JP2003023689A - Variable directivity ultrasonic wave speaker system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate arrangement of elements in an ultrasonic wave speaker that employs a piezoelectric element ultrasonic wave transducer to direct a beam. SOLUTION: An ultrasonic wave modulator modulates an ultrasonic wave signal with a central frequency (f) on the basis of an input of an audible signal. Many delay lines delay the modulated ultrasonic wave signal in order to direct a beam pattern toward an angle θ0 . A speaker outputs the ultrasonic wave signal by directing the beam pattern toward the angle θ0 . The ultrasonic wave speakers are placed so that a center distance (d) is greater than half the wavelength λ/2 of the ultrasonic wave. Thus, the beam pattern has a main lobe and at least one grating lobe. The main lobe has a mode K and the grating lobe has an attenuation Ga with respect to the attenuation of the main lobe. Expression (1) holds between the values d, f, K, Ga, and θ0 , where c is a sound velocity.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to an ultrasonic speaker system.

[0002]

2. Description of the Related Art In a conventional speaker that generates a sound wave, the generated sound wave is emitted in various directions. Although it is theoretically possible to use a speaker array to form a sharp and focused beam of sound waves, a very large number of speakers will be used. In addition, since the sound wave generated from the speaker has a relatively long wavelength in the range of 1.74 to 17.40 m, the speaker array has a large space due to the nature of the wave motion. Will occupy.

On the other hand, an ultrasonic transducer (speaker) array can form a beam with a narrow width. The ultrasonic waves are AM modulated by the audio signal and used to operate the ultrasonic transducer array to form an AM modulated ultrasonic beam. While propagating in the air, the AM-modulated ultrasonic beam is demodulated by the nonlinear propagation effect in the air, and only audible sound having a frequency other than the frequency band of the ultrasonic wave remains. Parametric arrays use this phenomenon to create a "sound spotlight" so that only those who are in the sound spotlight can hear it.

[0004] EP 0 973 152 A2 proposes reflecting an ultrasonic beam of narrow width by means of a rotatable reflecting surface. The ultrasonic beam is focused to a desired position by rotating the reflecting surface. When the ultrasonic beam is reflected toward the second surface, the ultrasonic frequency is absorbed and the audible sound frequency is reflected at the second surface, and the audible sound can be heard from the direction of the second surface. If the second surface is a movie screen, the reflective surface can be rotated to cause the movie character across the screen to follow the audible position.

European Patent No. 0 973 152 A
No. 2 discloses that a curved surface that absorbs ultrasonic frequencies and reflects audible sound frequencies can be used, and that this configuration directs the reflected audible sound to a specific viewing position.

European Patent No. 0 973 152 A
No. 2 discloses a transducer array having a large number of ultrasonic transducer modules, which are configured in two dimensions or three dimensions. A network adds variable correlated phases to the signals applied to the transducer modules to facilitate electrical focusing and steering. Since the signals are broadband, delays can be used to direct the beams from the array.

[0007]

The present inventor has tried to manufacture an ultrasonic speaker system capable of directing a beam by using a piezoelectric element type ultrasonic transducer. The piezoelectric element type ultrasonic transducer has a diameter in the range of 1 to 4 cm. When attempting to incorporate ultrasonic transducers into the array, the inventor found that it was very difficult to place the ultrasonic transducers at standard half wavelengths. This is because the half-wave length of ultrasonic waves is very short and is on the millimeter level. The inventor has found that such an arrangement is difficult to achieve in order to direct the ultrasonic beams generated by the ultrasonic transducer array. The elements of the ultrasonic transducer must be very small and must be separated by a very short distance.

An object of the present invention is to solve the above problems and to provide an ultrasonic speaker system capable of providing a degree of freedom of speaker arrangement in an array.

[0009]

An ultrasonic speaker system according to the present invention has an ultrasonic modulator, a large number of delay lines, and a large number of ultrasonic speakers connected to the delay lines in a one-to-one correspondence. . The ultrasonic modulator modulates the ultrasonic signal having the central frequency f based on the inputted audible signal and outputs the modulated ultrasonic signal. The delay line receives the modulated ultrasonic signal from the ultrasonic modulator, and ultrasonic waves are added to the modulated ultrasonic signal with different delays to direct the beam pattern to the angle θ ₀ . Output a signal. The ultrasonic speaker outputs ultrasonic waves in a state in which the beam pattern is directed to the angle θ ₀ based on the modulated ultrasonic signal delayed by the delay line. The beam pattern has a main lobe and at least one grating lobe by arranging the ultrasonic speaker so that the distance d between the centers thereof is larger than the half wavelength λ / 2 of the ultrasonic wave. The main lobe has a mode number K, and the grating lobe closest to the main lobe has an attenuation G _a with respect to the attenuation of the main lobe. The following relationship is established among the values of the distance d, the frequency f, the number of modes K, the attenuation G _a , and the angle θ ₀ , where c is the sound velocity. There is.

Since the distance d between the centers of the ultrasonic speaker is larger than the half wavelength λ / 2 of the ultrasonic wave, a very sharp beam can be formed and the sound can be directed to a selected viewer. it can.

The inventor of the present invention has recognized that the distance d, the frequency f, the number of modes K, the damping G _a , and the value of the angle θ ₀ have the relationship as shown in the claims. . The loudspeaker system designer specifies the allowable level of attenuation G _a and solves the above relationship to easily calculate the loudspeaker radiation pattern by solving the following equation for the number of modes K: Can be used.

The designer substitutes the number of modes K into the following equation to determine the radiation pattern required of the loudspeaker to suppress grating lobes.

Since the delay line adds a delay to the ultrasonic signal modulated before the ultrasonic wave is output from the ultrasonic speaker, the delay line should be controlled electrically rather than physically controlled. You can

It is preferable to provide a convex reflecting plate on the propagation path of the main lobe to widen the angle of pointing.
Since the convex reflecting surface is used, it is possible to widen the limit range of the desired angle, for example, from several degrees to 10 degrees or more.

Further, in order to reduce the distortion of the output audible sound, it is preferable to have a preprocessor for processing the audio signal input to the ultrasonic modulator.

The ultrasonic speaker is preferably of the piezoelectric element type. Piezoelectric ultrasonic speakers are not expensive. However, since the sizes are considerably large such as 1 to 4 cm, it is difficult to bring them close to each other. In the present invention, the piezoelectric element type ultrasonic speaker, which is not expensive, can be used because it is arranged at an interval larger than the half wavelength λ / 2.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION The above-mentioned objects of the present invention, other objects,
The features and advantages will become apparent by understanding the description of the following embodiments with reference to the accompanying drawings.

An ultrasonic speaker system 10 according to an embodiment of the present invention will be described with reference to the accompanying drawings. As shown in FIG. 1, the system 10 includes an ultrasonic speaker array 20, a convex reflector 30, and a signal processor 40.
And an input unit 50. The ultrasonic speaker array 20 is composed of five piezoelectric element type ultrasonic speakers 21a to 21e.

As shown in FIG. 2, the input unit 50 is
It has an audio signal input unit 51 and a desired angle θ ₀ input unit 52. The audio signal input unit 51 inputs an audio signal represented by an audible sound. The desired angle θ ₀ input unit 52 inputs data represented by the desired angle θ ₀ for specifying the beam direction, and outputs the data to the signal processor 40.

The signal processing circuit 40 includes a preprocessor 4
1, an ultrasonic modulator 42, and a variable delay unit 4
3 and an ultrasonic signal source 44. The preprocessor 41 performs pre-processing of the audio signal from the audio signal input unit 51 in order to reduce the distortion of the output audible sound. The preprocessor 42 is The Jour
nal of Audio Eng. Soc. ，
Vol. 47, No. 9.1999 Septemb
ER, "Use of Ultrasonic Waves in Air for General Audible Sound Beam". Joseph
It operates according to the method of Pompei. See the same publication for this content.

The ultrasonic signal source 44 outputs an ultrasonic signal to the ultrasonic modulator 42. Ultrasonic signal is 20KH
It has a central frequency f higher than z. The ultrasonic modulator 42 modulates the ultrasonic signal based on the audio signal input processed by the preprocessor 41 and outputs the modulated ultrasonic signal x (t) to the variable delay unit 43.

The variable delay unit 41 has a plurality of adjustable delay lines 43a to 43e. Delay line 43
a~43e respectively receive the modulated ultrasonic signal from the ultrasonic modulator 42, the desired angle theta ₀ to the desired angle theta ₀ inputted in the input unit 52 to direct the beam, different delays each additional The modulated and modulated ultrasonic signal is output. The delay time is calculated for each speaker 21a to 21e based on the desired beam angle θ ₀ and set for the variable delay circuit.

As shown in FIG. 3, each speaker 21 of the array 20 has a sound pressure directivity of several degrees to 10 degrees.
As shown in FIG. 4, the array 20 is configured by a one-dimensional array including five ultrasonic speakers 21a to 21e arranged in a straight line. 5 ultrasonic speakers 21
The a to 21e are connected to the delay lines 43a to 43e in a one-to-one correspondence. Therefore, the speakers 21a to 21e
Generates different ultrasonic waves based on the modulated ultrasonic signal delayed by the delay lines 43a to 43e. As shown in FIGS. 5, 6, and 7, the ultrasonic waves are directed at a desired angle θ ₀ (in FIGS. 5, 6, and 7, θ ₀ is −3, ₀ , and +5, respectively). Combined to generate an ultrasonic beam pattern 70 having lobes 71. Since the ultrasonic beam propagates in the air,
An audible sound is generated due to the nonlinear effect of air.

The ultrasonic speakers 21a to 21e are arranged such that the distance between their centers is d. d is a half wavelength λ of the ultrasonic wave generated from the ultrasonic speakers 21a to 21e.
Greater than / 2. The distance d between the centers is the half wavelength λ of the ultrasonic wave.
Since it is larger than / 2, the arrangement of the speakers 21a and 21e in the array 20 can be made extremely easy. Also, as shown in FIGS. 5, 6 and 7, the resulting beam pattern 70 has a preferred narrow width main lobe 71. The sharp main beam allows it to be aimed at a more specific audience. On the other hand, FIG. 8 shows a beam pattern generated when the distance d between the centers is equal to the half wavelength λ / 2 of the ultrasonic wave. Note that the main lobe is wide.

The main lobe 71 has a mode number K which affects the width of the rope 71. That is, FIG.
As shown in, the larger the value of K, the narrower the beam width.

As shown in FIGS. 5, 6 and 7, when the distance d between the centers is larger than the half wavelength λ / 2 of the ultrasonic wave,
The resulting beam pattern 70 has at least one grating lobe. Main robe 7
The grating lobe 72 close to 1 is the main lobe 7
It has a damping Ga for a damping of 1. All other grating lobes have much greater attenuation.

As will be described later, between the distance d between the centers, the frequency f, the number of modes K, the attenuation Ga, and the angle θ ₀ ,
It has the following relationships. Here, c is the speed of sound.

FIG. 10 shows a beam pattern generated from a speaker array in which the distance d between the centers of the individual speakers is 1.5 cm. These speakers generate ultrasonic waves with a frequency of 55 to 65 KHz, that is, a half wavelength λ / 2 of 2.6 to 3.1 mm. Therefore, the distance d between the centers is the half wavelength λ / 2 of the ultrasonic wave.
Much larger than Therefore, the grating lobes 72 are generated on the left side or the right side of the main (0) beam 71, and these are oriented in the direction perpendicular to the speaker surface.

In the example shown in FIG. 10, it is assumed that the sound waves are omnidirectional and are generated in all directions and have the same amplitude and phase. However, in reality, the ultrasonic speaker array has the sound pressure directivity characteristic SPD in a narrow range as shown by the arc-shaped line in FIG. Therefore, the actual beam pattern is as shown in FIG. That is, this beam pattern is obtained by combining the beam pattern of FIG. 10, which is assumed to be omnidirectional, and the sound pressure directional characteristic SPD of the speaker itself. The grating lobe is
By properly using the sound pressure directivity of the loudspeaker itself, it can be reduced, thus the beam is integrated in only one direction and the grating lobe can be eliminated.

The direction of the sound wave from the speaker array 20 is
By changing the delay time in the delay lines 43a to 43e, it can be swung to the left or right by the angle θ ₀ .
However, the range of the variable angle is limited by the sound pressure directivity characteristic of the speaker as shown in FIG. 5 or 7. In some cases, the radiation pattern of the ultrasonic array 20 limits it to a few degrees to 20 degrees. Also, the range of angles that can be pointed is limited by the range that the variables used in equation (1) can actually take. By providing the convex reflecting plate 30, it is possible to widen the range of the variable angle of the beam.

Here, an application example of the ultrasonic speaker system according to the present invention will be shown.

FIG. 11 shows an example of a hearing-impaired television viewer watching television with a person having normal hearing in one room. The loud audible sound beam is aimed at a hearing-impaired viewer. Since the beam is louder than the normal omnidirectional radiation of sound, it is easier for hearing-impaired viewers to hear. However, since the beam has a sharp directivity, it does not disturb surrounding television viewers.

FIG. 12 shows a group watching a television program that includes Japanese and English tracks. The beam of the English track is aimed at one of the group wishing to watch the English version, while the other is watching the Japanese version of the omnidirectional sound.

FIG. 13 shows a state in which surround sound is realized by a conventional method by using five conventional speakers and a subwoofer. Figure 14
It is shown that surround sound is realized by generating a highly directional sound beam from one speaker according to the present invention. By reflecting the beam off the wall as needed, the viewer can hear sound from five different directions. The beam direction can be easily adjusted by taking the shape of the room into consideration.

The basic theory behind the present invention will now be described.

Each of the adjustable delay lines 43a to 43e can be represented by Tn, where N is the number of speakers in the speaker array 20 and n = 1, 2, ..., N.
N is 5 in the examples. Adjustable delay line T
n is determined by the desired angle θ ₀ of the main lobe of the directional beam pattern shown below.

The fixed delay T ₀ is Is required. Excluding T ₀ , θ ₀
This is because a negative delay is obtained for a negative value of, but such a value cannot be satisfied. The signal received in the far field in the direction θ (−90 <θ <+90) is be equivalent to. Here, X _n (t) is the signal generated from the speaker n, τ _n is the delay caused by the difference in the distance between each element and the viewer, and A (θ) is each element and the propagation path. Is the total gain of.

The time delay τ _n in the example of FIG. 2 is be equivalent to. Where τ0 is a constant transmission delay of the first speaker x ₁ (t),
It is independent of the reception direction θ in the far field.

The gain A (θ) can be divided into two parts as follows. Here, A ₁ (θ) is the gain of the element depending on the angle, and A ₂ is the attenuation depending on the distance.

Substituting equations (4) and (5) into equation (3), To get Here, α ₀ = T ₀ + τ ₀ .

The frequency domain is Is.

From equation (7), it can be seen that beam patterns are formed in different directions at different frequencies. Figure 8
FIG. 4 is a graph showing a directional pattern formed by a beamforming speaker array with a center-to-center distance d equal to half-wavelength λ / 2 for 11 frequencies evenly distributed between 55 and 65 KHz. The example of FIG. 8 shows the amplitude normalized in equation (7), and the following conditions are substituted into equation (7). That is, the AM signal has a center frequency f of 60 KHz and a bandwidth of 10 K
And the far field reception direction θ ₀ is 10 °,
-90 <θ <+90, the number N of speakers is 10, the distance d between the centers of the speakers is 2.6 mm, the sound velocity c is 340 m / s, and the gain A ( θ) is equal to the distance-related attenuation A ₂ . It is considered that when the desired beam angle θ ₀ is the same as the far-field reception direction θ, the signal-to-noise ratio dB is the same at all frequencies, and the signal-to-noise ratio is completely independent of frequency. It is considered that the signal-to-noise ratio dB increases depending on the frequency as the difference between the beam angle θ ₀ and the far-field reception direction θ increases. However, the beam former has a fractional bandwidth of 10/6.
0, that is, 17%. For otherwise identical messages, only unmodulated, the fractional bandwidth is 5 / 2.5, or 200%. Therefore, since the ratio of the center frequency to the bandwidth is so large that it is difficult to form a beam, even if one beamformer is used for an unmodulated audio signal instead of a modulated ultrasonic wave. Useless. There is a need for beamformers that utilize more complex filters. Moreover, the distance d between the centers of the speakers is 2.6 mm, which makes it extremely difficult or possibly impossible to actually make speaker arrays.

The speaker array can be made more easily by increasing the distance d between the centers of the speakers than the half wavelength λ / 2. Also, by making it large in this way, a sharp main beam is obtained. Since the main beam is sharp, it is possible to direct it to a more specific viewer. On the other hand, by increasing the size in this way, the place where the viewer of the sound emitted from the speaker is usually located, that is, -9
At angles in the range 0 <θ <+90, undesirable grating lobes are formed.

In this way, the grating lobe
The angle and its presence are determined. Equation (7) is a corner
The gain A (θ), which has a degree dependency, decreases with distance.
Decline A _TwoFar-field reception direction θ
Is the desired beam angle θ₀Can be rewritten when
Wear. In other words, equation (7) shows that the main lobe is perfect.
Rewrite the equation
You can

Further, the relationship of the equation (8) holds for other angles θ _g which are grating lobes. In order to calculate the angle θ _{g of the} grating lobe, it is first determined from equation (7) as follows:

Equation (9) is now calculated for the grating lobe angle θ _g as follows: Here, m = ± 1, ± 2, ... Therefore,

The first grating lobe is obtained by setting m = ± 1. It can be seen that the beam-forming network does not generate grating lobes if there is no grating lobe angle θ _g with a value of −90 <θ <+90. The desired beam angle θ ₀ is ±
When taking a value of 90, the worst grating lobe occurs. From equation (11), the following relationship is determined.
That is, when the distance d between the centers of the speakers is smaller than the half wavelength λ / 2, the grating lobe does not occur.

[0048] The other is

It is interesting to note that the angle θ _{g of the} grating lobe is not a function of the number N of loudspeakers, but entirely on the distance d between the loudspeaker centers.

The distance d between the centers of the speakers is a half wavelength λ / 2.
It shows how a grating lobe appears when the value is larger than. FIG. 10 is a graph showing the same state as in FIG. 8, but the distance d between the centers of the speakers is 15 mm, and the distance d between the centers of the speakers is larger than the half wavelength λ / 2. As a result of increasing the distance between the speakers in this way, a grating lobe is generated. Frequency f of 60 KHz
The closest grating lobe to 33.5 to -1
It lies between 1.8 degrees and does not match equation (11).
The more the far-field reception angle θ differs from the desired angle θ ₀ , the more the beam pattern becomes more dependent on the frequency.

It should be noted that the main beam 71 shown in FIG. 10 is much narrower than the main beam shown in FIG. The equivalent equation can be easily derived. Inter-null beam widt
h) INBW is defined by the two closest null beams around the desired beam angle θ ₀ . Assuming that the intermediate null beam width INBW is zero, equation (7) establishes the following relationship. Here, m = ± 1, ± 2, ...

Two angles θ around the desired beam angle θ _0
1 and θ ₂ can be calculated from the equation (13) by setting m to m = + 1 and m = −1, and are as follows.

In this way, the two closest angles θ ₁ ,
The intermediate null beam width for θ ₂ is INBWΔθ =
θ ₂ θ _1, which is determined as follows.

[0054] Then, it is clear that the null beam does not appear on the left side or the right side of the desired beam angle θ ₀ . As a special case, the desired beam angle θ ₀ may be zero, in which case the intermediate null beam width INBW ₀ is Becomes

That is, it can be understood that by increasing the distance d between the centers of the speakers, the value of the intermediate null beam width INBW becomes smaller and a sharp beam is formed. Under the conditions used in the examples of FIGS.
You can easily try 6). Under these conditions, the intermediate null beam width IN in the example of FIGS.
The values of BW can be calculated as 25.6 and 4.4 degrees, respectively.

From the above description, it can be understood that a sharp beam can be obtained by increasing the distance d between the centers of the speakers, but many grating lobes are generated. Next, the speaker 21 of the speaker array 20
The directional pattern of will be described in more detail. The loudspeakers are each directional and have a normalized gain A with an angle dependence, as indicated by the following radiation pattern:
_{It has 1} (θ).

Where K is the radiation lob
It is the number of modes in e).

Ultrasonic speakers (transducers) having various frequency characteristics and directivity are produced. The different radiation patterns of the ultrasonic speaker are given by the equation (1
It can be roughly expressed by appropriately changing the number of modes K in 8). FIG. 9 shows the radiation pattern formed by one ultrasonic transducer (angle-dependent normalized gain A ₁ when the number of modes K is changed to several different values).
(Θ)) is shown.

Substituting equation (18) into equation (7) and calculating equation (7) assuming that the far-field reception angle θ is equal to the grating lobe θ _g , equation (10) yields m = ± 1. The following results are obtained.

Since it can be assumed that the element has directivity, the far-field reception direction θ ₀ and the grating lobe angle θ _g are respectively determined. Attenuation such as occurs. The required attenuation G _a dB at the angle of the closest grating lobe, relative to the attenuation at the desired beam angle θ _0, can be calculated using equation (1). That is,

The positive and negative signs (±) in the equation (20) are used for the negative and positive values in the far field reception direction θ, respectively.

The inclusion of the center frequency f in most equations indicates that the calculations and their results are frequency dependent. When required, the center frequency f of the carrier of the signal can be used as the estimation parameter, such as the number of modes K in equation (20).
For example, at the nearest grating lobe angle in the far field reception direction θ _{0 with} a value of ± 5,
When 30 dB is allowed as the value of the attenuation Ga, the value of the number of modes K is 85.7 in the array shown in FIG.
Is required. In FIGS. 5, 6, and 7, the number of modes K is 8
In the case of the value of 5.7, the whole directivity pattern of the speaker array when the desired beam angle θ ₀ is −3, _{0, and 5} is shown. By setting the mode number K to a high value, the attenuation G _{a of the} grating lobe increases, but at the same time, the range of the angle θ ₀ useful as the beam directivity narrows. Damping G _a and useful angle θ ₀
There is a trade relationship with. If the number of modes K is fixed and the desired attenuation G _a for the main lobe attenuation is set, then by trial and error to determine the range of useful angles θ ₀ for the corresponding speaker array, etc. Equation (20) can be solved for the desired beam angle θ ₀ .

Although the present invention has been described in detail with reference to specific embodiments, the present invention can be variously modified and improved within the scope described in the appended claims.

For example, in the present embodiment, the speaker array 20 has five rows of ultrasonic speakers.
The same effect can be obtained even with a 10-row ultrasonic speaker.

A two-dimensional speaker array 20 'as shown in FIG. 15 may be used instead. Speaker array 20 '
Has a speaker 21 of n rows and m columns. In this case, the signal processor 40 'has a delay line for each row of speakers, and the signal from each delay line is sent to all the speakers in the corresponding row. The vertically arranged speakers in each row increase the power of the sound waves generated.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing an ultrasonic speaker system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing the ultrasonic speaker system of FIG.

FIG. 3 is a conceptual diagram showing the sound pressure directivity of each ultrasonic speaker provided in the ultrasonic speaker system.

FIG. 4 is a conceptual diagram showing sound pressure directivity of an ultrasonic speaker array provided in the ultrasonic speaker system.

FIG. 5 is a graph showing an ultrasonic wave generated from a speaker array in which the distance between the centers of the speakers is larger than a half wavelength, in order to generate an ultrasonic beam pattern having a main lobe directivity with a narrow value of −3. The figure shows a state in which ultrasonic waves are combined, and also shows the range within the range allowed by the sound pressure directivity of the speaker array.

FIG. 6 is a graph showing ultrasonic waves emitted from a speaker array combined to generate an ultrasonic beam pattern with a narrow main lobe directivity of zero value.

FIG. 7 is a graph showing ultrasonic waves generated from a speaker array, showing a state in which ultrasonic waves are combined to generate an ultrasonic beam pattern having a main lobe directivity with a narrow value of +5; It shows the range within the range allowed by the sound pressure directivity of the speaker array.

FIG. 8 is a graph showing the directional patterns of the delayed beamformer for different frequencies when the speakers are arranged at half-wavelength intervals.

FIG. 9 is a graph showing a radiation pattern of one ultrasonic speaker when the number of modes is different.

FIG. 10 is a graph showing a state under the same conditions as in FIG. 8 except that the speaker spacing is larger than a half wavelength, and assuming that the speaker array is omnidirectional.

FIG. 11 is a perspective view showing a state in which the present invention is applied to selectively increase the volume of a television program for one person in one group.

FIG. 12 is a perspective view showing a state in which the present invention is applied to allow one of the groups to selectively hear sub-channel audio, for example, an English audio track.

FIG. 13 is a perspective view showing a configuration of a conventional surround.

FIG. 14 shows a surround configuration according to the present invention.

FIG. 15 is a conceptual diagram showing another embodiment of the present invention in which the speaker array is two-dimensional rather than one-dimensional.

[Explanation of symbols]

10 Ultrasonic speaker system 21a-21e ultrasonic speaker 30 convex reflector 41 preprocessor 42 Ultrasonic modulator 43a to 43e Delay line

Continued front page    (72) Inventor Mohammad Gabami             3-14-13 Higashi Gotanda, Shinagawa-ku, Tokyo Stock             Ceremony Company Sony Computer Science Research             In-house F-term (reference) 5D019 AA02 BB18                 5D020 AC11

Claims (3)

[Claims] 1. An ultrasonic modulator for modulating an ultrasonic signal having a center frequency f based on an input audible signal and outputting the modulated ultrasonic signal, and the modulated ultrasonic wave from the ultrasonic modulator. A plurality of delay lines for receiving the ultrasonic signal and outputting the ultrasonic signal with different delays added to the modulated ultrasonic signal to direct the beam pattern to the angle θ ₀ . And a plurality of ultrasonic speakers connected to the delay line in a one-to-one correspondence, the speaker forming a beam pattern based on the modulated ultrasonic signal delayed by the delay line by an angle θ _0.
The beam pattern is arranged so that the ultrasonic wave is output in a state of being directed to and the distance d between the centers is larger than the half wavelength λ / 2 of the ultrasonic wave, so that the beam pattern has a main lobe and at least one grating. , The main lobe has a mode number K, the grating lobe closest to the main lobe has an attenuation G _a with respect to the attenuation in the main lobe, and c is the speed of sound and the distance d is , The frequency f, the number of modes K, the attenuation G _a , and the value of the angle θ ₀ , the following relationship: An ultrasonic speaker system characterized in that there is. 2. The ultrasonic speaker system according to claim 1, wherein the propagation path of the main lobe is provided with a convex reflecting plate for widening a pointing angle. 3. The ultrasonic wave according to claim 1, further comprising a preprocessor for processing the audio signal input to the ultrasonic modulator in order to reduce the distortion of the output audible sound. Speaker system.