JP2014060584A - Speaker arrangement design support system, speaker arrangement design support device, speaker arrangement design support method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a speaker arrangement design support system, a speaker arrangement design support device, a speaker arrangement design support method and a program capable of determining a proper speaker arrangement by simple processing.SOLUTION: A speaker arrangement design support system 10 includes a speaker 110 the arrangement state of which is variable, a dummy head 200 including a microphone 220 attached to a dummy head 210, and a speaker arrangement support device 300 including an impulse response measurement unit 311 for measuring impulse response on the basis of a measurement signal outputted from the speaker 110 and inputted from the microphone 220, an evaluation data operation unit 312 for calculating the evaluation data for determining the arrangement of the speaker 110 on the basis of the impulse response, and a speaker arrangement determination unit 313 for determining the arrangement of the speaker 110 so that a sound image is localized at a desired position on the basis of the evaluation data.

Description

  The present invention relates to a speaker arrangement design support apparatus, a speaker arrangement design support apparatus, a speaker arrangement design support method, and a program for designing a speaker arrangement that can localize a sound image at a desired position.

  Patent Document 1 discloses an arrangement of an on-vehicle speaker for localizing a sound image. The speaker is arranged so that the main sound axis is oriented substantially parallel to the windshield and the instrument panel surface. Further, a pair of middle and high tone speakers are provided at the front lower part of the front door and / or the dash side. According to this configuration, it is stated that stereo separation can be secured and a sound image localization without bias can be formed.

Japanese Patent No. 2890764

  Although the method according to Patent Document 1 is suitable for music appreciation, it is difficult for a driver to localize a sound image at a desired position.

  Warning sound in driving a car was generated by a piezoelectric speaker built in the meter box, and the localization was ambiguous. It is very important to make them understand immediately what the warning sound is when driving. Therefore, localization of the warning sound in the target object or target direction of the warning should greatly help understanding.

  Therefore, the inventors have previously developed a sound field generating device of Japanese Patent Application No. 2011-089915 (not disclosed at the time of filing). In this sound field generation device, a speaker unit that has a speaker that outputs sound of two or more channels and outputs an acoustic signal toward a reflection unit that reflects the sound output from the speaker and transmits the sound to a listener And an audio signal in which a sound image of an acoustic signal output from the speaker unit and reflected by the reflecting unit to reach the listener is localized as a virtual sound source at the listener's position is input, and the input audio A crosstalk canceling unit that performs arithmetic processing so that crosstalk between the channels of the acoustic signal that is reflected by the reflection unit and reaches the listener is canceled at the position of the listener ing.

  In this sound field generator, in order to determine the speaker arrangement with the most localization effect, the vehicle shape, the interior shape around the driver (dashboard, meter box, etc.), the sound absorption rate of the interior material, etc. are used as input data. It is necessary to perform acoustic simulation. This took time and effort. The speaker arrangement that omits the simulation and is simply reflected has a problem that it is not localized in a desired direction.

  The present invention has been made in view of the above problems, and a speaker arrangement design support apparatus, a speaker arrangement design support apparatus, a speaker arrangement design support method, and a speaker arrangement design support apparatus capable of determining an appropriate speaker arrangement with simple processing, and The purpose is to provide a program.

  A speaker arrangement design support system (10) according to an aspect of the present invention includes a speaker (110) that can change an arrangement state and outputs a measurement signal, a dummy head (210), and left and right of the dummy head (210). A dummy head device (200) having a microphone (220) attached to each ear part, and an impulse response for measuring an impulse response based on the measurement signal output from the speaker (110) and input from the microphone (200) A measurement unit (311), an evaluation data calculation unit (312) for calculating evaluation data for determining the arrangement of the speakers (110) based on the impulse response measured by the impulse response measurement unit (311), and the evaluation Desired based on the evaluation data calculated by the data calculation unit (312) Speaker placement determining unit which determines the placement of the speaker sound in localization position is localized (110) (313), a speaker layout design support apparatus comprising a (300), those with a.

  The speaker arrangement design support device (300) according to one aspect of the present invention is an impulse based on the measurement signal output from the speaker (100) that outputs the measurement signal and input from the microphone (220) installed at the listening position. An impulse response measuring unit (311) for measuring a response, and an evaluation data calculation for calculating evaluation data for determining the arrangement of the speaker (100) based on the impulse response measured by the impulse response measuring unit (311) And a speaker arrangement determination unit (313) for determining the arrangement of the speaker (110) where the sound image is localized at a desired localization position based on the evaluation data calculated by the evaluation data calculation unit (312). ).

  The speaker arrangement design support method according to an aspect of the present invention measures an impulse response based on the measurement signal output from the speaker (110) that outputs the measurement signal and input from the microphone (220) installed at the listening position. An impulse response measuring step, an evaluation data calculating step for calculating evaluation data for determining an arrangement of the speakers (110) based on the impulse response measured in the impulse response measuring step, and an evaluation data calculating step A speaker arrangement determining step of determining an arrangement of the speakers (110) where the sound image is localized at a desired localization position based on the calculated evaluation data.

  A program according to an aspect of the present invention is output from a speaker (110) that outputs a measurement signal to a computer (310) included in the speaker arrangement design support device (300) and from a microphone (220) that is installed at a listening position. An impulse response measurement step for measuring an impulse response based on the input measurement signal, and evaluation data for determining the arrangement of the speakers (110) based on the impulse response measured in the impulse response measurement step. An evaluation data calculation step; and a speaker arrangement determination step for determining an arrangement of the speaker (110) where the sound image is localized at a desired localization position based on the evaluation data calculated in the evaluation data calculation step. Is.

  As described above, according to the present invention, it is possible to provide a speaker arrangement design support system, a speaker arrangement design support apparatus, a speaker arrangement design support method, and a program that can determine an appropriate speaker arrangement with simple processing. it can.

It is a figure which shows the structure which has arrange | positioned the speaker in the position of the meter box in a vehicle. (A) is a graph which shows the binaural response which is a measurement result of the impulse response in a laboratory, (b) is a graph which shows the binaural response in a vehicle. It is a figure which shows the path | route of the reflected sound at the time of putting a speaker in the position of a meter box. It is a figure which shows the path | route of the reflected sound at the time of placing a speaker in the back of a dashboard. It is a figure which shows the path | route of the reflected sound at the time of shifting a speaker from the position shown in FIG. It is a figure which shows the measurement result of the binaural response at the time of arrange | positioning a speaker in the position of a meter box. It is a figure which shows the measurement result of the binaural response at the time of arrange | positioning a speaker behind a dashboard. (A) is a figure which shows typically the state in which a listener exists in the right side of a left and right speaker, (b) is a figure which shows typically the state in which a listener exists in the middle of a left and right speaker. It is a side view which shows the example of arrangement | positioning of the speaker in the inside of a motor vehicle. It is a figure which shows the example of arrangement | positioning of the speaker in the vehicle interior of a motor vehicle. It is a graph which shows the frequency characteristic of the binaural response which measured the crosstalk cancellation effect at the time of directing the speaker of a meter box position to a driver. It is a graph which shows the frequency characteristic of the binaural response which measured the crosstalk cancellation effect at the time of putting a speaker in the back of a dashboard, and making it reflect on a windshield. It is a graph which shows the frequency characteristic of the binaural response which measured the crosstalk cancellation effect at the time of putting a speaker in a meter box position, and making it reflect above a windshield. It is a block diagram which shows the structure of the speaker arrangement | positioning design support system concerning this embodiment. It is explanatory drawing of the speaker in a speaker arrangement | positioning design support system. It is a flowchart figure which shows the speaker arrangement | positioning design support method concerning this embodiment. It is a flowchart figure which shows the process of the impulse response measurement step of an impulse response measurement part. It is a flowchart which shows the evaluation data calculation process in an evaluation data calculation part. It is a table which shows the evaluation data obtained by the evaluation data calculation process in an evaluation data calculation part. It is a table which shows the evaluation item by evaluation data. It is a flowchart which shows the speaker arrangement | positioning determination process in a speaker arrangement | positioning determination part. It is a flowchart which shows the evaluation data calculation process in an evaluation data calculation part.

(Preface)
The inventors have arranged a normal stereo speaker in the driver's seat in the vehicle, that is, in the front of the driver so that the driver is the center of the L channel speaker (Lsp) and the R channel speaker (Rsp). Even experienced that the normal stereo feeling is not obtained. The interior of an automobile is an environment that is asymmetrical with respect to the driver. For example, in the case of a right steering wheel, the distance from the driver to the right side glass is shorter than the distance to the left side glass. Therefore, the influence of the right side glass on the right ear of the driver is greatly affected, and the left and right reflections are uneven.

  The inventors have developed a sound field generation device of Japanese Patent Application No. 2011-089915 in order to generate a desired sound field. In this sound field generation device, the acoustic signal that reaches the driver after being reflected by the reflection unit with the speaker facing the reflection unit is reflected by the reflection unit and reaches the driver so that the acoustic signal that reaches the driver is equivalent to the acoustic signal from the virtual sound source. A crosstalk canceling unit that performs arithmetic processing so that the crosstalk between the channels of the acoustic signal is canceled at the position of the driver. However, it turned out that it is not only necessary to reflect.

  For example, as shown in FIG. 1, in a speaker arrangement in which a speaker is placed at a position in front of the meter box and the sound reflected above the windshield reaches the driver, the left localization is felt smaller than the desired angle. Compared to different types of cars, even with the same loudspeaker arrangement, a stable orientation was obtained with the wagon type, but with the hatchback type the sound image was blurred and felt smaller than the desired angle. From this, it can be understood that the localization of the sound image greatly affects the speaker arrangement.

[In-car sound field]
2A and 2B show impulse responses of both ears of the dummy head. FIG. 2A shows the measurement result in the laboratory (cuboid), and FIG. 2B shows the measurement result in the sealed vehicle. LspLe and LspRe are responses of the left ear (Le) and the right ear (Re) when the L channel speaker Lsp is driven with a pulse signal. RspLe and RspRe are responses of the left and right ears when the R channel speaker Rsp is driven with a pulse signal. For the dummy head, both speakers were installed. (A) In the laboratory, the response of the ear on the driven speaker side is large, whereas (b) there is a lot of reflected sound in the vehicle as a whole, and the right ear is more rich in reflected sound. Recognize.

  FIG. 3 is a diagram showing a route of arrival in which a speaker (sound source) arranged at the position of the meter box is arranged toward the driver and how sound reaches both ears (sound receiving points). In FIG. 3, two circles located at the upper part of the drawing indicate microphones attached to both ears of the dummy head, and two circles located at the lower part of the drawing indicate an L channel speaker and an R channel speaker. Is shown. Lines other than these indicate the shape and configuration inside the vehicle. It can be obtained by performing geometric acoustic simulation using the vehicle shape, the interior shape around the driver (dashboard, meter box, etc.), the sound absorption coefficient of the interior material, etc. as input data. The straight line from the speaker to the microphone in FIG. 3 is a direct sound, and the other lines indicate the 1st to 5th order reflected sounds that are reflected on glass or the ceiling. In the car, the reflected sound has more energy than the direct sound and affects how it is heard. Higher-order reflected sound reflects many reflecting parts, and the path to reach the receiving point is longer, so there is also sound absorption and distance attenuation at the reflecting part, and the energy is small, and it can be heard compared to lower-order reflected sounds. There is little impact on people. In a normal room, the way of arrival is left-right symmetric, whereas in a car, when the driver is seated on the right side, the driver is biased to the right side of the driver.

[About speaker placement]
Therefore, the inventors searched for the speaker arrangement in the vehicle in which the sound “arrival” is symmetrical in the same way as in the laboratory, by acoustic simulation. (Note: The “arrangement” of the speakers is not symmetrical.) 3 to 5 show three patterns for comparison.
1) A speaker placed at the above-mentioned meter box (Fig. 3)
2) A speaker placed behind the dashboard (Figure 4)
3) Placing a speaker in the back of the dashboard is the same as 2) but with a slight shift in position (Fig. 5)

  3 to 5 show the paths of the direct sound and the 1st to 5th order reflected sounds. The reflection of the right window as shown in FIG. 3 was observed at the position of the meter box. When the speaker was placed in the back of the dashboard (in front of the car), the influence of the right window was not seen as shown in FIG. Although the arrival path is not completely symmetrical, it can be said to be almost symmetrical. However, even in the back of the dashboard, the primary and secondary reflected sounds come again from the right window as shown in FIG.

  The measured responses of both ears are shown in FIGS. FIG. 6 shows the measurement results when a speaker is placed at the position of the meter box 1), and FIG. 7 shows the measurement results when the speaker is placed behind the dashboard 2). In FIG. 6, the amplitude level of LspRe is significantly higher than the others. This means that even if you play Lsp, you can hear it from the right. FIG. 7 shows that LspLe is higher when comparing LspLe and LspRe, and that RspRe is higher when comparing RspLe and RspRe, so that it can be heard from the reproduced side.

[Why is the path of reflected sound symmetrical?]
In the following, in order to assume ideal conditions, a time delay when only a direct sound such as an anechoic room is generated will be considered. As shown in FIG. 8A, the listener is not at a position equidistant from the left and right speakers but on the right side. The sound field generation device of Japanese Patent Application No. 2011-089915 reproduces a reverse phase signal for canceling the Rsp crosstalk signal (RspLe) from the left ear, but the path of LspLe is longer than RspLe because of the longer path than RspLe. Must be played in advance. To play in advance must be played before the warning sound presentation time. Since this is impossible, the reproduction from Rsp is delayed. However, since the warning sound is reproduced later than the time at which the warning is desired, there is a possibility that danger cannot be avoided. Therefore, it is ideal that this path difference is as small as possible, that is, it is ideal that it is symmetrical.

Next, consider the example in the car. As shown in FIG. 8B, even if the listener is at an equal distance from the left and right speakers, the crosstalk signal LspRe is delayed by a reflected sound larger than the direct sound of LspRe via the right window. The situation is the same as in the previous anechoic room. Furthermore, in order to cancel the large reflected sound with certainty, it is necessary to adjust the time so that the reverse phase signal of the same level is reproduced from Rsp and canceled at the Re position. However, since the reflected sound from the right window of RspRe arrives at Re again with a delay, a signal disturbing localization can be heard. It can be seen that the asymmetrical left / right as described above complicates the signal and makes it difficult to cancel the crosstalk.
From the above, it can be seen that it is preferable to select a speaker arrangement in which the reflected sound is as symmetric as possible. However, this does not necessarily require the arrival directions to be bilaterally symmetrical, as long as they arrive at almost the same time. This is because control is performed so that a desired signal is delivered to the position of both ears, that is, a point.

[Speaker orientation]
The geometric acoustic simulation is calculated with an omnidirectional sound source, but an actual speaker has directivity, and the in-vehicle sound field is a very narrow space. Considering the wave behavior, it is natural to think that many reflected sounds other than those seen in the simulation diagram have arrived. Then, just as if you were dealing directly with sound in an anechoic room, you created the most powerful, high-energy main reflected sound in the car, and took all the information on the transfer characteristics into the reflected sound. And weaken other reflected sounds. In this way, I thought that the same effect as an anechoic room might be obtained.

  Therefore, in FIG. 4 of the simulation result, the reflected sound reflected at a low position of the windshield is assumed as the main. For this reason, it was set so that the axis of the speaker was directed upward by 52 degrees and hit the reflection point of the windshield (FIGS. 9 and 10). FIG. 7 shows the responses of both ears of the dummy head in the driver's seat measured in such a state. When Lsp is driven, the amplitude of Le (LspLe) is larger than the amplitude of Re (LspRe), and when Rsp is driven, the amplitude of Re (RspRe) is larger than the amplitude of Le (RspLe), and there are many reflected sounds. It became a shape similar to the response of an anechoic chamber.

[Effect of sound localization]
Whether or not a sound image can be localized at a desired position can be determined by examining whether or not the effect of the crosstalk cancellation is exerted. That is, in Embodiment 2 of Japanese Patent Application No. 2011-089915, the Lch signal (Ls in Embodiment 2) is presented only to the left ear and the Rch signal (Rch in Embodiment 2) is presented only to the right ear. Check if it is made.

  In each of FIGS. 11 to 13, the left diagram shows frequency characteristics of binaural response when the Lch signal is driven as a pulse signal and the Rch signal as silence, and the right diagram shows the Lch signal as silence and the Rch signal. The frequency characteristic of the binaural response when driven as a pulse signal is shown. 11 to 13, LchLe is a measurement result of the left ear when the Lch signal from the left channel speaker is driven as a pulse signal and the Rch signal from the right channel speaker is driven as silence. LchRe is the measurement result of the right ear when the Lch signal from the left channel speaker is driven as a pulse signal and the Rch signal from the right channel speaker is driven as silence. RchLe is a measurement result of the left ear when the Lch signal from the left channel speaker is driven as silence and the Rch signal from the right channel speaker is driven as a pulse signal. RchRe is the measurement result of the right ear when the Lch signal from the left channel speaker is driven as silence and the Rch signal from the right channel speaker is driven as a pulse signal. In each figure, the horizontal axis represents frequency (Hz) and the vertical axis represents sound pressure level (dB).

  It can be determined that the larger the difference between the levels, the more effective the crosstalk cancellation. FIG. 11 shows the measurement results when the speaker at the meter box position is directed toward the driver. FIG. 12 shows the measurement results when the speaker is placed in the back of the dashboard and reflected by the windshield. FIG. 13 shows the measurement results when the speaker is placed at the meter box position and reflected above the windshield.

  In the measurement result of FIG. 11, the effect of the crosstalk cancellation is not obtained at 1 kHz to 3 kHz. However, in the measurement result of FIG. 12, the effect of the crosstalk cancellation is improved. In the measurement result of FIG. 13, compared with FIG. 11, the band that is not effective moves to a lower band.

  When actually auditioning, the localization effect was good in the order of the windshield reflection at the back of the dashboard of FIG. 12, the windshield reflection of the meter box of FIG. 13, and the front of the meter box of FIG.

[Foreword End]
As described above, in order to obtain the optimum sound image localization effect, the arrangement including the direction of the speaker is very important. However, since the in-vehicle sound field varies depending on the shape of the vehicle and the interior shape, the arrangement must be examined for each target vehicle. In other words, using the shape of the vehicle, the interior shape, the sound absorption rate of the interior material, etc. as input data, a plurality of speaker sound sources are placed, a geometric acoustic simulation is performed over several days, and the speaker position and inclination that take a symmetrical reflected sound path are determined. It is necessary to investigate. However, this is a laborious and time consuming operation. For this reason, there has been a demand for a design support apparatus that can easily obtain an optimum speaker arrangement. A speaker arrangement design support apparatus that determines an optimum speaker arrangement by actually measuring several minutes to several tens of minutes will be described.

(Embodiment)
FIG. 14 is a diagram illustrating a configuration of the speaker layout design support system 10. The speaker arrangement design support system 10 includes a speaker unit 100 including a speaker arrangement driving unit 120 and a speaker 110, a dummy head 210 including a dummy head 210 and microphones 220 respectively attached to both ears, and speaker arrangement design support. The apparatus 300 is configured. The speaker arrangement design support apparatus 300 according to the present embodiment is realized by a configuration including partial software using a personal computer or the like.

  In FIG. 14, the speaker arrangement design support apparatus 300 includes a control unit 310, a display unit 350, an operation unit 360, an amplification unit 370, and a storage unit 380 configured from a CPU (Central Processing Unit), a DSP (Digital Signal Processor) memory, and the like. Composed. The control unit 310 includes an impulse response measurement unit 311, an evaluation data calculation unit 312, a speaker arrangement determination unit 313, and a speaker arrangement control unit 314 that are realized by executing a program.

  The speaker arrangement driving unit 120 attached to the speaker 110 can freely control the rotation of the inclination of the speaker unit 100 using the driving force of a motor or the like under the control of the speaker arrangement control unit 314 (see FIG. 15). That is, when the speaker arrangement driving unit 120 is driven, the installation angle of the speaker 110 is controlled. The control of the speaker arrangement driving unit 120 may be automatically controlled based on a program, or the speaker arrangement control unit 314 may be controlled based on an operation input from the operation unit 360.

  The amplifying unit 370 amplifies the measurement signal output from the impulse response measuring unit 311 and outputs the amplified signal to the speaker 110. The microphone 220 is attached to both ears of the dummy head 210 installed in the driver's seat, and the measurement signal input to the microphone 220 is amplified by a microphone amplifier (not shown) and input to the impulse response measurement unit 311. . The microphone amplifier may be provided in the dummy head unit 200 or may be provided in the speaker arrangement design support device 300.

  FIG. 16 is a flowchart showing a rough flow of processing executed by the speaker arrangement design support device 300. The impulse response measurement (step S1) performed by the impulse response measurement unit 311 measures the binaural response at any speaker position (including changes in tilt). In the evaluation data calculation (step S2) performed by the evaluation data calculation unit 312, evaluation data is calculated from the binaural responses at all speaker positions. Speaker placement determination (step S3) performed by the speaker placement determination unit 313 compares the evaluation data obtained by the evaluation data calculation, and determines the speaker placement that is optimal for sound image localization.

  Details will be described. First, the user installs the dummy head unit 200 and the speaker unit 100 in the car. Here, the space in which the dummy head unit 200 and the speaker unit 100 are installed may be the same space as the interior of the automobile in which the speaker arrangement is actually designed, or a space equivalent thereto. The user installs the dummy head unit 200 incorporating the microphones 220 in both ears in the driver's seat. The speaker 110 to which the speaker arrangement driving unit 120 is attached is placed on the dashboard. The interior of an automobile is an asymmetric space with respect to a driver who is a listener.

  At this time, it is preferable to place the speaker unit 100 on the back side of the dashboard for the reason described above (refer to the preface). Further, since the initial angle of the speaker 110 is measured after changing the angle, it is easy to understand if it is level with respect to the ground. If the speaker unit 100 is replaced with another place and measured again, the best speaker position can be determined from all measured data. The position of the speaker unit 100 may be controlled by a computer, for example, by placing the speaker unit 100 on a rail.

  Since the sound field generation device of Japanese Patent Application No. 2011-089915 uses at least two or more speakers, it is necessary to measure at least two or more positions. Moreover, in this speaker arrangement design support apparatus, it is possible to provide outputs of 2ch or more and perform measurement in order.

  After the speaker unit 100 and the dummy head unit 200 are installed, impulse response measurement is performed (step S1). The detailed operation in the impulse response measuring unit 311 is shown in FIG. FIG. 17 is a flowchart showing processing in the impulse response measurement unit 311.

  When the impulse response measurement is started, first, the current angle of the speaker 110 is set to 0 degree, and the angle α = 0 is set (step S11). In this state, the speaker arrangement driving unit 120 is driven, and the impulse response is measured with the microphones 220 for both ears (step S12). In the present embodiment, the measured responses are referred to as spLe and spRe, respectively, with the speaker as sp, the left ear as Le, and the right ear Re.

  The impulse response is measured by driving a signal having a narrow pulse characteristic (15-20 μs, 50-100 V), averaging a number of times, and driving an M-sequence noise or pink noise signal. Cross-spectrum method, time-stretched pulse signal, which is obtained by measuring the frequency transfer characteristics from the power spectrum of the sound source and the cross power spectrum between the sound source and the sound receiving point using both the received signal and the drive signal, and measuring by inverse Fourier transform There is a time stretched pulse method in which a transfer function is obtained by convolution of a response signal and a drive signal whose time axis is inverted.

  The measured binaural responses spLe and spRe are stored in the storage unit 380 with a file name that allows the position and inclination of the speaker 110 to be known (step S13). For example, spLe and spRe are stored in a folder at A position of 0 degree.

  Next, the angle α is updated to 10 degrees (step S14). In this embodiment, the angle of the speaker 110 changed by the speaker arrangement driving unit 120 is set every 10 degrees. However, the change angle of the speaker 110 can be set by the speaker arrangement control unit 314 before the measurement. Here, when the measurement is terminated because the speaker 110 cannot be tilted any more (NO in step S15), the process proceeds to the evaluation data calculation in step S2 of FIG.

  When the measurement is continued (YES in step S15), the speaker arrangement control unit 314 controls the speaker arrangement driving unit 120 to rotate the speaker 110 upward by 10 degrees (step S16). Note that a range of angles to be measured may be specified in advance using the operation unit 360, and the speaker placement control unit 314 may automatically control until the measurement is completed.

  The impulse response is again measured with the angle of the speaker 110 changed (step S12). The measurement data is stored in the folder at the A position of 10 degrees so that the rotation angle can be understood as in the previous time. In this way, Steps S12 to S16 are repeated, and the binaural response is sequentially measured and stored while rotating the angle of the speaker 110.

  For example, when the angle α is measured from 0 ° to 180 °, 19 folders (A0, A1,..., A170, A180) are created for the A position. Furthermore, when it is desired to change the position of the speaker unit 100 and perform measurement, the speaker unit 100 is appropriately installed at a different B position, measurement is started from an angle α = 0, and the folder name is changed to B0,. Save. As described above, the impulse response measuring unit 311 measures the impulse response while changing the position of the speaker 110 or the angle of the speaker 110.

  When all N measurements of the target position and inclination are completed, the evaluation data calculation unit 312 next calculates evaluation data for determining the speaker arrangement (step S2 in FIG. 16).

The evaluation data calculation unit 312 includes a delay time calculation unit 3121 and an energy ratio calculation unit 3122 as shown in FIG. The evaluation data calculation unit 312 repeats the operation of the flowchart shown in FIG. 18 for each folder in order to calculate evaluation data for each target position, that is, for each folder. In the process described later, each symbol indicates the following value.
Vmax_s: The maximum amplitude value of the response on the driven speaker side in the binaural response Vmax_a: The maximum amplitude value of the response on the opposite side to the driven speaker in the binaural response Tmax_s: Time taken for Vmax_s Tmax_a: Take Vmax_a Time Cv = Vmax_s / Vmax_a
Dt = Tmax_s−Tmax_a
E0: Total energy before the maximum amplitude E1: Total energy after the maximum amplitude Re = E0 / E1

  When the processing of the delay time calculation unit 3121 is started, first, the direct sound having the first amplitude is searched for each binaural response, and the direct sound arrival time Tdl of spLe and the direct sound arrival time Tdr of spRe are respectively set. Calculate (step S21). When the speaker unit 100 is installed on the left side, Tdl <Tdr, and when it is on the right side, Tdl> Tdr. Therefore, the flag sp_flag is set by comparing Tdl and Tdr. For example, if Tdl <Tdr, sp_flag = 0, and if Tdl> Tdr, sp_flag = 1. Tdl = Tdr is set to sp_flag = 2 and stored in the storage unit 380 in association with the folder name (step S22).

  The binaural responses spLe and spRe can be determined as LspLe and LspRe when sp_flag = 0, so that they can be determined as LspLe and LspRe. In the subsequent steps, evaluation data is calculated as LspLe and RspRe for the ear response on the speaker side, and LspRe and RspLe for the response on the opposite side.

  Next, the maximum amplitude is searched, and the maximum amplitude value Vmax_s of the ear response on the speaker side and the maximum amplitude arrival time Tmax_s, the maximum amplitude value Vmax_a of the response on the opposite ear and the maximum amplitude arrival time Tmax_a are obtained and values are obtained. Save (step S23). Further, the maximum amplitude ratio Cv = Vmax_s / Vmax_a and the time difference Dt = Tmax_s−Tmax_a are calculated from the above values, and the values are stored in the storage unit 380 (step S24).

  Next, the energy ratio calculation unit 3122 calculates the total energy E0 before the maximum amplitude of each response, the total energy E1 after that including the maximum amplitude, calculates the ratio Re = E0 / E1, and stores the value It is stored in 380 (step S25). The evaluation data may be stored in a table as shown in FIG. That is, evaluation data is stored in a folder having the above folder name.

Using these calculated evaluation data, the speaker arrangement determining unit 313 determines an optimum speaker arrangement (step S3 in FIG. 3). The speaker arrangement determination unit 313 evaluates the evaluation items shown in FIG. 20 and determines an optimum speaker arrangement for sound image localization in the vehicle interior sound field. FIG. 20 shows the conditions for determining the optimum speaker arrangement. The binaural response at each position is evaluated in order from evaluation item 1. As described above (refer to the preface), in a vehicle interior sound field with a lot of reflection, the localization effect is high if the speaker arrangement is such that the reflected sound arrives symmetrically. Furthermore, it is ideal that the response in the laboratory shown in FIG. That is, an arrangement having the following features 1) to 3) must be determined.
1) The response of the ear driving the speaker has a larger maximum amplitude.
2) The main reflected sound (reflected sound having the maximum amplitude) reaches the ear immediately, and the level of extra reflected sound (including direct sound) reaching the main reflected sound is small and the number is small.
3) The sound arrives symmetrically.

  The speaker arrangement determining unit 313 first determines whether or not each binaural response is a target for selecting an optimal arrangement, and then evaluates and determines the combination of Lsp and Rsp. The flowchart of the process of the speaker arrangement | positioning determination part 313 is shown in FIG.21 and FIG.22.

  When the processing of the speaker arrangement determination unit 313 is started, first, the index is initialized to 0 (step S31). Then, using the evaluation data at A0 (A position 0 degree), it is determined whether the maximum amplitude of the ear response on the speaker side is larger than the maximum amplitude of the response on the opposite side (evaluation item 1), that is, Cv> 1. (Step S32).

  If Cv> 1 (YES in step S32), select_flag data indicating that it is a selection target of the optimum arrangement is set to 1 (step S33). If Cv ≦ 1 (NO in step S32), select_flag = 0 is set (step S35). The response with select_flag = 0 is excluded from the selection target of the optimum speaker arrangement. Then, the index is incremented (step S36), and the process proceeds to the evaluation of the response at the next position (folder).

  If select_flag = 1 is set in step S33, is the difference between the time of the maximum amplitude of the ear response on the speaker side and the time of the maximum amplitude of the opposite response negative (evaluation item 2), that is, Dt ≦ 0? Is determined (step S34). If Dt ≦ 0 (YES in step S34), the process proceeds to step S36. On the other hand, if Dt> 0 (NO in step S34), it is rewritten to select_flag = 0 (step S35). Thereby, it deviates from the object of optimal speaker arrangement. Thereafter, the index is incremented (step S36).

  The index incremented in step S36 is compared with N (N is an integer of 2 or more) (step S37). N is a number corresponding to the number of times of impulse response measurement. If index <N (YES in step S37), the process returns to step S32 to proceed to evaluation of a response at the next position. Then, for all N responses, the processes in steps S32 to S36 described above are executed to obtain select_flag.

  When select_flag is obtained for all N responses (NO in step S37), only for the response of select_flag = 1, the sum of the maximum amplitude arrival times of binaural responses, Tmax_s + Tmax_a, is sorted in ascending order, and the result is stored in memory 380 as sortT. (Step S38, evaluation item 3).

  Similarly, only for the response of select_flag = 1, Re = (total energy E0 before the maximum amplitude / total energy E1 including the maximum amplitude) is sorted in ascending order, and the result is stored as sortE (step S39, evaluation item) 4). If Re has a large value, it means that many reflected sounds exist before the main reflected sound, which is not preferable as a speaker arrangement for sound image localization.

  Responses ranked higher in the evaluation item 3 and the evaluation item 4 are candidates for selection of the optimal arrangement. Further, in order to give each response a superiority or inferiority, the rank is scored as it is, sorted in ascending order of the total score P, and the result is stored as sortP (step S40). In this embodiment, the rank is used as a score as it is, but it may be weighted.

  At this time, the operation unit 360 may be used to set the upper speaker position to be selected. For example, by looking at sp_flag, the same number of 0 (Lsp) and 1 (Rsp) may be selected. If the number of selection targets is too large, the upper half may be set as the target range. At this time, the response out of the selection target is updated to select_flag = 0. In the present embodiment, sortPh is stored with the upper half of sp_flag = 0 and the upper half of sp_flag = 1 as selection targets for optimal arrangement (step S41).

  Next, the combination of Lsp and Rsp is evaluated. From the selected response, the time Tmax_s that is the maximum amplitude of the ear response on the speaker side is approximately equal on the Lsp side and the Rsp side, and the values are small, that is, the reflected sound arrives early . One candidate is selected from each of sp_flag = 1 and sp_flag = 0, and the optimal arrangement of the left and right speakers is made.

  For this reason, Tmax_s with sp_flag = 0 is compared with Tmax_s with sp_flag = 1. In addition, it is good to sort in order with a small value in sp_flag = 0 and 1 respectively (step S42), and make the minimum Tmax_s Tmax_smin and leave a response within Tmax_smin + 0.2 msec as a selection target (evaluation item 5). The interior of the vehicle is very narrow, and it is difficult to control the sound field if the right and left objects are slightly shifted. Therefore, a response satisfying each Tmax_s <Tmax_smin + 0.2 msec is left as a selection target (step S43).

  In the next step, the time difference Dt between the time for taking the maximum amplitude of the ear on the speaker side and the time for taking the maximum amplitude of the ear on the opposite side is selected to be approximately equal on the Lsp side and the Rsp side (Evaluation Item 6). If the evaluation items 5 and 6 are satisfied, it can be said that the reflected sounds of Lsp and Rsp arrive symmetrically, so that the optimum speaker arrangement can be determined. That is, Dt with sp_flag = 0 is compared with Dt with sp_flag = 1. Then, it is preferable to sort in descending order of the absolute value of the difference between Dt (step S44) and leave a response whose absolute value of the difference between Dt is within 0.1 msec as a selection target. Therefore, a combination of sp_flag = 0 and 1 whose absolute value of the difference in Dt is within 0.1 msec is left as a selection target (step S45).

  The next step is to prioritize pairing. Since the combinations of sp_flag = 0 and 1 remaining in step S45 are sorted in ascending order of the absolute value of the difference in Dt, the optimal arrangement is arranged from the top. That is, the speaker arrangement is determined with priority given to the arrival of the reflected sound in both ears symmetrically (step S46). For example, the first combination is displayed on the display unit 350 as a folder name “A50, C30”. That is, the left speaker 220L and the right speaker 220R are optimally arranged in a pair of the A position of 50 degrees and the right speaker 220R of the C position of 30 degrees, respectively. Thereby, the optimal position and angle of the left and right speakers are determined respectively.

  If the selection target becomes 0 in step S45, a display is made on the display unit 350 so as to prompt the measurement at another position. This is because the combination of L and R is very important, and if the evaluation item is not satisfied, the localization effect cannot be expected.

  As described above (see the introduction), when the speaker is placed in front of the dashboard (driver side), the direct sound is received to a certain extent. Since a large reflected sound arrives behind the direct sound, Re = E0 / E1 takes a large value and is rarely the target as the optimum position. Since the windshield is curved, it does not go as calculated on the desk, but if the main reflected sound reflected on the windshield is delivered to both ears horizontally with the ground, the speaker should be at the back of the dashboard (front) ).

  Furthermore, due to the relationship between the windshield and both ear positions, if the angle formed by the windshield and the horizontal plane is greater than 45 degrees, and the angle is greater than 45 degrees, the speaker back is directed to the driver and the speaker unit is directed toward the windshield. . When the angle formed by the windshield and the horizontal plane is 45 degrees, the speaker faces directly above (perpendicular to the horizontal plane). As the angle between the windshield and the horizontal plane decreases from 45 degrees, the speaker should be directed from the top to the driver. The measurement time is further shortened by selecting the examination position with reference to the above. In addition, this embodiment is an example and is not limited to the evaluation items and numerical values shown in FIG.

  Normally, acoustic simulation is performed using vehicle shape and interior shape as input data. In this embodiment, the impulse response is automatically measured, and the optimum arrangement is determined from the measured data. For sound image localization in a car with a lot of reflected sound, automatic measurement and optimum arrangement are determined simply by placing a speaker. Normally, the input work and calculation time of geometric acoustic simulation, which took several days to several tens of days per vehicle type, can be greatly reduced. Since it is only necessary to determine an optimum arrangement for each type of automobile, the optimum arrangement can be easily determined.

  In the present embodiment, the optimal arrangement is determined only from the binaural response without measuring whether there is a localization effect. Accordingly, it is possible to determine a speaker arrangement with a high sound image localization effect without requiring time for verifying the localization effect. Furthermore, the optimum arrangement is conditioned and utilized for speaker arrangement determination. Therefore, a sound image position having a high localization effect can be obtained by outputting a signal obtained by performing acoustic signal processing so that sound can be heard from a desired position in the vehicle.

  Using the above speaker layout design support method, the speaker layout in the asymmetric vehicle can be optimized in a short time. Of course, it is also possible to determine the speaker arrangement for vehicles and spaces other than automobiles. A speaker position in the vehicle is determined based on the optimum speaker arrangement obtained by the above calculation. Thereby, it is possible to obtain a speaker system that can reliably localize a sound image even in a space asymmetric with respect to a listener.

  This speaker system has left and right speakers that are subjected to signal processing to localize a sound image at a desired position and are reflected from at least one speaker to a reflecting portion and delivered to a listener. The difference between the binaural responses related to the left and right speakers is substantially equal in the left and right speakers with a difference between (the maximum amplitude arrival time of the response of the ear on the speaker side) and (when the maximum amplitude of the response of the opposite ear is reached) The time when the maximum amplitude of the response of the ear on the speaker side is substantially equal in the left and right speakers, and in each speaker, the maximum amplitude value of the response of the ear on the speaker side is larger than the maximum amplitude value of the response of the ear on the opposite side, In each speaker, the speakers are arranged such that the difference between (maximum amplitude arrival time of the response of the ear on the speaker side) and (maximum arrival time of the response of the ear response on the opposite side) is negative. By doing so, it is possible to reliably localize a sound image even in an asymmetric space.

  Note that the above-described calculation may be executed by a computer program. The arithmetic program described above can be stored using various types of non-transitory computer readable media and supplied to a computer. Non-transitory computer readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (for example, flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (for example, magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / W, semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (Random Access Memory)) are included. The program may also be supplied to the computer by various types of transitory computer readable media. Examples of transitory computer readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire and an optical fiber, or a wireless communication path.

  In addition to the case where the function of the above-described embodiment is realized by the computer executing the program that realizes the function of the above-described embodiment, this program is not limited to the OS ( The case where the functions of the above-described embodiment are realized in cooperation with the Operating System) or application software is also included in the embodiment of the present invention.

  Although the present invention has been described with reference to the exemplary embodiments, the present invention is not limited to the above. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the invention.

  DESCRIPTION OF SYMBOLS 100: Speaker part, 110: Speaker, 120: Speaker arrangement drive part, 200: Dummy head part, 210: Dummy head, 220: Microphone, 300: Speaker arrangement design support apparatus, 310: Control part, 311: Impulse response measurement part 312: Evaluation data calculation unit 313: Speaker arrangement determination unit 314: Speaker arrangement control unit 350: Display unit 360: Operation unit 370: Amplification unit 380: Storage unit

Claims (6)

A speaker whose arrangement state can be changed and which outputs a measurement signal;
A dummy head device comprising a dummy head and a microphone mounted on each of left and right ears of the dummy head;
An impulse response measurement unit that measures an impulse response based on the measurement signal output from the speaker and input from the microphone, and an evaluation for determining the placement of the speaker based on the impulse response measured in the impulse response measurement unit A speaker arrangement comprising: an evaluation data calculation unit that calculates data; and a speaker arrangement determination unit that determines the arrangement of the speakers where a sound image is localized at a desired localization position based on the evaluation data calculated by the evaluation data calculation unit A speaker arrangement design support system comprising a design support device. The speaker arrangement design support device further includes a speaker arrangement control unit that controls the direction of the speaker,
The speaker further includes a speaker arrangement driving unit capable of changing the direction of the speaker under the control of the speaker arrangement control unit,
The speaker arrangement design support system according to claim 1, wherein the impulse response measurement unit measures the impulse response for each direction of the speaker controlled by the speaker arrangement control unit. An impulse response measurement unit that measures an impulse response based on the measurement signal that is output from a speaker that outputs a measurement signal and is input from a microphone installed at a listening position;
An evaluation data calculation unit that calculates evaluation data for determining the arrangement of the speakers based on the impulse response measured in the impulse response measurement unit;
A speaker arrangement design support device, comprising: a speaker arrangement determining unit that determines an arrangement of the speakers in which a sound image is localized at a desired localization position based on the evaluation data calculated by the evaluation data calculation unit. The evaluation data calculation unit includes:
A delay time calculation unit that calculates a direct sound arrival time difference, a maximum amplitude arrival time difference, and a maximum amplitude value ratio in the measurement signals input to the microphones attached to the left and right ears;
An energy ratio calculation unit that calculates an energy ratio before and after the maximum amplitude in the measurement signal input to each of the microphones attached to the left and right ears;
The speaker arrangement design support apparatus according to claim 3, wherein the evaluation data is calculated based on the arrival time difference and the energy ratio. An impulse response measuring step of measuring an impulse response based on the measurement signal output from a speaker that outputs a measurement signal and input from a microphone installed at a listening position;
An evaluation data calculation step for calculating evaluation data for determining the arrangement of the speakers based on the impulse response measured in the impulse response measurement step;
A speaker arrangement design support method comprising: a speaker arrangement determination step for determining an arrangement of the speakers where a sound image is localized at a desired localization position based on the evaluation data calculated in the evaluation data calculation step. In the computer provided in the speaker arrangement design support device,
An impulse response measuring step of measuring an impulse response based on the measurement signal output from a speaker that outputs a measurement signal and input from a microphone installed at a listening position;
An evaluation data calculation step for calculating evaluation data for determining the arrangement of the speakers based on the impulse response measured in the impulse response measurement step;
A speaker arrangement determining step for determining an arrangement of the speaker in which the sound image is localized at a desired localization position based on the evaluation data calculated in the evaluation data calculation step;
A program that executes