of the present disclosure includes a plurality of speakers arranged side by side so that their diaphragms are positioned coaxially with an axis extending in a front-rear direction as an axis, and a processing unit that performs signal processing on a sound source signal input from a sound source, wherein the plurality of speakers include a first speaker that emits sound forward and a second speaker that is positioned behind the first speaker, and the first speaker has a first horn that has an opening in front of it, and the processing unit includes a first signal processing unit that performs signal processing on the sound source signal to generate a first playback signal and input the first playback signal to the first speaker, and a second signal processing unit that performs signal processing on the sound source signal to generate a second playback signal and input the second playback signal to the second speaker, and the first speaker is a first control point is a position a certain distance in front of and in front of a reference position on the axis based on the position of the first speaker, and a second control point is a position away from the reference position in a direction different from the direction toward the first control point; the first signal processing unit generates the first reproduction signal and the second signal processing unit generates the second reproduction signal so that the reproduced sound based on the sound source signals reproduced from the multiple speakers achieves a first target characteristic in a predetermined frequency band at the first control point and the reproduced sound achieves a second target characteristic in the predetermined frequency band at the second control point; and the second target characteristic is a characteristic such that the level of the reproduced sound is lower than the first target characteristic.

[0023] This allows the sound reproduced from the two speakers, the first speaker and the second speaker, to be reproduced according to the first target characteristic at the first control point in front of the first speaker, while the level of the reproduced sound is reduced according to the second target characteristic at the second control point located other than in front of the first speaker, thereby achieving directivity of the reproduced sound. Furthermore, controlling the reproduced sound using the sound reproduced from the two speakers is particularly effective for low-frequency components that are prone to sound interference, while the acoustic effect of the first horn can enhance the directivity of the reproduced sound for high-frequency components. Therefore, the speaker system according to this aspect can achieve directivity of the reproduced sound over a wide frequency band from low to high frequencies while improving the quality of the reproduced sound at the first control point and suppressing an increase in size.

[0024] Also, for example, a speaker system according to a second aspect of the present disclosure is a speaker system according to the first aspect, wherein the second control point is located forward of the first speaker and at a certain distance from the reference position in a direction forming a predetermined angle with respect to the axis.

[0025] This allows the reproduced sound to have directivity at the front side of the speaker system.

[0026] Also, for example, a speaker system according to a third aspect of the present disclosure is the speaker system according to the first aspect, wherein the second control point is located behind the second speaker.

[0027] This makes it possible to lower the level of the reproduced sound at the second control point behind the speaker system, so that even if there is a wall or ceiling near the rear of the speaker system, the reproduced sound can be prevented from reflecting off the wall or ceiling and affecting the quality of the sound reproduced in front of the speaker system.

[0028] Furthermore, for example, a speaker system according to a fourth aspect of the present disclosure is the speaker system according to any one of the first to third aspects, wherein the first signal processing unit includes a low-frequency extraction crossover filter that extracts low-frequency components below a predetermined frequency from the sound source signal, a high-frequency extraction crossover filter that extracts high-frequency components above a predetermined frequency from the sound source signal, a level adjuster that adjusts the level of an output signal from the high-frequency extraction crossover filter, a delay unit that delays the output signal from the level adjuster, a first FIR filter that performs convolution processing on the output signal from the low-frequency extraction crossover filter, and a playback signal adder that generates the first playback signal by adding the output signal from the delay unit and the output signal of the first FIR filter, and the second signal processing unit includes a second FIR filter that generates the second playback signal by performing convolution processing on the output signal from the low-frequency extraction crossover filter.

[0029] This allows the signal processing load of the first signal processing unit and the second signal processing unit to be reduced by the low-frequency extraction crossover filter and the high-frequency extraction crossover filter, while the sound quality of the reproduced sound can be improved by the level adjuster and delay device.

[0030] Furthermore, for example, a speaker system according to a fifth aspect of the present disclosure is the speaker system according to the first or second aspect, wherein the plurality of speakers includes a third speaker located behind the second speaker and emitting sound rearward, the processing unit has a third signal processing unit that processes the sound source signal to generate a third reproduction signal and inputs the third reproduction signal to the third speaker, the third speaker emits sound based on the third reproduction signal, and when a position behind the third speaker is set as a third control point, the reproduced sound is The first signal processing unit generates the first reproduction signal, the second signal processing unit generates the second reproduction signal, and the third signal processing unit generates the third reproduction signal so that the first target characteristic is realized in the specified frequency band at the first control point, the reproduced sound realizes the second target characteristic in the specified frequency band at the second control point, and the reproduced sound realizes the third target characteristic in the specified frequency band at the third control point, and the third target characteristic is a characteristic such that the level of the reproduced sound is lower than the first target characteristic.

[0031] This makes it possible to reduce the level of the reproduced sound at the third control point behind the speaker system, so that even if there is a wall or ceiling near the rear of the speaker system, the reproduced sound can be prevented from reflecting off the wall or ceiling and affecting the quality of the sound reproduced in front of the speaker system.

[0032] Furthermore, for example, a speaker system according to a sixth aspect of the present disclosure is the speaker system according to the fifth aspect, wherein the first signal processing unit includes a low-frequency extraction crossover filter that extracts low-frequency components below a predetermined frequency from the sound source signal, a high-frequency extraction crossover filter that extracts high-frequency components above a predetermined frequency from the sound source signal, a level adjuster that adjusts the level of an output signal from the high-frequency extraction crossover filter, a delay unit that delays the output signal from the level adjuster, and a delay circuit that adjusts the output signal from the low-frequency extraction crossover filter. the second signal processing unit includes a second FIR filter that performs convolution processing on the output signal from the low-frequency extraction crossover filter to generate the second reproduction signal, and the third signal processing unit includes a third FIR filter that performs convolution processing on the output signal from the low-frequency extraction crossover filter to generate the third reproduction signal.

[0033] This allows the signal processing load of the first signal processing unit, the second signal processing unit, and the third signal processing unit to be reduced by the low-frequency extraction crossover filter and the high-frequency extraction crossover filter, while the sound quality of the reproduced sound can be improved by the level adjuster and delay device.

[0034] Also, for example, a speaker system according to a seventh aspect of the present disclosure is a speaker system according to the fifth or sixth aspect, in which at least one of the following is satisfied: (i) the second target characteristic is a characteristic that minimizes the level of the reproduced sound at the second control point; and (ii) the third target characteristic is a characteristic that minimizes the level of the reproduced sound at the third control point.

[0035] This makes it possible to reduce the level of the reproduced sound at at least one of the second control point and the third control point.

[0036] Also, for example, a speaker system according to an eighth aspect of the present disclosure is a speaker system according to the fifth or sixth aspect, in which the third target characteristic is a characteristic in which the frequency characteristic is lowered by a certain level compared to the first target characteristic.

[0037] This makes it possible to reduce the level of the reproduced sound evenly across a predetermined frequency band at the third control point.

[0038] Also, for example, a speaker system according to a ninth aspect of the present disclosure is the speaker system according to any one of the fifth to eighth aspects, wherein the third speaker has a horn with an opening at the rear.

[0039] This makes it possible to prevent the high frequency components of the sound emitted by the third speaker from leaking to the front of the speaker system.

[0040] Furthermore, for example, a speaker system according to a tenth aspect of the present disclosure is the speaker system according to any one of the first to ninth aspects, wherein the second speaker emits sound forward.

[0041] This reduces the level of sound emitted from the second speaker at the rear of the speaker system.

[0042] Also, for example, a speaker system according to an eleventh aspect of the present disclosure is the speaker system according to the tenth aspect, wherein the second speaker has a horn having an opening in front thereof.

[0043] This makes it possible to further reduce the level of the high-frequency components of the sound emitted from the second speaker at the rear of the speaker system.

[0044] Furthermore, for example, a speaker system according to a twelfth aspect of the present disclosure is the speaker system according to any one of the first to ninth aspects, wherein the second speaker emits sound rearward.

[0045] This allows the sound output position of the first speaker and the sound output position of the second speaker to be spaced apart without increasing the size of the speaker system.

[0046] Furthermore, for example, a speaker system according to a thirteenth aspect of the present disclosure is the speaker system according to the twelfth aspect, wherein the second speaker has a horn having an opening at the rear.

[0047] This makes it possible to prevent the high frequency components of the sound emitted by the second speaker from leaking to the front of the speaker system.

[0048] Also, for example, a speaker system according to a fourteenth aspect of the present disclosure is a speaker system according to any one of the first to thirteenth aspects, wherein the first target characteristic is a characteristic that improves the level of low frequencies below a predetermined frequency of the reproduced sound at the first control point and flattens the frequency characteristics.

[0049] This makes it possible to improve the quality of the sound reproduced at the first control point.

[0050] Also, for example, a speaker system according to a fifteenth aspect of the present disclosure is a speaker system according to any one of the first to fourteenth aspects, in which the second target characteristic is a characteristic in which the frequency characteristic is lowered by a certain level compared to the first target characteristic.

[0051] This makes it possible to reduce the level of the reproduced sound evenly across a predetermined frequency band at the second control point.

[0052] Hereinafter, embodiments will be described with reference to the drawings as appropriate. However, more detailed explanations than necessary may be omitted. For example, detailed explanations of well-known matters or duplicate explanations of substantially identical configurations may be omitted. This is to avoid unnecessary redundancy in the following explanation and to facilitate understanding by those skilled in the art. Furthermore, each figure is a schematic diagram and is not necessarily an exact illustration. Furthermore, in each figure, substantially identical configurations are assigned the same reference numerals, and duplicate explanations may be omitted or simplified.

[0053] The embodiments described below each illustrate a preferred specific example of the present disclosure. The components, the arrangement and connection of the components, the order of operations, and the like shown in the following embodiments are merely examples. They are not intended to limit the subject matter described in the claims.

[0054] Therefore, among the components in the following embodiments, components that are not recited in the independent claims will be described as optional components that constitute more preferred embodiments.

[0055] Furthermore, in this specification, the terms "front," "rear," and "front-rear direction" are used to define the direction in which sound is emitted in a speaker system and the relative direction in which multiple speakers are lined up. Specifically, the direction in which a first speaker emits sound is defined as "front," and the direction opposite to the direction in which the first speaker emits sound is defined as "rear." Furthermore, the "front-rear direction" refers to the direction in which multiple speakers are lined up. The terms "front," "rear," and "front-rear direction" do not limit the position of a speaker system according to the present disclosure during manufacture or use.

[0056] Furthermore, in this specification, unless otherwise specified, ordinal numbers such as "first" and "second" do not refer to the number or order of components, but are used to avoid confusion between and distinguish between components of the same type. Furthermore, the order of the explanation does not necessarily match the order of the ordinal numbers.

[0057] (Embodiment 1) [Configuration] First, the configuration of a speaker system according to embodiment 1 will be described using Figures 1A and 1B. Figure 1A is a block diagram showing the configuration of the speaker system according to embodiment 1. Figure 1B is a plan view showing the appearance of the speaker system according to embodiment 1. Figure 1B shows a plan view of speaker system 100 as viewed from a direction perpendicular to axis L. In Figure 1B and the plan views of embodiment 1 and each of the modifications below, diaphragms 101a and 102a of speaker units housed inside housings 101b and 102b, respectively, are shown by dashed lines.

[0058] 1A and 1B , speaker system 100 according to the first embodiment includes a plurality of speakers, namely, first speaker 101 and second speaker 102, and processing unit 1000. Speaker system 100 is a directional speaker, and has directionality for reproduced sound. For example, speaker system 100 can reproduce sound at a high sound pressure level at first control point CP1, while making the reproduced sound difficult to hear at second control point CP2.

[0059] The first speaker 101 and the second speaker 102 are arranged in a straight line at a fixed distance from each other. More specifically, as shown in Fig. 1B, the first speaker 101 and the second speaker 102 are arranged side by side in the front-to-rear direction so that their respective diaphragms (diaphragms 101a and 102a described below) are positioned coaxially about an axis L. The axis L is an imaginary straight line extending in the front-to-rear direction.

[0060] The first speaker 101 includes a diaphragm 101a, a housing 101b, and a horn 101c, which are parts of a speaker unit. The horn 101c is an example of a first horn. In addition to the diaphragm 101a, the speaker unit of the first speaker 101 includes, for example, a support member for supporting the diaphragm 101a, a magnet, a voice coil, and the like (not shown).

[0061] The diaphragm 101a faces forward, and the first speaker 101 emits sound forward as the diaphragm 101a vibrates. The diaphragm 101a is disposed so that an axis L passes through the center thereof.

[0062] The housing 101b is, for example, a cylindrical housing whose axial direction is the extension direction of the axis line L. An opening in which the diaphragm 101a is installed is provided at the front end of the housing 101b, and the opening is closed by the diaphragm 101a. In addition, the rear end of the housing 101b is closed by a plate-like member.

[0063] The horn 101c is disposed in front of the diaphragm 101a and has an opening at the front. That is, the horn 101c has an opening in the direction in which the first speaker 101 emits sound. The horn 101c functions to provide directivity to the sound emitted from the first speaker 101, primarily in the high frequencies. That is, the sound emitted from the first speaker 101 has a characteristic in which low to high frequencies are reproduced evenly at the first control point CP1 located directly in front of the first speaker 101. However, at a position such as the second control point CP2, which is a certain angle away from the first control point CP1 when viewed from the reference position SP1, the high frequency level is reduced relative to the low frequency level. The reference position SP1 is a position on the axis L based on the position of the first speaker 101. In the example shown in FIG. 1B, the reference position SP1 is the center of the opening of the horn 101c, but this is not limited thereto. The reference position SP1 may be, for example, a position between the center of the opening of the horn 101c and the diaphragm 102a of the second speaker 102.

[0064] Here, the first control point CP1 is located a certain distance r in front of the reference position SP1. The second control point CP2 is located in a direction different from the direction from the reference position SP1 toward the first control point CP1. In the example shown in FIG. 1B , the second control point CP2 is located forward of the first speaker 101 and a certain distance r from the reference position SP1 in a direction that forms a certain angle θ (0 degrees < θ ≦ 90 degrees) with respect to the axis L. The certain angle θ can also be said to be the angle between the first control point CP1 and the second control point CP2 as viewed from the reference position SP1. Furthermore, the second control point CP2 is located, for example, outside the space enclosed by the extension of the horn 101c when it is virtually extended forward. In this case, the certain angle θ is greater than the angle at which the inner surface of the horn 101c is inclined with respect to the axis L. The distance from the reference position SP1 to the second control point CP2 is, for example, the same as the distance r from the reference position SP1 to the first control point CP1, but may be different from the distance r.

[0065] The second speaker 102 is located behind the first speaker 101. The second speaker 102 has a diaphragm 102a, which is part of a speaker unit, and a housing 102b. In addition to the diaphragm 102a, the speaker unit of the second speaker 102 includes, for example, a support member for supporting the diaphragm 102a, a magnet, a voice coil, etc. (not shown). In the example shown in FIG. 1B , the second speaker 102 is integrated with the first speaker 101 by a connecting rod, and a space 110 for the second speaker 102 to emit sound is formed in front of the diaphragm 102a of the second speaker 102, i.e., between the first speaker 101 and the second speaker 102.

[0066] The diaphragm 102a faces forward, and the second speaker 102 emits sound forward as the diaphragm 102a vibrates. The diaphragm 102a is positioned so that the axis L passes through its center. Therefore, the diaphragm 101a of the first speaker 101 and the diaphragm 102a of the second speaker 102 are positioned coaxially with the axis L as their axis.

[0067] The housing 102b is, for example, a cylindrical housing whose axial direction is the extension direction of the axis line L. An opening in which the diaphragm 102a is installed is provided at the front end of the housing 102b, and the opening is closed by the diaphragm 102a. In addition, the rear end of the housing 102b is closed by a plate-like member.

[0068] As shown in FIG. 1A , the processing unit 1000 processes a sound source signal input from the sound source 10. The processing unit 1000 controls the characteristics of the reproduced sound based on the sound source signal at the first control point CP1 and the second control point CP2 by processing the sound source signal. The processing unit 1000 includes a first signal processing unit 1100 and a second signal processing unit 1200. The processing unit 1000 is a processing circuit realized, for example, by one or more memories that store programs and one or more processors that execute the programs. The processing unit 1000 may be a dedicated logic circuit that performs signal processing, which will be described later. The processing unit 1000 is formed on, for example, a substrate, and the substrate is housed in the housing 101b or the housing 102b, or another housing not shown.

[0069] The first signal processing unit 1100 generates a first reproduction signal by performing signal processing on the sound source signal from the sound source 10, and inputs the first reproduction signal to the first speaker 101. The first speaker 101 emits sound based on the first reproduction signal input from the first signal processing unit 1100. The second signal processing unit 1200 generates a second reproduction signal by performing signal processing on the sound source signal from the sound source 10, and inputs the second reproduction signal to the second speaker 102. The second speaker 102 emits sound based on the second reproduction signal input from the second signal processing unit 1200.

[0070] In speaker system 100, first signal processing unit 1100 generates a first reproduction signal and second signal processing unit 1200 generates a second reproduction signal so that reproduced sound based on a sound source signal reproduced from first speaker 101 and second speaker 102 achieves a first target characteristic in a predetermined frequency band at first control point CP1 and achieves a second target characteristic in a predetermined frequency band at second control point CP2. In speaker system 100, the reproduced sound from first speaker 101 and second speaker 102 is a composite sound of a sound emitted from first speaker 101 based on the first reproduction signal and a sound emitted from second speaker 102 based on the second reproduction signal. The predetermined frequency band may be a portion of the frequency band of the sound source signal (for example, a frequency band equal to or lower than a predetermined frequency) or the entire frequency band of the sound source signal.

[0071] The first signal processing unit 1100 includes a first FIR (Finite Impulse Response) filter 1110 (FIR1 in the figure), a low-frequency extraction crossover filter 1120, a high-frequency extraction crossover filter 1130, a level adjuster 1140, a delay unit 1150, and a reproduced signal adder 1160. The second signal processing unit 1200 includes a second FIR filter 1210 (FIR2 in the figure).

[0072] [Signal Processing] With continued reference to FIGS. 1A and 1B, the signal processing by the processing unit 1000 will be described in detail below.

[0073] The low-frequency extraction crossover filter 1120 extracts low-frequency components of the sound source signal below a predetermined frequency and inputs the extracted low-frequency components as output signals to the first FIR filter 1110 and the second FIR filter 1210. The high-frequency extraction crossover filter 1130 extracts high-frequency components of the sound source signal above a predetermined frequency and inputs the extracted high-frequency components as output signals to the level adjuster 1140. The extraction of low-frequency components by the low-frequency extraction crossover filter 1120 and the extraction of high-frequency components by the high-frequency extraction crossover filter 1130 are performed in parallel, for example. The low-frequency extraction crossover filter 1120 is, for example, a low-pass filter (LPF). The high-frequency extraction crossover filter 1130 is, for example, a high-pass filter (HPF).

[0074] Here, the cutoff frequencies of the low-frequency extraction crossover filter 1120 and the high-frequency extraction crossover filter 1130 are set, for example, taking into consideration the frequency characteristics of the directionality of the horn 101 c. For example, when comparing the sounds emitted from the first speaker 101 at the first control point CP1 and the second control point CP2, if the frequency characteristics of the directionality of the horn 101 c have a large drop in level (in other words, sound pressure) in the frequency band of 5 kHz or higher (for example, if the difference in sound level between the first control point CP1 and the second control point CP2 in the frequency band of 5 kHz or higher is 10 dB or higher), then the cutoff frequencies are set to 5 kHz.

[0075] The first FIR filter 1110 performs convolution processing with a coefficient on the output signal of the low-frequency extraction crossover filter 1120, which is the low-frequency component of the sound source signal extracted by the low-frequency extraction crossover filter 1120. The output signal of the first FIR filter 1110 passes through the reproduction signal adder 1160 to become a first reproduction signal, which is input to the first speaker 101.

[0076] The second FIR filter 1210 generates a second reproduction signal by convolving the output signal of the low-frequency extraction crossover filter 1120, which is the low-frequency component of the sound source signal extracted by the low-frequency extraction crossover filter 1120, with a coefficient. The generated second reproduction signal is input to the second speaker 102. The convolution process by the first FIR filter 1110 and the convolution process by the second FIR filter 1210 are performed in parallel, for example.

[0077] The first FIR filter 1110 and the second FIR filter 1210 are adaptive filters that are set to filter characteristics designed by, for example, adaptive filter design. The first FIR filter 1110 and the second FIR filter 1210 adjust the phase and amplitude of the input signal for each frequency. The first FIR filter 1110 and the second FIR filter 1210 multiply the phase and amplitude of the input signal for each frequency by coefficients and output the results.

[0078] The coefficients of the first FIR filter 1110 and the second FIR filter 1210 have characteristics for realizing the first target characteristic at the first control point CP1 and the second target characteristic at the second control point CP2 in the reproduced sound based on the sound source signal reproduced from the first speaker 101 and the second speaker 102. As a result, the reproduced sound from the first speaker 101 and the second speaker 102 can simultaneously realize the first target characteristic at the first control point CP1 and the second target characteristic at the second control point CP2.

[0079] The first target characteristic is, for example, a characteristic that improves the level of low frequencies below a predetermined frequency of the reproduced sound at the first control point CP1 and flattens the frequency characteristics. In other words, the first target characteristic is a characteristic that has a higher low frequency level and a flatter frequency characteristic than the characteristics of the reproduced sound before control. The second target characteristic is a characteristic that reduces the level of the reproduced sound more than the first target characteristic. For example, the second target characteristic is a characteristic in which the frequency characteristics are lowered by a certain level compared to the first target characteristic.

[0080] Here, the first target characteristic and the second target characteristic will be described while showing the results of measuring the frequency characteristics of the reproduced sound actually emitted by the speaker system 100. FIG. 2A is a diagram showing the frequency characteristics before and after control of the reproduced sound at the first control point. FIG. 2B is a diagram showing the frequency characteristics after control of the reproduced sound at the first control point and the second control point. In FIGS. 2A and 2B, the horizontal axis represents frequency, and the vertical axis represents level (sound pressure). The characteristics before control shown in FIG. 2A are the characteristics of the reproduced sound when reproduced only from the first speaker 101 without performing signal processing on the sound source signal. The characteristics after control shown in FIGS. 2A and 2B are the characteristics of the reproduced sound reproduced from the first speaker 101 and the second speaker 102 after signal processing by the processing unit 1000. Furthermore, the cutoff frequencies of the low-frequency extraction crossover filter 1120 and the high-frequency extraction crossover filter 1130 during this signal processing are 5 kHz.

[0081] As shown in FIG. 2A, by setting the first target characteristic shown by the dashed line to the characteristic of the reproduced sound before control at the first control point CP1, the characteristic of the reproduced sound after control shown by the solid line can be realized. Also, as shown in FIG. 2B, by setting the second target characteristic shown by the dashed line, which has a frequency characteristic 10 dB lower than the first target characteristic shown by the dashed line, the characteristic of the reproduced sound after control at the first control point CP1 shown by the solid line and the characteristic of the reproduced sound after control at the second control point CP2 shown by the dotted line can be realized. Controlling the reproduced sound using the first speaker 101 and the second speaker 102 is particularly effective for low-frequency components that are prone to sound interference. The first and second target characteristics are designed to cut off high frequencies above 5 kHz, which is the cutoff frequency of the low-frequency extraction crossover filter 1120 and the high-frequency extraction crossover filter 1130. However, the high-frequency characteristics are not particularly limited and can be set as appropriate. For example, the first target characteristic and the second target characteristic may be flat characteristics in the high frequency range.

[0082] The method for designing the coefficients of the first FIR filter 1110 and the second FIR filter 1210 will be explained in detail later.

[0083] The above has been a description of signal processing for the low frequency components of the sound source signal. Next, signal processing for the high frequency components of the sound source signal will be described.

[0084] 1A again, level adjuster 1140 adjusts the level of the output signal of high-frequency extraction crossover filter 1130, which is the high-frequency component of the sound source signal extracted by high-frequency extraction crossover filter 1130. Delay device 1150 delays the output signal of level adjuster 1140 and inputs it to playback signal adder 1160.

[0085] The reproduction signal adder 1160 generates a first reproduction signal by adding the output signal from the delay unit 1150 and the output signal from the first FIR filter 1110. The generated first reproduction signal is input to the first speaker 101, and sound based on the first reproduction signal resulting from the addition by the reproduction signal adder 1160 is reproduced from the first speaker 101. In other words, only the low-frequency components of the sound source signal are reproduced from the second speaker 102, while the entire frequency band components of the sound source signal are reproduced from the first speaker 101. As a result, due to the synergistic effect with the acoustic characteristics of the horn 101c, a reproduced sound with a high sound pressure level can be achieved over a wide frequency band from low to high at the first control point CP1. On the other hand, due to the effect of the acoustic characteristics of the horn 101c, the high-frequency characteristics are reduced at the second control point CP2, and the low-frequency characteristics are also reduced to achieve the second target characteristics. The low frequency characteristics have already been described above, but the effect of signal processing on the sound source signal by the processing unit 1000 will be described again with reference to FIGS. 2A and 2B.

[0086] 2A, the characteristics of the reproduced sound before control at the first control point CP1 indicated by the dashed line have a low level in the frequency band below 900 Hz, and in the frequency band above 900 Hz, the frequency characteristics are significantly disturbed, with noticeable peak and dip characteristics. In contrast, as a result of controlling the reproduced sound by setting the first target characteristics indicated by the dashed line, the characteristics of the reproduced sound after control, indicated by the solid line, show an improvement in the level in the frequency band below 900 Hz, and suppression of the peak and dip characteristics in the frequency band above 900 Hz.

[0087] Here, first FIR filter 1110 and second FIR filter 1210 do not process signals in the high frequency range above 5 kHz as in the first target characteristic, so the high frequency components added by playback signal adder 1160 achieve high frequency characteristics above 5 kHz. Level adjuster 1140 and delay unit 1150 perform signal processing so that the synthesis of the low frequency characteristics and high frequency characteristics resulting from this addition does not cause distortion in the frequency characteristics or phase. Level adjuster 1140 and delay unit 1150 are adjusted, for example, based on the results (e.g., impulse response) of reproducing only the low frequency components from first speaker 101, which have been acquired in advance.

[0088] Next, as shown in Figure 2B, when the reproduced sound is controlled by setting a second target characteristic shown by a two-dot chain line, which has a frequency characteristic 10 dB lower than the first target characteristic shown by a dotted line, the reproduced sound characteristic after control at the second control point CP2 shown by a dotted line is obtained compared to the reproduced sound characteristic after control at the first control point CP1 shown by a solid line. The difference in characteristics is roughly equal to the difference between the first target characteristic and the second target characteristic, achieving about 10 dB. Furthermore, even in the high frequency range above 5 kHz, the level of the reproduced sound at the second control point CP2 is significantly lower than the level of the reproduced sound at the first control point CP1 due to the acoustic effect of the horn 101c.

[0089] As described above, speaker system 100 controls the reproduced sound to achieve the first target characteristic at first control point CP1 and the second target characteristic at second control point CP2 as shown and explained in Figures 2A and 2B, thereby achieving an improvement in the low-frequency sound pressure level of the reproduced sound at first control point CP1, as well as improvements in the peak and dip characteristics of the reproduced sound, while achieving directionality of the reproduced sound over a wide frequency band from low to high frequencies.

[0090] Here, Fig. 3 shows the measurement results of the directional characteristics when the predetermined angle θ shown in Fig. 1B is set to 60 degrees. Fig. 3 is a diagram showing an example of the directional characteristics of the reproduced sound of the speaker system according to embodiment 1. The angles shown in Fig. 3 are angles seen from the center of the opening of horn 101c, when the direction of first control point CP1 (first control point direction) seen from the center of the opening of horn 101c is set to 0 degrees.

[0091] As shown in FIG. 3, if the frequency level in the 0-degree direction, which is the first control point direction, is taken as 0 dB, the levels in the +60-degree direction, which is the second control point direction, are -10 dB or less from low to high frequencies. The same characteristics are also obtained in the -60-degree direction. This is because, as shown in FIG. 1B, the diaphragm 101a of the first speaker 101 and the diaphragm 102a of the second speaker 102 are configured to be positioned coaxially, and the reproduced sound reproduced from each speaker is radiated concentrically, thereby achieving characteristics that are symmetrical with respect to the axis L passing through the center of the speaker system 100, i.e., the line passing through 0 degrees and 180 degrees in FIG. 3. This means that the directional characteristics shown in FIG. 3 are realized as a sphere that includes both horizontal and vertical planes.

[0092] Therefore, at the front of the speaker system 100, i.e., at the first control point CP1 in the 0 degree direction, a control effect is obtained for the reproduced sound such that the frequency characteristics are flattened and the level is increased over a wide frequency range from low to high. At the same time, at the ±60 degree direction of the speaker system 100 corresponding to the direction of the second control point CP2, a control effect is obtained for the reproduced sound such that the level is reduced by 10 dB or more from the frequency characteristics in the 0 degree direction over a wide frequency band.

[0093] The first target characteristic can be set to any characteristic that is optimal for the intended use, such as not only a characteristic that flattens the frequency characteristics as shown in Figures 2A and 2B, but also a characteristic that emphasizes the level of a certain frequency band, such as a characteristic that boosts the low range, or, conversely, a characteristic that lowers the level of a certain frequency band.

[0094] [FIR Filter Coefficient Design] Next, the design of the coefficients of the first FIR filter 1110 and the second FIR filter 1210 to obtain the above-described control effect will be described with reference to FIGS. 4A and 4B. FIG. 4A is a block diagram showing the configuration of the speaker system according to embodiment 1 in coefficient design mode. FIG. 4B is a diagram for explaining the placement of microphones in coefficient design mode. FIG. 4B shows a plan view of the speaker system 100 as viewed from a direction perpendicular to the axis L. The processing unit 1000 of the speaker system 100 has, for example, both the functional configuration shown in FIG. 1A for executing the service mode and the functional configuration shown in FIG. 4A for executing the coefficient design mode. The service mode is an operating mode for playing music for a listener. The coefficient design mode is an operating mode for determining the control characteristics of the first signal processing unit 1100 and the second signal processing unit 1200 (specifically, the coefficients of the first FIR filter 1110 and the second FIR filter 1210).

[0095] 4A and 4B, a first microphone 1701 and a second microphone 1702 are used to design the coefficients of the first FIR filter 1110 and the second FIR filter 1210. As shown in Fig. 4B, the first microphone 1701 is placed at the position of the first control point CP1 shown in Fig. 1B, and the second microphone 1702 is placed at the position of the second control point CP2 shown in Fig. 1B.

[0096] In the coefficient design mode, the processing unit 1000 includes a first signal processing unit 1100, a second signal processing unit 1200, a first target characteristic filter 1501 (first target characteristic F1 in the figure), a second target characteristic filter 1502 (second target characteristic F2 in the figure), a first error signal adder 1601, and a second error signal adder 1602. In the coefficient design mode, the first signal processing unit 1100 also includes a first FIR filter 1110, a first coefficient updater 1111 (LMS11 in the figure), a second coefficient updater 1112 (LMS12 in the figure), a first speaker characteristic correction filter 1121 (Fx11 in the figure), and a second speaker characteristic correction filter 1122 (Fx12 in the figure). In addition, in the coefficient design mode, the second signal processing unit 1200 includes a second FIR filter 1210, a fourth coefficient updater 1211 (LMS21 in the figure), a fifth coefficient updater 1212 (LMS22 in the figure), a fourth speaker characteristics correction filter 1221 (Fx21 in the figure), and a fifth speaker characteristics correction filter 1222 (Fx22 in the figure).

[0097] In the coefficient design mode, the first FIR filter 1110, the first speaker characteristic compensation filter 1121, and the second speaker characteristic compensation filter 1122 of the first signal processing unit 1100 each receive a sound source signal from the sound source 10 and perform convolution processing of the input sound source signal with a coefficient. The output signal of the first FIR filter 1110 is reproduced from the first speaker 101.

[0098] Similarly, in the coefficient design mode, the second FIR filter 1210, the fourth speaker characteristic compensation filter 1221, and the fifth speaker characteristic compensation filter 1222 of the second signal processing unit 1200 each receive the sound source signal from the sound source 10 and perform convolution processing of the input sound source signal with the coefficients. The output signal of the second FIR filter 1210 is reproduced from the second speaker 102.

[0099] Then, the reproduced sounds reproduced from the first speaker 101 and the second speaker 102 are detected by the first microphone 1701 and the second microphone 1702. The detection signal of the reproduced sounds detected by the first microphone 1701 is input to the first error signal adder 1601. The detection signal of the reproduced sounds detected by the second microphone 1702 is input to the second error signal adder 1602.

[0100] On the other hand, the sound source signal from sound source 10 is input to first target characteristic filter 1501, in which a first target characteristic is set as a control coefficient, and is subjected to signal processing together with the control coefficient in first target characteristic filter 1501. Also, the sound source signal from sound source 10 is input to second target characteristic filter 1502, in which a second target characteristic is set as a control coefficient, and is subjected to signal processing together with the control coefficient in second target characteristic filter 1502. The output signal of first target characteristic filter 1501 is input to first error signal adder 1601. The output signal of second target characteristic filter 1502 is input to second error signal adder 1602.

[0101] Then, first error signal adder 1601 adds the detection signal from first microphone 1701 and the output signal from first target characteristic filter 1501, and inputs the output signal generated by the addition to first coefficient updater 1111 and fourth coefficient updater 1211. As a result, the difference between the first target characteristic and the detection result of the reproduced sound by first microphone 1701 is input to first coefficient updater 1111 and fourth coefficient updater 1211. First error signal adder 1601 finds the difference by, for example, inverting the sign of one of the signals and then adding the detection signal from first microphone 1701 and the output signal from first target characteristic filter 1501. Similarly, second error signal adder 1602 adds the detection signal from second microphone 1702 and the output signal from second target characteristic filter 1502, and inputs the output signal generated by the addition to second coefficient updater 1112 and fifth coefficient updater 1212. As a result, the difference between the second target characteristic and the detection result of the reproduced sound by second microphone 1702 is input to second coefficient updater 1112 and fifth coefficient updater 1212. Second error signal adder 1602 finds the difference by, for example, inverting the sign of one of the signals and then adding the detection signal from second microphone 1702 and the output signal from second target characteristic filter 1502.

[0102] On the other hand, the first speaker characteristics correction filter 1121 has a coefficient set in advance that represents the transfer characteristic from the first speaker 101 to the first microphone 1701. The first speaker characteristics correction filter 1121 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the first coefficient updater 1111. Similarly, the second speaker characteristics correction filter 1122 has a coefficient set in advance that represents the transfer characteristic from the first speaker 101 to the second microphone 1702. The second speaker characteristics correction filter 1122 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the second coefficient updater 1112.

[0103] Here, for example, the first coefficient updater 1111 and the second coefficient updater 1112 use the LMS algorithm as described in Patent Document 2.

[0104] In this case, first coefficient updater 1111 uses the output signal from first speaker characteristic compensation filter 1121 and the output signal from first error signal adder 1601 to update the coefficients of first FIR filter 1110 so as to minimize the output signal from first error signal adder 1601. Also, second coefficient updater 1112 uses the output signal from second speaker characteristic compensation filter 1122 and the output signal from second error signal adder 1602 to update the coefficients of first FIR filter 1110 so as to minimize the output signal from second error signal adder 1602.

[0105] On the other hand, the fourth speaker characteristics correction filter 1221 has a coefficient set in advance that represents the transfer characteristic from the second speaker 102 to the first microphone 1701. The fourth speaker characteristics correction filter 1221 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the fourth coefficient updater 1211. Similarly, the fifth speaker characteristics correction filter 1222 has a coefficient set in advance that represents the transfer characteristic from the second speaker 102 to the second microphone 1702. The fifth speaker characteristics correction filter 1222 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the fifth coefficient updater 1212.

[0106] Here too, for example, the LMS algorithm as described in Patent Document 2 is used for the calculations of the fourth coefficient updater 1211 and the fifth coefficient updater 1212.

[0107] In this case, fourth coefficient updater 1211 uses the output signal from fourth speaker characteristic compensation filter 1221 and the output signal from first error signal adder 1601 to update the coefficients of second FIR filter 1210 so as to minimize the output signal from first error signal adder 1601. Furthermore, fifth coefficient updater 1212 uses the output signal from fifth speaker characteristic compensation filter 1222 and the output signal from second error signal adder 1602 to update the coefficients of second FIR filter 1210 so as to minimize the output signal from second error signal adder 1602.

[0108] The above-described processing designs the coefficients of the first FIR filter 1110 and the second FIR filter 1210. Furthermore, as necessary, the coefficients of the first FIR filter 1110 and the coefficients of the second FIR filter 1210 are repeatedly updated until a reproduced sound with the desired frequency characteristics is obtained.

[0109] The coefficients thus obtained are set in the first FIR filter 1110 and the second FIR filter 1210 in the service mode shown in Fig. 1A. In the service mode, music is played back for the listener.

[0110] In other words, when music is played for a listener, the above-described coefficient design has already been performed, and therefore the functional configuration of the processing unit 1000 during coefficient design and music playback differs as shown in FIG. 4A and FIG. 1A . In this manner, the processing unit 1000 operates in a coefficient design mode when determining the control characteristics of the first signal processing unit 1100 and the second signal processing unit 1200, and in a service mode when playing music for a listener, and changes the signal processing configuration according to each mode. For example, if the processing unit 1000 is configured with a processor and a memory, programs corresponding to both the service mode and the coefficient design mode are stored in the memory, and the processor executes the necessary program according to the mode to realize the function of each mode. Also, for example, if the processing unit 1000 is configured with a dedicated logic circuit, the processing unit 1000 has circuits corresponding to both the service mode and the coefficient design mode, and the necessary circuits operate according to the mode to realize the function of each mode.

[0111] [Another Example of the Position of the Second Control Point] Note that, in the example shown in Fig. 1B, the second control point CP2 is located forward of the first speaker 101, but this is not limiting and the second control point CP2 may be located at a position other than in front of the first speaker 101. Fig. 5A is a diagram showing another example of the position of the second control point. Fig. 5B is a diagram showing another example of the position of the second microphone. Figs. 5A and 5B show plan views of the speaker system 100 as viewed from a direction perpendicular to the axis L.

[0112] 5A, the second control point CP2 is located behind the second speaker 102. Therefore, in the coefficient design mode, the second microphone 1702 is also located behind the second speaker 102. Therefore, the signal processing operation by the processing unit 1000 is the same as above, but the position at which the second target characteristic is realized in the reproduced sound is behind the speaker system 100.

[0113] For example, when a directional characteristic such as that shown in Figure 3 is obtained, if the reproduced sound is not affected by the environment behind speaker system 100, for example, if there is no wall or ceiling near the rear of speaker system 100, then ignoring the decrease in the rear directional characteristic, where the level is higher behind the +90 degree and -90 degree directions as shown in Figure 3, and reducing the level in the direction of a predetermined angle θ in front of speaker system 100 (the ±60 degree direction in Figure 3), will allow a listener located in the 0 degree direction to experience reproduced sound of better sound quality.

[0114] However, if the influence of the environment behind speaker system 100 cannot be ignored, for example, if there is a wall or ceiling near the rear of speaker system 100, the high-level reproduced sound to the rear as shown in Fig. 3 may be reflected off the wall or ceiling, adversely affecting the sound quality of the reproduced sound experienced by a listener positioned in a 0-degree direction in front of speaker system 100. In that case, it is sufficient to install second microphone 1702 at the position shown in Fig. 5B and design the coefficients of first FIR filter 1110 and second FIR filter 1210 so that second control point CP2 is located behind speaker system 100 as shown in Fig. 5A.

[0115] In this way, taking into consideration the installation conditions of the speaker system 100, the position of the second control point CP2 (distance and angle relative to the first speaker 101) can be appropriately determined so as not to adversely affect listeners located directly in front of the speaker system 100.

[0116] (Variations of Embodiment 1) Next, variations of Embodiment 1 will be described. Note that in the following descriptions of the variations, differences between Embodiment 1 and each variation of Embodiment 1 will be mainly described, and descriptions of commonalities will be omitted or simplified. Furthermore, in each variation described below, the positions of second control point CP2 and second microphone 1702 are the positions shown in FIGS. 1B and 4B, but are not particularly limited thereto, and may be, for example, the positions shown in FIGS. 5A and 5B.

[0117] [Variation 1] Fig. 6 is a plan view showing the appearance of a speaker system according to Variation 1 of Embodiment 1. Fig. 6 shows a plan view of speaker system 100A as viewed from a direction perpendicular to axis L. In Fig. 6, the shape of speaker system 100A with horn 102c removed is shown by a solid line, and horn 102c is shown by a two-dot chain line.

[0118] As shown in Fig. 6 , speaker system 100A according to Variation 1 of Embodiment 1 differs from speaker system 100 according to Embodiment 1 mainly in that speaker system 100A includes second speaker 112 instead of second speaker 102. Second speaker 112 includes a horn 102c in addition to the configuration of second speaker 102. In the example shown in Fig. 6 , second speaker 112 emits sound forward.

[0119] Horn 102c is installed in front of diaphragm 102a and has an opening at the front. That is, horn 102c has an opening in the direction in which second speaker 112 emits sound. Horn 102c serves to impart directionality to the sound emitted from second speaker 112, mainly in the high frequency range. That is, the sound emitted from second speaker 112 has a characteristic in which the sound is reproduced evenly from low to high frequencies in front of speaker system 100A, but has a characteristic in which the level of the high frequency range is reduced relative to the low frequency range in the rear of speaker system 100A.

[0120] Since the second speaker 112 has the horn 102c, the level of the high-frequency components of the sound emitted from the second speaker 112 at the rear of the speaker system 100A can be reduced, thereby suppressing the deterioration of the rear directional characteristics as shown in Figure 3.

[0121] [Modification 2] Fig. 7 is a plan view showing the appearance of a speaker system according to Modification 2 of Embodiment 1. Fig. 7 shows a plan view of speaker system 100B as seen from a direction perpendicular to axis L.

[0122] 7 , speaker system 100B according to Variation 2 of Embodiment 1 differs from speaker system 100 according to Embodiment 1 mainly in that second speaker 102 emits sound rearward. In speaker system 100B, diaphragm 102a of second speaker 102 is oriented rearward, and second speaker 102 emits sound rearward by vibrating diaphragm 102a. Note that if the orientation of diaphragm 102a of second speaker 102 changes, the transfer characteristics from second speaker 102 to first control point CP1 (= position of first microphone 1701) and the transfer characteristics from second speaker 102 to second control point CP2 (= position of second microphone 1702) change. Therefore, the coefficients of the first FIR filter 1110 and the second FIR filter 1210 are designed to have characteristics according to the orientation of the diaphragm 102a of the second speaker 102. The design of the coefficients of the first FIR filter 1110 and the second FIR filter 1210 is as described above with reference to FIGS. 4A and 4B . By having the second speaker 102 emit sound backward, the difference between the distance from the sound output position of the first speaker 101 to the second control point CP2 and the distance from the sound output position of the second speaker 102 to the second control point CP2 can be increased without increasing the size of the speaker system 100B. This makes it easier to reduce the level of the reproduced sound based on the sound source signal at the second control point CP2.

[0123] [Variation 3] Fig. 8 is a plan view showing the appearance of a speaker system according to Variation 3 of Embodiment 1. Fig. 8 shows a plan view of speaker system 100C as seen from a direction perpendicular to axis L.

[0124] As shown in FIG. 8 , speaker system 100C according to Variation 3 of Embodiment 1 differs from speaker system 100 according to Embodiment 1 mainly in that it includes second speaker 112 that emits sound rearward, instead of second speaker 102 that emits sound forward. In speaker system 100C, diaphragm 102a of second speaker 112 faces rearward, and second speaker 112 emits sound rearward by vibrating diaphragm 102a. In speaker system 100C, horn 102c is installed behind diaphragm 102a and has an opening at the rear. In speaker system 100C, sound emitted from second speaker 112 has a characteristic of reproducing sounds from low to high frequencies evenly at the rear of speaker system 100C, but has a characteristic of lowering the high frequency level relative to the low frequency at the front of speaker system 100C.

[0125] As described above, the second speaker 112 reproduces the second reproduction signal generated by processing only the low-frequency components of the sound source signal from the sound source 10 in the second signal processing unit 1200, and therefore the sound reproduced by the second speaker 112 should contain fewer high-frequency components. However, the sound reproduced by the second speaker 112 may contain high-frequency components that have not been completely removed. Therefore, by providing the second speaker 112 with the horn 102c that opens to the rear, the directional characteristics of the high frequencies are controlled by the effect of the horn 102c, and it is possible to prevent the high-frequency components of the sound reproduced by the second speaker 112 from leaking into the first control point CP1 and the second control point CP2.

[0126] [Variation 4] Fig. 9 is a block diagram showing the configuration of a speaker system according to Variation 4 of Embodiment 1 in coefficient design mode. As shown in Fig. 9, speaker system 100D according to Variation 4 of Embodiment 1 differs from speaker system 100 according to Embodiment 1 mainly in that, instead of processing unit 1000, processing unit 1000D having a second target characteristic different from that of processing unit 1000 is provided. Note that speaker system 100D may include second speaker 112 instead of second speaker 102, as in Variation 1 or 3 above. Furthermore, in speaker system 100D, second speaker 102 may emit sound backward, as in Variation 2 above.

[0127] As shown in Fig. 9, the processing unit 1000D in the coefficient design mode has a configuration in which the second target characteristic filter 1502 and the second error signal adder 1602 are removed from the processing unit 1000 in the coefficient design mode shown in Fig. 4A. The configuration of the processing unit 1000D in the service mode is the same as the configuration of the processing unit 1000 in the service mode shown in Fig. 1A. In the processing unit 1000D in the service mode, the coefficients of the first FIR filter 1110 and the second FIR filter 1210 are different from those of the processing unit 1000, so that the first signal processing unit 1100 generates a first reproduced signal and the second signal processing unit 1200 generates a second reproduced signal to realize a second target characteristic different from that of the processing unit 1000.

[0128] In processing unit 1000D, since second target characteristic filter 1502 and second error signal adder 1602 are not provided, the detection signal from second microphone 1702 is directly input to second coefficient updater 1112 and fifth coefficient updater 1212. In this case, second coefficient updater 1112 updates the coefficient of first FIR filter 1110 so as to minimize the detection signal from second microphone 1702. Furthermore, fifth coefficient updater 1212 updates the coefficient of second FIR filter 1210 so as to minimize the detection signal from second microphone 1702.

[0129] As a result, in the service mode, the first signal processing unit 1100 generates a first reproduction signal, and the second signal processing unit 1200 generates a second reproduction signal, so that the reproduced sound based on the sound source signals reproduced from the first speaker 101 and the second speaker 102 achieves the first target characteristic in a predetermined frequency band at the first control point CP1 and is minimized at the second control point CP2. That is, in the processing unit 1000D, the second target characteristic is a characteristic that minimizes the level of the reproduced sound at the second control point CP2. As a result, in the speaker system 100D, by eliminating the second target characteristic filter 1502 and the second error signal adder 1602, the amount of calculation and memory capacity can be reduced, and the level of the reproduced sound at the second control point CP2 can be reduced. The characteristic that minimizes the level of the reproduced sound at the second control point CP2 is a characteristic that is achieved by designing the signal processing coefficients so as to minimize the level of the reproduced sound using a predetermined algorithm, such as the LMS algorithm, as described above. In addition, in the processing unit 1000, even by setting the control coefficient of the second target characteristic filter 1502 to zero, the second target characteristic becomes a characteristic that minimizes the level of the reproduced sound at the second control point CP2, and the first signal processing unit 1100 generates the first reproduced signal and the second signal processing unit 1200 generates the second reproduced signal so that the reproduced sound is minimized at the second control point CP2.

[0130] (Embodiment 2) Next, a description will be given of the configuration of a speaker system according to embodiment 2. In the following description of embodiment 2, differences from embodiment 1 and the modifications of embodiment 1 will be mainly described, and descriptions of commonalities will be omitted or simplified.

[0131] Fig. 10A is a block diagram showing the configuration of a speaker system according to embodiment 2 in a service mode. Fig. 10B is a plan view showing the appearance of speaker system 200 according to embodiment 2. Fig. 10B shows a plan view of speaker system 200 as viewed from a direction perpendicular to axis L. In Fig. 10B and Fig. 13B described later, diaphragms 101a, 102a, and 103a housed in housings 101b, 102b, and 103b, respectively, are shown by dashed lines. Also, in Fig. 10B and Fig. 13B described later, the shape of speaker system 200 with horn 102c removed is shown by solid lines, and horn 102c is shown by a two-dot chain line.

[0132] 10A and 10B, speaker system 200 according to embodiment 2 differs from speaker system 100 according to embodiment 1 mainly in that it includes second speaker 112 and processing unit 2000 instead of second speaker 102 and processing unit 1000, and in that it further includes third speaker 103. Furthermore, the appearance of speaker system 200 is the same as that of speaker system 100A described above, with third speaker 103 added.

[0133] The first speaker 101, the second speaker 112, and the third speaker 103 are arranged on a straight line. More specifically, as shown in Fig. 10B, the first speaker 101, the second speaker 112, and the third speaker 103 are arranged side by side in the front-to-rear direction so that their respective diaphragms (diaphragm 101a, diaphragm 102a, and diaphragm 103a described below) are positioned coaxially with respect to axis L.

[0134] The third speaker 103 is located behind the first speaker 101 and the second speaker 112. The third speaker 103 has a diaphragm 103a, which is part of a speaker unit, a housing 103b, and a horn 103c. In addition to the diaphragm 103a, the speaker unit of the third speaker 103 also includes, for example, a support member for supporting the diaphragm 103a, a magnet, a voice coil, and the like (not shown).

[0135] The diaphragm 103a faces rearward, and the third speaker 103 emits sound rearward as the diaphragm 103a vibrates. The diaphragm 103a is disposed so that the axis L passes through its center. The diaphragm 101a of the first speaker 101, the diaphragm 102a of the second speaker 112, and the diaphragm 103a of the third speaker 103 are positioned coaxially with the axis L as their axis.

[0136] Housing 103b is, for example, a cylindrical housing whose axial direction is the extension direction of axis line L. An opening in which diaphragm 103a is installed is provided at the rear end of housing 103b, and this opening is closed by diaphragm 103a. In addition, the rear end of housing 103b is closed by a plate-like member.

[0137] Horn 103c is installed behind diaphragm 103a and has an opening at the rear. That is, horn 103c has an opening in the direction in which third speaker 103 emits sound. Horn 103c serves to impart directionality mainly to the high frequencies of the sound emitted from third speaker 103. That is, the sound emitted from third speaker 103 has a characteristic in which the sound is reproduced evenly from low to high frequencies behind speaker system 200, but has a characteristic in which the level of the high frequencies is reduced relative to the low frequencies in the front of speaker system 200.

[0138] In speaker system 200, horn 101c of first speaker 101 and horn 102c of second speaker 112 are intended to suppress the high-frequency level of the sound emitted from first speaker 101 and second speaker 112 relative to third control point CP3. Horn 103c of third speaker 103 is intended to suppress the high-frequency level of the sound emitted from third speaker 103 relative to first control point CP1 and second control point CP2. Here, third control point CP3 is located rearward of third speaker 103. In the example shown in FIG. 10B , third control point CP3 is located rearward of third speaker 103 and away from reference position SP2 in a direction that forms a predetermined angle γ (0 degrees≦γ≦90 degrees) with respect to axis L. Reference position SP2 is a position on axis L based on the position of third speaker 103. In the example shown in Fig. 10B, reference position SP2 is the center of the opening of horn 103c, but is not limited to this. Reference position SP2 may be, for example, a position between the center of the opening of horn 103c and the center of the opening of horn 101c.

[0139] 10A, processing unit 2000 has first signal processing unit 2100, second signal processing unit 2200, and third signal processing unit 2300. The functional configurations of first signal processing unit 2100 and second signal processing unit 2200 in the service mode are the same as the functional configurations of first signal processing unit 1100 and second signal processing unit 1200 shown in FIG. 1A.

[0140] The first signal processing unit 2100 performs signal processing on the sound source signal from the sound source 10 to generate a first reproduction signal, and inputs the first reproduction signal to the first speaker 101. The first speaker 101 emits sound based on the first reproduction signal input from the first signal processing unit 2100. The second signal processing unit 2200 performs signal processing on the sound source signal from the sound source 10 to generate a second reproduction signal, and inputs the second reproduction signal to the second speaker 112. The second speaker 112 emits sound based on the second reproduction signal input from the second signal processing unit 2200. The third signal processing unit 2300 performs signal processing on the sound source signal from the sound source 10 to generate a third reproduction signal, and inputs the third reproduction signal to the third speaker 103. The third speaker 103 emits sound based on the third reproduction signal input from the third signal processing unit 2300 .

[0141] In speaker system 200, first signal processing unit 2100 generates a first playback signal, second signal processing unit 2200 generates a second playback signal, and third signal processing unit 2300 generates a third playback signal so that playback sound based on sound source signals played from first speaker 101, second speaker 112, and third speaker 103 achieves a first target characteristic in a predetermined frequency band at first control point CP1, the playback sound achieves a second target characteristic in a predetermined frequency band at second control point CP2, and the playback sound achieves a third target characteristic in a predetermined frequency band at third control point CP3. In speaker system 200, the sound reproduced from first speaker 101, second speaker 102, and third speaker 103 is a composite sound of the sound emitted by first speaker 101 based on the first playback signal, the sound emitted by second speaker 112 based on the second playback signal, and the sound emitted by third speaker 103 based on the third playback signal.

[0142] The third signal processing unit 2300 includes a third FIR filter 1310 (FIR3 in the figure). The third FIR filter 1310 generates a third reproduction signal by convolving the output signal of the low-frequency extraction crossover filter 1120, which is the low-frequency component of the sound source signal extracted by the low-frequency extraction crossover filter 1120, with a coefficient. The generated third reproduction signal is input to the third speaker 103. The convolution process by the first FIR filter 1110, the convolution process by the second FIR filter 1210, and the convolution process by the third FIR filter 1310 are performed in parallel, for example.

[0143] The third FIR filter 1310 is an adaptive filter that is set to filter characteristics determined by, for example, adaptive filter design. The third FIR filter 1310 adjusts the phase and amplitude of the input signal for each frequency. The third FIR filter 1310 multiplies the phase and amplitude of the input signal for each frequency by a coefficient and outputs the result.

[0144] The coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 have characteristics that enable the reproduced sound based on the sound source signal reproduced from the first speaker 101, the second speaker 112, and the third speaker 103 to achieve the first target characteristic at the first control point CP1, the second target characteristic at the second control point CP2, and the third target characteristic at the third control point CP3. As a result, the reproduced sound from the first speaker 101, the second speaker 112, and the third speaker 103 can simultaneously achieve the first target characteristic at the first control point CP1, the second target characteristic at the second control point CP2, and the third target characteristic at the third control point CP3.

[0145] The first target characteristic and the second target characteristic are, for example, the same as those of the speaker system 100 according to the first embodiment. Specifically, the first target characteristic is, for example, a characteristic for improving the level of a low frequency range below a predetermined frequency of the reproduced sound at the first control point CP1 and flattening the frequency characteristic. Furthermore, the second target characteristic is, for example, a characteristic in which the frequency characteristic is lowered by a certain level compared to the first target characteristic.

[0146] Furthermore, the third target characteristic is a characteristic that reduces the level of the reproduced sound more than the first target characteristic. For example, the third target characteristic is a characteristic that minimizes the level of the reproduced sound at the third control point CP3. In this case, the reproduced sound reproduced from the first speaker 101, the second speaker 112, and the third speaker 103 can be minimized at the third control point CP3. In other words, the first signal processing unit 2100 generates the first reproduced signal, the second signal processing unit 2200 generates the second reproduced signal, and the third signal processing unit 2300 generates the third reproduced signal so that the reproduced sound is minimized at the third control point CP3. The characteristic that minimizes the level of the reproduced sound at the third control point CP3 is a characteristic that is realized by designing signal processing coefficients so as to minimize the level of the reproduced sound using a predetermined algorithm such as an LMS algorithm, as will be described later.

[0147] Furthermore, as with speaker system 100, the output signal of high-frequency extraction crossover filter 1130 is used only to generate the first reproduction signal, and therefore the high-frequency components of the sound source signal are reproduced from first speaker 101. That is, only the low-frequency components of the sound source signal are reproduced from second speaker 112 and third speaker 103, but the entire frequency band of the sound source signal is reproduced from first speaker 101. As a result, due to a synergistic effect with the acoustic characteristics of horn 101c, it is possible to achieve reproduced sound with a high sound pressure level over a wide frequency band from low to high, which is the first target characteristic, at first control point CP1. On the other hand, due to the effect of the sound emission direction of the first speaker 101 and the acoustic characteristics of the horn 101c, the high-frequency characteristics are reduced in level at the second control point CP2 and the third control point CP3, and the low-frequency characteristics are also reduced in level to achieve the second target characteristics and the third target characteristics.

[0148] Here, the first target characteristic, the second target characteristic, and the third target characteristic will be described while showing the results of measuring the frequency characteristics of the reproduced sound actually emitted by the speaker system 200. FIG. 11A is a diagram showing the frequency characteristics before and after control of the reproduced sound at the first control point. FIG. 11B is a diagram showing the frequency characteristics after control of the reproduced sound at the first control point and the second control point. FIG. 11C is a diagram showing the frequency characteristics before and after control of the reproduced sound at the third control point. In FIGS. 11A, 11B, and 11C, the horizontal axis represents frequency, and the vertical axis represents level (sound pressure). The characteristics before control shown in FIGS. 11A and 11C are the characteristics of the reproduced sound when reproduced only from the first speaker 101 without performing signal processing on the sound source signal. 11A, 11B, and 11C are the characteristics of the sound reproduced from the first speaker 101, the second speaker 112, and the third speaker 103 as a result of signal processing performed by the processing unit 2000. In addition, the cutoff frequencies of the low-frequency extraction crossover filter 1120 and the high-frequency extraction crossover filter 1130 during this signal processing are 5 kHz.

[0149] 11A, when the first target characteristic shown by the dashed line is set to the characteristic of the reproduced sound before control at the first control point CP1 shown by the dashed line, the characteristic of the reproduced sound after control shown by the solid line can be realized. More specifically, the characteristic of the reproduced sound before control at the first control point CP1 shown by the dashed line has a low level in the frequency band below 900 Hz, and in the frequency band above 900 Hz, the frequency characteristics are significantly disturbed, with noticeable peak and dip characteristics. In contrast, as a result of setting the first target characteristic shown by the dashed line and controlling the reproduced sound, the characteristic of the reproduced sound after control shown by the solid line shows that the level in the frequency band below 900 Hz is improved, and the peak and dip characteristics in the frequency band above 900 Hz are also suppressed.

[0150] Here, first FIR filter 1110, second FIR filter 1210, and third FIR filter 1310 do not control the high frequency range above 5 kHz as in the first target characteristic, so the high frequency components added by playback signal adder 1160 achieve high frequency characteristics above 5 kHz. Similar to speaker system 100, signal processing is performed by level adjuster 1140 and delay unit 1150 to prevent disturbances in frequency characteristics and phase from occurring when the low frequency characteristics and high frequency characteristics are combined by this addition.

[0151] 11B, when the reproduced sound is controlled by setting a second target characteristic indicated by a two-dot chain line, which has a frequency characteristic 10 dB lower than the first target characteristic indicated by a dot-dash line, it becomes possible to achieve the characteristics of the reproduced sound after control at the first control point CP1 indicated by the solid line and the characteristics of the reproduced sound after control at the second control point CP2 indicated by the dotted line. The difference in characteristics is roughly equal to the difference between the first target characteristic and the second target characteristic, and is about 10 dB.

[0152] Furthermore, by controlling the reproduced sound to be minimized at the third control point CP3, it is possible to achieve the characteristics of the reproduced sound after control shown by the solid line, which has a lower level than the characteristics of the reproduced sound before control at the third control point CP3 shown by the dashed line, as shown in Figure 11C.

[0153] The method for designing the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 will be described in detail later.

[0154] As described above, speaker system 100 controls the reproduced sound to achieve the first target characteristic at first control point CP1, the second target characteristic at second control point CP2, and the third target characteristic at third control point CP3 as shown and explained in Figures 11A, 11B, and 11C, thereby achieving an improvement in the low-frequency sound pressure level of the reproduced sound at first control point CP1, as well as improvements in the peak and dip characteristics of the reproduced sound, while achieving directionality of the reproduced sound over a wide frequency band from low to high frequencies.

[0155] FIG. 12 shows the measurement results of the directional characteristics when the predetermined angle θ shown in FIG. 10B is set to 60 degrees and the predetermined angle γ is set to 45 degrees. FIG. 12 is a diagram illustrating an example of the directional characteristics of the reproduced sound of the speaker system according to embodiment 2. The angles shown in FIG. 12 are angles seen from the center of the opening of horn 101c, when the direction of first control point CP1 (first control point direction) as seen from the center of the opening of horn 101c is set to 0 degrees. As shown in FIG. 12, when the frequency levels in the 0-degree direction, which is the first control point direction, are set to 0 dB, the levels in the +60-degree direction, which is the second control point direction, are -10 dB or less from low to high frequencies. Furthermore, because the diaphragm 101a of first speaker 101, the diaphragm 102a of second speaker 112, and the diaphragm 103a of third speaker 103 are configured to be positioned coaxially, the same characteristics are obtained even in the -60-degree direction. Similarly, the level in the direction of -135 degrees (=-180+45) which is the direction of the third control point is -10 dB or less except for 700 Hz and 1 kHz, and the same characteristics are obtained in the direction of +135 degrees.

[0156] Here, the levels of 700 Hz and 1 kHz are only about -6 dB in the ±135-degree direction. However, compared to the directional characteristics of speaker system 100 shown in FIG. 3, the directional characteristics of speaker system 200 shown in FIG. 12 have lower levels in the range from +90 degrees to +120 degrees. The same is true for the range from -90 degrees to -120 degrees. Thus, when comparing the directional characteristics of speaker system 100 shown in FIG. 3 with the directional characteristics of speaker system 200 shown in FIG. 12, the directional characteristics of speaker system 200 have lower overall levels in the +90-degree direction and the rearward direction behind the -90-degree direction. This can be said to be the effect of minimizing the playback sounds from first speaker 101, second speaker 112, and third speaker 103 at third control point CP3, as shown in FIG. 11C.

[0157] As described above, at the front of speaker system 200, i.e., at first control point CP1 in the 0-degree direction, a control effect of flattening the frequency characteristics and increasing the level of the reproduced sound is obtained over a wide frequency range from low to high. At the same time, at ±60 degrees of speaker system 200, corresponding to the direction of second control point CP2, a control effect of reducing the level of the reproduced sound by 10 dB or more from the frequency characteristics in the 0-degree direction over a wide frequency band is obtained. Furthermore, by setting third control point CP3 at a position behind speaker system 200, a control effect of reducing the level of the reproduced sound over a wide frequency range in directions behind the +90-degree and -90-degree directions is obtained. Therefore, even when the influence of the environment behind speaker system 200 cannot be ignored, for example, even when there is a wall or ceiling near the rear of speaker system 200, the influence of the reflected sound from the wall or ceiling on the sound quality of the reproduced sound in front of speaker system 200 can be suppressed.

[0158] As explained with reference to FIG. 3, the directional characteristics shown in FIG. 12 are also realized as a sphere including the horizontal and vertical planes.

[0159] 10B, the predetermined angle γ that defines the position of the third control point CP3 is set to 45 degrees, but it is not limited to this angle and may be set to any angle that achieves appropriate directional characteristics in accordance with the installation conditions of the speaker system 200. The same applies to the distance from the reference position that defines the position of the third control point CP3 to the third control point CP3.

[0160] Furthermore, the first target characteristic can be set to any characteristic that is optimal for the intended use, such as not only a characteristic that flattens the frequency characteristics as shown in Figures 11A and 11B, but also a characteristic that emphasizes the level of a certain frequency band, such as a characteristic that boosts the low range, or, conversely, a characteristic that lowers the level of a certain frequency band.

[0161] Next, the design of coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 to obtain the above-described control effect will be described with reference to FIGS. 13A and 13B . FIG. 13A is a block diagram showing the configuration of a speaker system according to embodiment 2 in a coefficient design mode. FIG. 13B is a diagram for explaining the placement of microphones in the coefficient design mode. FIG. 13B shows a plan view of the speaker system 200 as viewed from a direction perpendicular to the axis L. The processing unit 2000 of the speaker system 200 has, for example, a service mode having the functional configuration shown in FIG. 10A and a coefficient design mode having the functional configuration shown in FIG. 13A as its operating modes. The service mode is an operating mode for playing music for a listener, etc. The coefficient design mode is an operating mode for determining the control characteristics of the first signal processing unit 2100, the second signal processing unit 2200, and the third signal processing unit 2300 (specifically, the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310).

[0162] 13A and 13B, a first microphone 1701, a second microphone 1702, and a third microphone 1703 are used to design the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310. As shown in Fig. 13B, the first microphone 1701 is placed at the position of the first control point CP1 shown in Fig. 10B, the second microphone 1702 is placed at the position of the second control point CP2 shown in Fig. 10B, and the third microphone 1703 is placed at the position of the third control point CP3 shown in Fig. 10B.

[0163] In the coefficient design mode, the processing unit 2000 has a first signal processing unit 2100, a second signal processing unit 2200, a third signal processing unit 2300, a first target characteristic filter 1501, a second target characteristic filter 1502, a first error signal adder 1601, and a second error signal adder 1602.

[0164] In the coefficient design mode, the first signal processing unit 2100 includes a third coefficient updater 1113 (LMS13 in the figure) and a third speaker characteristics compensation filter 1123 (Fx13 in the figure) in addition to the configuration of the first signal processing unit 1100 shown in Fig. 4A. In the coefficient design mode, the second signal processing unit 2200 includes a sixth coefficient updater 1213 (LMS23 in the figure) and a sixth speaker characteristics compensation filter 1223 (Fx23 in the figure) in addition to the configuration of the second signal processing unit 1200 shown in Fig. 4A. In the coefficient design mode, the third signal processing unit 2300 includes a third FIR filter 1310, a seventh coefficient updater 1311 (LMS31 in the figure), an eighth coefficient updater 1312 (LMS32 in the figure), a ninth coefficient updater 1313 (LMS33 in the figure), a seventh speaker characteristics correction filter 1321 (Fx31 in the figure), an eighth speaker characteristics correction filter 1322 (Fx32 in the figure), and a ninth speaker characteristics correction filter 1323 (Fx33 in the figure).

[0165] In the coefficient design mode, the first FIR filter 1110, the first speaker characteristic compensation filter 1121, the second speaker characteristic compensation filter 1122, and the third speaker characteristic compensation filter 1123 of the first signal processing unit 2100 each receive a sound source signal from the sound source 10 and perform convolution processing of the input sound source signal with a coefficient. The output signal of the first FIR filter 1110 is reproduced from the first speaker 101.

[0166] Similarly, in the coefficient design mode, the second FIR filter 1210, the fourth speaker characteristic compensation filter 1221, the fifth speaker characteristic compensation filter 1222, and the sixth speaker characteristic compensation filter 1223 of the second signal processing unit 2200 each receive the sound source signal from the sound source 10 and perform convolution processing of the input sound source signal with the coefficients. The output signal of the second FIR filter 1210 is reproduced from the second speaker 112.

[0167] Furthermore, in the coefficient design mode, the third FIR filter 1310, the seventh speaker characteristic compensation filter 1321, the eighth speaker characteristic compensation filter 1322, and the ninth speaker characteristic compensation filter 1323 of the third signal processing unit 2300 each receive the sound source signal from the sound source 10 and perform convolution processing of the input sound source signal with the coefficients. The output signal of the third FIR filter 1310 is reproduced from the third speaker 103.

[0168] Then, the sounds reproduced from first speaker 101, second speaker 112, and third speaker 103 are detected by first microphone 1701, second microphone 1702, and third microphone 1703. A detection signal of the reproduced sound detected by first microphone 1701 is input to first error signal adder 1601. A detection signal of the reproduced sound detected by second microphone 1702 is input to second error signal adder 1602.

[0169] On the other hand, the sound source signal from sound source 10 is input to first target characteristic filter 1501, in which a first target characteristic is set as a control coefficient, and is subjected to signal processing together with the control coefficient in first target characteristic filter 1501. Also, the sound source signal from sound source 10 is input to second target characteristic filter 1502, in which a second target characteristic is set as a control coefficient, and is subjected to signal processing together with the control coefficient in second target characteristic filter 1502. The output signal of first target characteristic filter 1501 is input to first error signal adder 1601. The output signal of second target characteristic filter 1502 is input to second error signal adder 1602.

[0170] Then, the first error signal adder 1601 adds the detection signal from the first microphone 1701 and the output signal from the first target characteristic filter 1501, and inputs the output signal generated by the addition to the first coefficient updater 1111, the fourth coefficient updater 1211, and the seventh coefficient updater 1311. As a result, the difference between the first target characteristic and the detection result of the reproduced sound by the first microphone 1701 is input to the first coefficient updater 1111, the fourth coefficient updater 1211, and the seventh coefficient updater 1311. Similarly, the second error signal adder 1602 adds the detection signal from the second microphone 1702 and the output signal from the second target characteristic filter 1502, and inputs the output signal generated by the addition to the second coefficient updater 1112, the fifth coefficient updater 1212, and the eighth coefficient updater 1312. As a result, the difference between the second target characteristic and the detection result of the reproduced sound by the second microphone 1702 is input to the second coefficient updater 1112, the fifth coefficient updater 1212, and the eighth coefficient updater 1312. On the other hand, the detection signal of the reproduced sound detected by the third microphone 1703 is directly input to the third coefficient updater 1113, the sixth coefficient updater 1213, and the ninth coefficient updater 1313.

[0171] The first speaker characteristic correction filter 1121 convolves the sound source signal from the sound source 10 with a coefficient, and inputs the output signal generated by the convolution to the first coefficient updater 1111. Similarly, the second speaker characteristic correction filter 1122 convolves the sound source signal from the sound source 10 with a coefficient, and inputs the output signal generated by the convolution to the second coefficient updater 1112. Furthermore, the third speaker characteristic correction filter 1123 has a coefficient that is preset to represent the transfer characteristic from the first speaker 101 to the third microphone 1703. The third speaker characteristic correction filter 1123 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the third coefficient updater 1113.

[0172] Here, for example, the first coefficient updater 1111, the second coefficient updater 1112, and the third coefficient updater 1113 use the LMS algorithm as described in Patent Document 2.

[0173] In this case, first coefficient updater 1111 uses the output signal from first speaker characteristic compensation filter 1121 and the output signal from first error signal adder 1601 to update the coefficients of first FIR filter 1110 so as to minimize the output signal from first error signal adder 1601. Also, second coefficient updater 1112 uses the output signal from second speaker characteristic compensation filter 1122 and the output signal from second error signal adder 1602 to update the coefficients of first FIR filter 1110 so as to minimize the output signal from second error signal adder 1602. Furthermore, third coefficient updater 1113 uses the output signal from third speaker characteristic compensation filter 1123 and the detection signal from third microphone 1703 to update the coefficients of first FIR filter 1110 so as to minimize the detection signal from third microphone 1703.

[0174] The fourth speaker characteristics correction filter 1221 convolves the sound source signal from the sound source 10 with a coefficient, and inputs the output signal generated by the convolution to the fourth coefficient updater 1211. Similarly, the fifth speaker characteristics correction filter 1222 convolves the sound source signal from the sound source 10 with a coefficient, and inputs this output signal to the fifth coefficient updater 1212. Furthermore, the sixth speaker characteristics correction filter 1223 has a coefficient set in advance that is the transfer characteristic from the second speaker 112 to the third microphone 1703. The sixth speaker characteristics correction filter 1223 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the result to the sixth coefficient updater 1213.

[0175] Here too, for example, the LMS algorithm as described in Patent Document 2 is used for the calculations of the fourth coefficient updater 1211, the fifth coefficient updater 1212, and the sixth coefficient updater 1213.

[0176] In this case, fourth coefficient updater 1211 uses the output signal from fourth speaker characteristic compensation filter 1221 and the output signal from first error signal adder 1601 to update the coefficients of second FIR filter 1210 so as to minimize the output signal from first error signal adder 1601. Furthermore, fifth coefficient updater 1212 uses the output signal from fifth speaker characteristic compensation filter 1222 and the output signal from second error signal adder 1602 to update the coefficients of second FIR filter 1210 so as to minimize the output signal from second error signal adder 1602. Furthermore, sixth coefficient updater 1213 uses the output signal from sixth speaker characteristic compensation filter 1223 and the detection signal from third microphone 1703 to update the coefficients of second FIR filter 1210 so as to minimize the detection signal from third microphone 1703.

[0177] The seventh speaker characteristics correction filter 1321 has a coefficient set in advance that represents the transfer characteristic from the third speaker 103 to the first microphone 1701. The seventh speaker characteristics correction filter 1321 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the seventh coefficient updater 1311. Similarly, the eighth speaker characteristics correction filter 1322 has a coefficient set in advance that represents the transfer characteristic from the third speaker 103 to the second microphone 1702. The eighth speaker characteristics correction filter 1322 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution to the eighth coefficient updater 1312. Furthermore, the ninth speaker characteristics correction filter 1323 has a coefficient set in advance that represents the transfer characteristic from the third speaker 103 to the third microphone 1703. The ninth speaker characteristic correction filter 1323 convolves the sound source signal from the sound source 10 with this coefficient, and inputs the output signal generated by the convolution process to the ninth coefficient updater 1313 .

[0178] Here too, for example, the seventh coefficient updater 1311, the eighth coefficient updater 1312, and the ninth coefficient updater 1313 use the LMS algorithm as described in Patent Document 2.

[0179] In this case, the seventh coefficient updater 1311 uses the detection signal from the seventh speaker characteristic compensation filter 1321 and the output signal from the first error signal adder 1601 to update the coefficients of the third FIR filter 1310 so as to minimize the output signal from the first error signal adder 1601. The eighth coefficient updater 1312 uses the output signal from the eighth speaker characteristic compensation filter 1322 and the output signal from the second error signal adder 1602 to update the coefficients of the third FIR filter 1310 so as to minimize the output signal from the second error signal adder 1602. The ninth coefficient updater 1313 uses the output signal from the ninth speaker characteristic compensation filter 1323 and the detection signal from the third microphone 1703 to update the coefficients of the third FIR filter 1310 so as to minimize the detection signal from the third microphone 1703.

[0180] The above-described processing designs the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310. Furthermore, as necessary, the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 are repeatedly updated until a reproduced sound with the desired frequency characteristics is obtained.

[0181] The coefficients thus obtained are set in the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 in the service mode shown in Fig. 10A. In the service mode, music is played back for the listener.

[0182] In other words, when music is played back for the listener, the above-described coefficient design has already been performed, and therefore the functional configuration of processing unit 2000 during coefficient design and music playback differs as shown in Fig. 13A and Fig. 10A. In this way, processing unit 2000 operates in a coefficient design mode when determining the control characteristics of first signal processing unit 2100, second signal processing unit 2200, and third signal processing unit 2300, and in a service mode when playing music for the listener, and changes the signal processing configuration according to each mode.

[0183] In speaker system 200, third speaker 103 has horn 103c, but is not limited to this. Since the third playback signal is generated by processing only the low-frequency components of the sound source signal from sound source 10 in third signal processing unit 2300, third speaker 103 does not need to have horn 103c in cases where the high-frequency components of the sound emitted by third speaker 103 do not affect the playback sound at first control point CP1 and second control point CP2. This also applies to the modified example of embodiment 2 described below.

[0184] Furthermore, in speaker system 200, second speaker 112 has horn 102c, but this is not limited to this. In cases where the high-frequency components of the sound emitted by second speaker 112 do not affect the sound reproduced at third control point CP3, second speaker 112 does not have to have horn 102c. In other words, speaker system 200 may be provided with second speaker 102 instead of second speaker 112. Furthermore, in speaker system 200, diaphragm 102a may face rearward, as in Variation 2 or Variation 3 of Embodiment 1. This also applies to the variations of Embodiment 2 described below.

[0185] (Modifications of Embodiment 2) Next, a description will be given of modifications of Embodiment 2. Note that in the following description of the modifications, differences between Embodiment 1, each modification of Embodiment 1, and Embodiment 2 and each modification of Embodiment 2 will be mainly described, and descriptions of commonalities will be omitted or simplified.

[0186] [Variation 1] Fig. 14 is a block diagram showing the configuration in coefficient design mode of a speaker system according to Variation 1 of Embodiment 2. As shown in Fig. 14, speaker system 200A according to Variation 1 of Embodiment 2 differs from speaker system 200 according to Embodiment 2 mainly in that, instead of processing unit 2000, processing unit 2000A having a second target characteristic different from that of processing unit 2000 is provided.

[0187] As shown in Fig. 14, processing unit 2000A in the coefficient design mode has a configuration in which second target characteristic filter 1502 and second error signal adder 1602 are removed from processing unit 2000 in the coefficient design mode shown in Fig. 13A. Note that the configuration of processing unit 2000A in the service mode is the same as the configuration of processing unit 2000 in the service mode shown in Fig. 10A. In processing unit 2000A in the service mode, first signal processing unit 2100 generates a first reproduced signal, second signal processing unit 2200 generates a second reproduced signal, and third signal processing unit 2300 generates a third reproduced signal so as to realize a second target characteristic different from that of processing unit 2000 by using different coefficients for first FIR filter 1110, second FIR filter 1210, and third FIR filter 1310.

[0188] In processing unit 2000A, since second target characteristic filter 1502 and second error signal adder 1602 are not provided, the detection signal from second microphone 1702 is directly input to second coefficient updater 1112, fifth coefficient updater 1212, and eighth coefficient updater 1312. In this case, second coefficient updater 1112 updates the coefficient of first FIR filter 1110 so as to minimize the detection signal from second microphone 1702. Furthermore, fifth coefficient updater 1212 updates the coefficient of second FIR filter 1210 so as to minimize the detection signal from second microphone 1702. Furthermore, eighth coefficient updater 1312 updates the coefficient of third FIR filter 1310 so as to minimize the detection signal from second microphone 1702. As a result, the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 are designed.

[0189] As a result, in the service mode, the first signal processing unit 2100 generates a first playback signal, the second signal processing unit 2200 generates a second playback signal, and the third signal processing unit 2300 generates a third playback signal so that the playback sound reproduced from the first speaker 101, the second speaker 112, and the third speaker 103 is minimized at the second control point CP2. That is, in the processing unit 2000A, the second target characteristic is a characteristic that minimizes the level of the playback sound at the second control point CP2. Note that, in the processing unit 2000, the second target characteristic also becomes a characteristic that minimizes the level of the playback sound at the second control point CP2 by setting the control coefficient of the second target characteristic filter 1502 to zero.

[0190] [Variation 2] Fig. 15 is a block diagram showing the configuration in coefficient design mode of a speaker system according to Variation 2 of Embodiment 2. As shown in Fig. 15, speaker system 200B according to Variation 2 of Embodiment 2 differs from speaker system 200 according to Embodiment 2 mainly in that, instead of processing unit 2000, it includes processing unit 2000B having a third target characteristic different from that of processing unit 2000.

[0191] 15, the processing unit 2000B in the coefficient design mode has a configuration in which a third target characteristic filter 1503 (third target characteristic F3 in the figure) and a third error signal adder 1603 are added to the processing unit 2000 in the coefficient design mode shown in FIG. 13A. The configuration of the processing unit 2000B in the service mode is the same as the configuration of the processing unit 2000 in the service mode shown in FIG. 10A. In the processing unit 2000B in the service mode, the first signal processing unit 2100 generates a first reproduced signal, the second signal processing unit 2200 generates a second reproduced signal, and the third signal processing unit 2300 generates a third reproduced signal, so as to realize a third target characteristic different from that of the processing unit 2000, by using different coefficients for the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310.

[0192] In processing unit 2000B, a detection signal of reproduced sound detected by third microphone 1703 is input to third error signal adder 1603. On the other hand, a sound source signal from sound source 10 is input to third target characteristic filter 1503, in which the third target characteristic is set as a control coefficient, and is subjected to signal processing together with the control coefficient in third target characteristic filter 1503. An output signal of third target characteristic filter 1503 is input to third error signal adder 1603. Third error signal adder 1603 adds the detection signal from third microphone 1703 and the output signal from third target characteristic filter 1503, and inputs the result to third coefficient updater 1113, sixth coefficient updater 1213, and ninth coefficient updater 1313. As a result, the difference between the third target characteristic and the detection result of the reproduced sound by the third microphone 1703 is input to the third coefficient updater 1113, the sixth coefficient updater 1213, and the ninth coefficient updater 1313. The third error signal adder 1603 finds the difference by, for example, adding the detection signal from the third microphone 1703 and the output signal from the third target characteristic filter 1503 after inverting the sign of one of the signals.

[0193] Here, the third target characteristic set as a control coefficient in the third target characteristic filter 1503 is, for example, a characteristic in which the frequency characteristic is lower by a certain level than the first target characteristic, like the second target characteristic. Also, the third target characteristic is, for example, a characteristic in which the level is lower than that of the second target characteristic.

[0194] The third coefficient updater 1113 uses the output signal from the third speaker characteristic compensation filter 1123 and the output signal from the third error signal adder 1603 to update the coefficients of the first FIR filter 1110 so as to minimize the output signal from the third error signal adder 1603. Furthermore, the sixth coefficient updater 1213 uses the output signal from the sixth speaker characteristic compensation filter 1223 and the output signal from the third error signal adder 1603 to update the coefficients of the second FIR filter 1210 so as to minimize the output signal from the third error signal adder 1603. Furthermore, the ninth coefficient updater 1313 uses the output signal from the ninth speaker characteristic compensation filter 1323 and the output signal from the third error signal adder 1603 to update the coefficients of the third FIR filter 1310 so as to minimize the output signal from the third error signal adder 1603. As a result, the coefficients of the first FIR filter 1110, the second FIR filter 1210, and the third FIR filter 1310 are designed.

[0195] As a result, in the service mode, the first signal processing unit 2100 generates a first playback signal, the second signal processing unit 2200 generates a second playback signal, and the third signal processing unit 2300 generates a third playback signal so that the playback sound reproduced from the first speaker 101, the second speaker 112, and the third speaker 103 achieves the third target characteristic at the third control point CP3.

[0196] In addition, in the processing unit 2000B, when the control coefficient of the third target characteristic filter 1503 is set to zero, the third target characteristic becomes a characteristic that minimizes the level of the reproduced sound at the third control point CP3. In other words, the third target characteristic becomes the same characteristic as that of the processing unit 2000 shown in FIG. 13A.

[0197] Furthermore, processing unit 2000B may not have second target characteristic filter 1502 and second error signal adder 1602. In this case, similar to processing unit 2000A shown in FIG. 14 , the detection signal from second microphone 1702 is directly input to second coefficient updater 1112, fifth coefficient updater 1212, and eighth coefficient updater 1312. As a result, in the service mode, first signal processing unit 2100 generates a first reproduction signal, second signal processing unit 2200 generates a second reproduction signal, and third signal processing unit 2300 generates a third reproduction signal so that the reproduced sounds reproduced from first speaker 101, second speaker 112, and third speaker 103 are minimized at second control point CP2. In other words, the second target characteristic is a characteristic that minimizes the level of the reproduced sound at second control point CP2.

[0198] (Other Embodiments) As described above, the embodiments (including modified examples) have been described as examples of the technology disclosed in this application. However, the technology in this disclosure is not limited to these, and can be applied to embodiments in which modifications, substitutions, additions, omissions, etc. are made as appropriate. Furthermore, it is also possible to combine the components described in the above embodiments to create new embodiments.

[0199] Furthermore, for example, in the above-described embodiment, the processing units 1000, 1000D, 2000, 2000A, and 2000B have a service mode and a coefficient design mode, but this is not limited to this. The processing units 1000 and 2000 may not have a coefficient design mode. In this case, a processing device such as a computer may design the coefficients of the first FIR filter 1110 and the second FIR filter 1210 (and also the third FIR filter 1310 in the second embodiment), and the designed coefficients may be set in the processing units 1000, 1000D, 2000, 2000A, and 2000B.

[0200] Furthermore, for example, the method of designing the coefficients of the first FIR filter 1110 and the second FIR filter 1210 (and the third FIR filter 1310 in the second embodiment) is not limited to the examples described in the above embodiments, and other methods such as a method using simulation or the like may also be used.

[0201] Furthermore, for example, in the above-described embodiments, the first signal processing units 1100 and 2100 extract low-frequency components and high-frequency components from the sound source signal from the sound source 10 using the low-frequency extraction crossover filter 1120 and the high-frequency extraction crossover filter 1130. However, this is not limiting. For example, the sound source signal from the sound source 10 may be input directly to the first FIR filter 1110 and the second FIR filter 1210 (and further to the third FIR filter 1310 in the second embodiment).

[0202] In the above embodiments, the processes executed by a specific processing circuit such as a processor may be executed by another processing circuit. The order of multiple processes may be changed, or multiple processes may be executed in parallel. The program may be executed by a single processor or multiple processors. In other words, centralized processing or distributed processing may be performed.

[0203] Furthermore, the general or specific aspects of the present disclosure may be realized as a system, device, method, integrated circuit, computer program, or computer-readable recording medium such as a CD-ROM. Also, they may be realized as any combination of a system, device, method, integrated circuit, computer program, computer program product, and recording medium. Furthermore, for example, a speaker system may be realized as a single device or may be realized by multiple devices. Furthermore, when a speaker system is realized by multiple devices, the components of the speaker system may be distributed among the multiple devices in any manner.

[0204] The present disclosure can be used in a directional speaker system or the like.

[0205] 10 Sound source 100, 100A, 100B, 100C, 100D, 200, 200A Speaker system 101 First speaker 101a, 102a, 103a Diaphragm 101b, 102b, 103b Housing 101c, 102c, 103c Horn 102, 112 Second speaker 103 Third speaker 1000, 1000D, 2000, 2000A, 2000B Processing unit 1100, 2100 First signal processing unit 1110 First FIR filter 1111 First coefficient updater 1112 Second coefficient updater 1113 Third coefficient updater 1120 Low frequency extraction crossover filter 1121 First speaker characteristic compensation filter 1122 Second speaker characteristic compensation filter 1123 Third speaker characteristic compensation filter 1130 High frequency extraction crossover filter 1140 Level adjuster 1150 Delay unit 1160 Reproduction signal adder 1200, 2200 Second signal processing unit 1210 Second FIR filter 1211 Fourth coefficient updater 1212 Fifth coefficient updater 1213 Sixth coefficient updater 1221 Fourth speaker characteristic compensation filter 1222 Fifth speaker characteristic compensation filter 1223 Sixth speaker characteristic compensation filter 1310 Third FIR filter 1311 Seventh coefficient updater 1312 Eighth coefficient updater 1313 Ninth coefficient updater 1321 Seventh speaker characteristic compensation filter 1322 Eighth speaker characteristic compensation filter 1323 Ninth speaker characteristic compensation filter 1501 First target characteristic filter 1502 Second target characteristic filter 1503 Third target characteristic filter 1601 First error signal adder 1602 Second error signal adder 1603 Third error signal adder 1701 First microphone 1702 Second microphone 1703 Third microphone 2300 Third signal processing unit CP1 First control point CP2 Second control point CP3 Third control point SP1, SP2 Reference positions

Claims

1. A device comprising: a plurality of speakers arranged side by side so that their diaphragms are positioned coaxially with an axis extending in the front-to-rear direction; and a processing unit that performs signal processing on a sound source signal input from a sound source, wherein the plurality of speakers include a first speaker that emits sound forward and a second speaker located behind the first speaker, wherein the first speaker has a first horn with an opening in front of it, and the processing unit comprises a first signal processing unit that processes the sound source signal to generate a first playback signal and inputs the first playback signal to the first speaker, and a second signal processing unit that processes the sound source signal to generate a second playback signal and inputs the second playback signal to the second speaker, wherein the first speaker emits sound based on the first playback signal, and the second speaker emits sound based on the second playback signal, a first control point is a position that is a certain distance in front of and in front of a reference position on the axis based on the position of the first speaker, and a second control point is a position that is away from the reference position in a direction different from the direction toward the first control point, wherein the first signal processing unit generates the first reproduction signal and the second signal processing unit generates the second reproduction signal so that reproduced sound based on the sound source signals reproduced from the plurality of speakers achieves first target characteristics in a predetermined frequency band at the first control point and the reproduced sound achieves second target characteristics in the predetermined frequency band at the second control point, and the second target characteristics are characteristics that result in a lower level of the reproduced sound than the first target characteristics.

2. The speaker system according to claim 1, wherein the second control point is located in front of the first speaker and at a certain distance from the reference position in a direction forming a predetermined angle with respect to the axis.

3. The speaker system according to claim 1, wherein the second control point is located behind the second speaker.

4. The speaker system of claim 1, wherein the first signal processing unit includes a low-frequency extraction crossover filter that extracts low-frequency components below a predetermined frequency from the sound source signal, a high-frequency extraction crossover filter that extracts high-frequency components above a predetermined frequency from the sound source signal, a level adjuster that adjusts the level of the output signal from the high-frequency extraction crossover filter, a delay unit that delays the output signal from the level adjuster, a first FIR filter that performs convolution processing on the output signal from the low-frequency extraction crossover filter, and a playback signal adder that generates the first playback signal by adding the output signal from the delay unit and the output signal of the first FIR filter, and the second signal processing unit includes a second FIR filter that generates the second playback signal by convolution processing the output signal from the low-frequency extraction crossover filter.

5. The plurality of speakers includes a third speaker located behind the second speaker and emitting sound rearward, the processing unit has a third signal processing unit that processes the sound source signal to generate a third reproduction signal and inputs the third reproduction signal to the third speaker, the third speaker emits sound based on the third reproduction signal, and when a position behind the third speaker is set as a third control point, the first signal processing unit generates the first reproduction signal, the second signal processing unit generates the second reproduction signal, and the third signal processing unit generates the third reproduction signal so that the reproduced sound achieves the first target characteristic in the predetermined frequency band at the first control point, the reproduced sound achieves the second target characteristic in the predetermined frequency band at the second control point, and the reproduced sound achieves the third target characteristic in the predetermined frequency band at the third control point, and the third target characteristic is a characteristic such that the level of the reproduced sound is lower than the first target characteristic.

2. The speaker system according to claim 1.

6. The speaker system of claim 5, wherein the first signal processing unit includes: a low-frequency extraction crossover filter that extracts low-frequency components below a predetermined frequency from the sound source signal; a high-frequency extraction crossover filter that extracts high-frequency components above a predetermined frequency from the sound source signal; a level adjuster that adjusts the level of the output signal from the high-frequency extraction crossover filter; a delay unit that delays the output signal from the level adjuster; a first FIR filter that convolution-processes the output signal from the low-frequency extraction crossover filter; and a playback signal adder that generates the first playback signal by adding the output signal from the delay unit and the output signal of the first FIR filter; the second signal processing unit includes a second FIR filter that convolution-processes the output signal from the low-frequency extraction crossover filter to generate the second playback signal; and the third signal processing unit includes a third FIR filter that convolution-processes the output signal from the low-frequency extraction crossover filter to generate the third playback signal.

7. A speaker system as described in claim 5, wherein at least one of the following is satisfied: (i) the second target characteristic is a characteristic that minimizes the level of the reproduced sound at the second control point; and (ii) the third target characteristic is a characteristic that minimizes the level of the reproduced sound at the third control point.

8. The speaker system according to claim 5, wherein the third target characteristic is a characteristic in which the frequency characteristic is lower by a certain level than the first target characteristic.

9. The speaker system according to claim 5, wherein the third speaker has a horn with an opening at the rear.

10. A speaker system according to any one of claims 1 to 9, wherein the second speaker emits sound forward.

11. The speaker system according to claim 10, wherein the second speaker has a horn with an opening at the front.

12. A speaker system according to any one of claims 1 to 9, wherein the second speaker emits sound rearward.

13. The speaker system according to claim 12, wherein the second speaker has a horn with an opening at the rear.

14. A speaker system according to any one of claims 1 to 9, wherein the first target characteristic is a characteristic that improves the level of low frequencies below a predetermined frequency of the reproduced sound at the first control point and flattens the frequency characteristics.

15. A speaker system according to any one of claims 1 to 9, wherein the second target characteristic is a characteristic in which the frequency characteristic is lowered by a certain level compared to the first target characteristic.

Images