JP2004135023A - Sound outputting appliance, system, and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable optimum sound image setting with rear speakers and meet the movements of the listener's head. <P>SOLUTION: A sound signal outputted from a speaker behind a listener is corrected according to a head sound transfer function and then a sound image is localized at a specified position in a specified direction, e.g. in the space in front of the listening position. Further, an optimum head sound transfer function is selected for correction while the position relation between the speaker and the head of the listener is detected. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an audio output device, an audio output system, and an audio output method capable of localizing a virtual sound image in an arbitrary space in an AV playback device (audio / visual playback device), a multimedia playback device, and the like. Things.
[0002]
[Prior art]
2. Description of the Related Art In recent years, in the audio / video field and the multimedia field, it is desired to reproduce an audio / video having a three-dimensional effect and a sense of reality, and hardware corresponding to this is being actively developed. In particular, in sound reproduction, the sound signal can be heard from the side, the rear, or the upper or lower side without arranging the speaker in the side or the rear of the listener, or in the direction above or below, or a combination thereof. A sound reproducing apparatus for localizing a virtual sound image has attracted attention. In the sound image control, a technique for interactively controlling the sound image localization has also attracted attention.
[0003]
In general, in a stereophonic system that has been conventionally used, a sound image position is controlled using a plurality of speakers to give a listener a sense of presence.
However, in a conventional speaker system in which speakers are installed on the left and right in front of a listener, a sound image can be formed only between two speakers.
Furthermore, when headphones are used, a specific phenomenon (intra-head localization) in which a sound image is formed in the head occurs, and the listener is often tired.
[0004]
On the other hand, with the development of hardware such as an audio signal processing technology and a DSP (Digital Signal Processor), even in a system in which left and right speakers are installed in front of a listener, sound image localization to an arbitrary position in front of the listener is performed. Became.
Also, in systems using headphones, attempts have been made to localize the sound image outside the head.
Conventionally, as an acoustic signal processing technique for localizing a sound image by two-channel speaker reproduction, a technique using convolution and crosstalk cancellation of a head acoustic transfer function in a time domain is known.
[0005]
In the acoustic signal processing technology for localizing a sound image, an operation related to sound image localization is performed using a head acoustic transfer function that means a transfer characteristic of a sound from a front sound source to a listener's ear.
FIG. 9 is a block diagram illustrating an example of a conventional sound reproducing device that controls sound image localization.
In FIG. 9, an audio signal of the L and R channels is input to an audio signal input unit 31. The input audio signals of the L and R channels are respectively converted into digital audio signals by the A / D converter 32 and input to the digital filter 33.
After being processed by the digital filter 33, the signal is converted into an analog sound signal by the D / A converter 34, and is supplied to the L and R channel speakers 36 via the amplifier 35. The speakers 36 are speakers installed on the left and right in front of the listener.
[0006]
In this sound reproducing device, a sound image localization controller 37 is connected to the digital filter 33 for controlling a sound image position.
When an arbitrary predetermined sound image position is input by the sound image localization controller 37, the digital filter 33 adds the two-channel signals output from the A / D converter 32 to the sound image position set by the sound image localization controller 37. Convolve the corresponding coefficients. Then, the convolved sound signal is amplified by the amplifier 35, and the sound signal is reproduced by the speakers 37 installed on the front left and right sides of the listener. By performing such an acoustic signal process, it is possible to control the sound image position at an arbitrary position on the left and right in front of the listener.
[0007]
Further, in the above-described crosstalk canceling technique, it is possible to localize a sound image outside a speaker by mixing sounds of opposite phases.
For example, in FIG. 10, it is assumed that a sound signal of sound pressure E is emitted from a sound image at a virtual position, and the sound signals give sound pressures of yL and yR to the left and right ears of the listener, respectively.
The head acoustic transfer function from the virtual sound image to the left ear is TL (Z), and the head acoustic transfer function from the virtual sound image to the right ear is TR (Z). When the relationship between the virtual sound image and yL, yR is represented using head-related acoustic transfer functions TL (Z), TR (Z),
yL = TL (Z) × E (Equation 1)
yR = TR (Z) × E (Equation 2)
It becomes.
[0008]
In order to localize the sound image at the virtual position in FIG. 10 (the position outside the actual left and right speakers), the sound pressure yR obtained by (Equation 1) and (Equation 2) in the digital filter 33 of FIG. , YL are corrected from the virtual sound image by a convolution operation. Further, the sound from the front left and right speakers actually installed for the listener is canceled.
That is, a virtual sound image as shown in FIG. 10 can be created by canceling the sound reaching the right ear from the left speaker and the sound reaching the left ear from the right speaker (these are referred to as “crosstalk sounds”).
Specifically, using the above TL (Z), TR (Z) and the head-related acoustic transfer function from the actual speaker to the ear, two directional localization filters having common inputs are used to determine the direction of the sound image. The sound image is localized at the virtual position in FIG. 10 through the acoustic signal by the localizer and a crosstalk canceller having a filter having a filter characteristic for removing the influence of the crosstalk sound.
[0009]
For example, the techniques related to the sound image localization described above are known, and the following literatures show the techniques related to the sound image localization.
[0010]
[Patent Document 1] JP-A-10-224900
[0011]
[Problems to be solved by the invention]
However, in the conventional sound reproducing device as described above, since the speaker is installed in the space in front of the listener, the speaker cannot be mounted at a position necessary to create an ideal sound field and sound image for the listener. There were many.
[0012]
For example, in the case of a listener sitting on a seat and watching a video on a display, it is necessary to mount a speaker in a place where viewing is not hindered, and the installation location of the speaker is restricted. Therefore, in some cases, the speaker must be installed in an undesired place in front of the listener, and the sound image can be freely moved from the position of the listener to an arbitrary space, for example, to the side or back of the listener, or in the vertical direction. I couldn't freely localize.
[0013]
Accordingly, the present applicant has previously provided, as a prior art, a sound reproducing apparatus capable of obtaining a natural sound image localization in an arbitrary space using a speaker installed in a space behind a listener in Japanese Patent Application No. 2001-246360. .
According to the present invention, a virtual sound source can be placed in a place where mounting is impossible, and it can be perceived that a reproduced sound is output from the virtual sound source. And can produce sound images.
However, when the listener's head moves during playback, it is difficult to obtain a sufficient sound image localization feeling because an optimal head-related acoustic transfer function cannot be applied.
[0014]
[Means for Solving the Problems]
The present invention has been made in view of such a problem, and realizes a sound reproducing apparatus that can obtain a natural sound image localization in an arbitrary space by using a speaker installed in a space behind a listener. It is an object of the present invention to be able to obtain a sense of sound image stability corresponding to the movement of the listener's head at that time.
[0015]
Therefore, the sound output device of the present invention is a sound output device that supplies a sound output signal of a plurality of channels to a speaker of a plurality of channels installed in a space behind a listening position of a listener. Correction filter means for performing a correction process based on a head acoustic transfer function so as to localize a sound image in a predetermined direction space with respect to the listening position with respect to the input signal, and a signal of a plurality of channels corrected by the correction filter means Signal output means for supplying a sound output signal to the speakers of the plurality of channels, a sensor means for detecting a positional relationship between a position of the speaker and a head position of the listener, and a detection result of the sensor means Control means for selecting a head-related sound transfer function according to the above, and controlling the correction processing of the correction filter means. That.
In this case, the space in the predetermined direction in which the sound image is localized is a space in front of the listening position.
Further, an operating means for variably instructing a predetermined direction space for sound image localization is further provided, and the control means selects a head acoustic transfer function according to a detection result of the sensor means and operation information of the operating means. Then, the correction processing of the correction filter means is controlled.
[0016]
The sound output system of the present invention includes the sound output device having the above-described configuration, and a plurality of channels of speakers installed in a space behind a listening position of a listener.
[0017]
According to the sound output method of the present invention, when a plurality of channels of sound output signals are supplied to a plurality of channels of speakers installed in a space behind the listener's listening position, the position of the speakers and the head of the listener Detecting the positional relationship with the position of the head, selecting a head-related acoustic transfer function according to the detection result, and for each of the digital audio input signals of a plurality of channels, to localize a sound image in a predetermined direction space with respect to the listening position. And performing a correction process based on the selected head-related acoustic transfer function, and supplying the corrected multi-channel signals to the multi-channel speakers as sound output signals.
[0018]
According to the present invention as described above, the sound signal output from the speaker installed in the space behind the listener is set as the sound signal subjected to the correction processing based on the head-related sound transfer function, so that the sound signal in front of the listening position is obtained. A sound image is localized at a position in a predetermined direction such as a space. Furthermore, a stable sound image localization state is obtained by selecting an appropriate head acoustic transfer function while detecting the position of the listener's head in real time.
The rear space of the listener includes the left and right ear openings of the listener, includes a plane parallel to the central axis of the listener's body, includes the plane, and is divided into two spaces by the plane. It means one space including the back of the listener.
The head acoustic transfer function is a function representing a transmission system from a sound source to one ear (eardrum) of a human, and reflects reflection, diffraction, resonance, and the like of the head, pinna, shoulder, and the like. The head acoustic transfer function can be obtained by measurement. Each head-related transfer function is represented by a complex number.
[0019]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, sound reproducing apparatuses according to first and second embodiments of the present invention will be described in detail with reference to the drawings.
[0020]
<First embodiment>
FIG. 1 shows a configuration of a sound reproducing apparatus according to a first embodiment.
This sound reproducing apparatus includes an audio digital signal input unit 1, a correction filter unit 2, an audio digital signal output unit 3, a D / A converter 4, an output amplifier 5, a left and right speaker 6 (left channel speaker 6L, right channel speaker 6R). , A CPU 11, a storage unit 13, and an infrared sensor 17.
[0021]
The audio digital signal input unit 1 receives, as a source of digital audio data, digital audio data of L and R channels output from a playback device for AV media such as CD / MD / DVD.
Alternatively, as a source of the analog audio signal, an analog audio signal output from, for example, an FM tuner is converted into digital audio data by an A / D converter (not shown), and the digital audio data is input to the audio digital signal input unit 1. May be supplied.
[0022]
The digital audio data of the L and R channels supplied to the audio digital signal input unit 1 is supplied to the correction filter unit 2 and a predetermined correction process is performed.
Although the details of the correction filter unit 2 will be described later, the correction filter unit 2 is located at a target position (the position of the virtual sound image 9 in FIG. 1) of the sound images reproduced by the left and right channel speakers 6L and 6R. The digital audio data is corrected as described above.
Further, the correction filter unit 2 is controlled based on real-time detection information of the position of the listener's head by the infrared sensor 17, and is configured to always execute an optimum correction process.
[0023]
The digital audio data corrected by the correction filter unit 2 is supplied to a D / A converter 4 via an audio digital signal output unit 3 and converted into analog audio signals of L and R channels.
The analog audio signal is supplied to the left and right channel speakers 6L and 6R through an attenuator circuit process and a power amplifier process for adjusting the volume in the output amplifier unit 5.
[0024]
The speakers 6L and 6R are arranged in a space behind the listener 8. The rear space includes the left and right ear openings of the listener 8, includes the plane 10 in a plane 10 parallel to the central axis of the body of the listener 8, and includes the plane 10 in two spaces divided by the plane 10. This means one space including the occiput of the listener 8.
[0025]
The speakers 6L and 6R arranged in the rear space are arranged at the position of the headrest of the seat 14 (or can be arranged at the position of the headrest) as shown in FIG. 2, for example.
That is, assuming that the listener sitting on the seat 14 is the target, the speakers 6L and 6R are arranged on the headrest or at the left and right positions in the headrest in this manner, and thus are arranged in the space behind the listener 8. It will be.
The speakers 6L and 6R output audio signals such as CDs, MDs, DVDs, and broadcasts. The sound image formed by the audio signals is obtained when the speakers 6L and 6R are at the positions of the headrests in FIG. Also, the correction processing of the correction filter unit 2 causes the virtual sound image 9 to be positioned, for example, in front of the listener 8. For example, as shown in FIG. 2, the display device 15 is disposed in front, and when the user sits down and watches an image, the position of the display device 15 can be set as the position of the virtual sound image 9.
[0026]
1 and 2 show speakers 12L and 12R as virtual sound sources. The virtual sound source referred to here is a speaker that realizes an ideal sound field or sound image, but is not actually installed in a space, that is, a speaker that does not actually exist in the speakers 12L and 12R.
On the other hand, if the actually installed speakers are called real sound sources, the speakers 6L and 6R are the real sound sources.
[0027]
FIGS. 3 and 4 show appropriate positions of the speakers 6L and 6R in the rear space.
FIG. 3 is a top view showing the horizontal direction of the rear space.
As shown in FIG. 3, the rear space includes the plane 10 and includes all the space behind the listener from the plane 10. Therefore, the speakers 6L and 6R are arranged at arbitrary positions in the rear space.
Also, as shown in FIG. 3, an orthogonal projection angle that forms a plane 10 with respect to a straight line connecting the positions of the speakers 6L and 6R to the left and right ear openings 7L and 7R of the listener 8 is defined as an angle θ1, and the speakers 6L and An orthographic angle formed with the central axis J of 6R is defined as an angle θ2. In this case, the angle θ1 and the angle θ2 are
0 ° ≦ θ1 ≦ 90 °
-30 ° ≤θ2≤ + 30 °
(However, counterclockwise is defined as +)
Is desirable.
By arranging the speakers 6L and 6R so that the angle θ1 and the angle θ2 satisfy the above conditions, the sound image formed by the acoustic signals from the speakers 6L and 6R can be more effectively processed by the correction processing of the correction filter unit 2. , For example, in front of the listener 8 as the virtual sound image 9.
[0028]
FIG. 4 is a side view showing the vertical direction of the rear space.
As shown in FIG. 4, the rear space includes the plane 10 and includes all spaces behind the listener from the plane 10. Therefore, the speakers 6L and 6R are arranged at arbitrary positions in the rear space.
As shown in FIG. 4, a straight line connecting the positions of the speakers 6L and 6R to the left and right ear openings 7L and 7R of the listener 8 includes the left and right ear openings of the listener and is perpendicular to the plane 10. Is an angle θ3, and the orthogonal angle formed between the central axes J of the speakers 6L and 6R is an angle θ4.
In this case, the angle θ3 and the angle θ4 are
-90 ° ≦ θ3 ≦ 90 °
-30 ° ≦ θ4 ≦ + 30 °
(However, counterclockwise is defined as +)
Is desirable.
By arranging the speakers 6L and 6R so that the angle θ3 and the angle θ4 satisfy the above-described conditions, the sound image formed by the acoustic signals from the speakers 6L and 6R can be more effectively processed by the correction processing of the correction filter unit 2. , For example, in front of the listener 8 as the virtual sound image 9.
[0029]
Further, in the sound reproducing device of the present example, an infrared sensor 17 is provided as shown in FIG. This infrared sensor is a sensor that detects the positional relationship between the speakers 6L and 6R and the head of the listener 8, specifically, the distance.
For example, as shown in FIG. 2, the infrared sensor is disposed at the center of the headrest of the seat 14 or the like, and detects the positional relationship between the positions of the speakers 6L and 6R and the head of the listener 8 sitting on the seat 14.
Since the infrared sensor 17 includes at least an infrared light emitting unit and an infrared light receiving unit, the distance between the speakers 6L and 6R and the head of the listener 8 can be detected as the positional relationship. The light emitting unit and the light receiving unit may be configured separately, or may be integrally configured such as a photocoupler.
[0030]
FIGS. 5A, 5B, and 5C show examples of the positional relationship detected by the infrared sensor 17, that is, the distance d.
As shown in the drawing, even when the listener 8 is sitting on the seat 14, the distance d between the head position and the positions of the speakers 6L and 6R varies depending on the posture of the listener 8.
The infrared sensor 17 detects this distance d and supplies this detection information to the CPU 11 shown in FIG.
[0031]
By arranging a plurality of infrared sensors 17 in the height direction, the height position of the head of the seated listener 8 can be detected together with the distance d (the distance in the horizontal direction). For example, as shown in FIGS. 5D and 5E, when the listener 8 is an adult and a child, the height position of the head greatly differs.
In this case, even if the horizontal distance d between the speakers 6L and 6R and the head is the same, if the height position of the head is different between the child and the adult, the actual distance from the speakers 6L and 6R to the head Will be different. In view of this, a plurality of infrared sensors 17 are arranged in the height direction so that the height of the head, that is, the distance h between the speakers 6L and 6R and the head in the vertical direction can be detected as shown in FIG. It is possible to more accurately detect the positional relationship between the speakers 6L and 6R and the head corresponding to the listeners 8 having various heights and sitting heights.
[0032]
Although not shown, since the head of the listener 8 may shift left and right on the seat 14, a plurality of infrared sensors 17 are arranged in the lateral direction of the backrest of the seat 14, and the shift in the left and right direction ( It is also preferable that the distance can be detected.
[0033]
In this example, the infrared sensor 17 is arranged at the center of the headrest of the seat 14, but the infrared sensor 17 may be arranged at any position. Further, the present invention is not limited to the infrared sensor.
That is, any configuration may be used as long as it can detect at least the positional relationship (at least the distance) between the speakers 6L and 6R arranged in the rear space and the head of the listener 8.
For example, in the case of a sound reproduction system in a car, the sound reproduction system may be arranged in a headrest in a driver's seat, or may be arranged in a ceiling or a side wall of the car.
[0034]
The CPU 11 of FIG. 1 constantly monitors the detection information from the infrared sensor 17, that is, the positional relationship between the speakers 6L and 6R and the head of the listener 8.
The storage unit 13 stores a head acoustic transfer function measured in advance in various positional relationships in a database.
The head acoustic transfer function is a function representing a transmission system from a sound source to one ear (eardrum) of a human, and reflects reflection, diffraction, resonance, and the like of the head, pinna, shoulder, and the like. The head acoustic transfer function can be obtained by measurement. Each head-related transfer function is represented by a complex number.
[0035]
When the CPU 11 determines the positional relationship (distance) based on the detection information from the infrared sensor 17, the CPU 11 searches the database in the storage unit 13 based on the positional relationship and obtains a head acoustic transfer function corresponding to the positional relationship. I do.
Then, the acquired head acoustic transfer function (or the correction parameter value calculated from the head acoustic transfer function) is supplied to the correction filter unit 2, and parameters for the correction process (transfer characteristics Hp (Z), Hm (described later) Z))) is set.
[0036]
That is, in the present example, the sound reproduced from the speakers 6L and 6R arranged in the rear space is localized in the virtual sound image 9 in the front space by the correction processing in the correction filter unit 2 as described above. Furthermore, by variably controlling the parameters of the correction processing in real time according to the positional relationship between the speakers 6L and 6R and the head of the listener 8, the localization state to the virtual sound image 9 regardless of the movement of the posture of the listener 8 and the like. Is to stabilize.
[0037]
Hereinafter, the correction processing using the head acoustic transfer function will be described.
First, a method of measuring the head acoustic transfer function will be described.
As described above, the correction filter unit 2 corrects the digital acoustic signal so that the sound image is located in an arbitrary space. The correction is performed by taking into account the auditory characteristics from the speaker to the eardrum of the listener. This is realized by using a partial acoustic transfer function.
[0038]
This head-related acoustic transfer function can be generally measured as follows. That is,
(A) A speaker and a dummy head in the shape of a human head are arranged.
(B) An impulse signal that becomes flat on the frequency axis when Fourier-transformed is input to the speaker as a test signal. The test signal may be a signal having the property of an impulse function such as a time stretched pulse signal.
(C) Measure the impulse response of the dummy head at the artificial ear. This is the head-related acoustic transfer function when the impulse response has the positional relationship (a).
[0039]
Therefore, for example, in the case of using the head acoustic transfer function in the sound reproducing device of the present example in which the speakers 6L and 6R are arranged on the sheet 14 as shown in FIG.
(A) A dummy head in the shape of a human head is placed on a standard sheet or a representative sheet.
(B) The speakers are arranged at actual speaker positions, for example, the positions of the speakers 6L and 6R, that is, the positions of the headrests of the seat 14, and the head acoustic transfer function when the speakers are located at the positions of the speakers 6L and 6R is obtained.
(C) A position where a virtual sound source is desired to be realized, for example, positions of virtual speakers 12L and 12R, that is, heads when speakers are arranged on the left and right of the display 15 in FIG. 2 and the speakers are located at the positions of the speakers 12L and 12R. Find the acoustic transfer function.
[0040]
In this way, the head-related transfer function for correcting the position of the sound image by the correction filter unit 2 can be measured.
That is, the correction filter unit 2 corrects the digital acoustic signal based on the head acoustic transfer functions (B) and (C), and by this data correction, the correction is performed on the headrest of the seat 14 as described above. The sound image by the attached speakers 6L and 6R is corrected to the position of the sound image by the speakers 12L and 12R at the position of the virtual sound source.
[0041]
The measurement of the head acoustic transfer function in the above (B), that is, the head acoustic transfer function of the speakers 6L and 6R and the dummy head, is based on the head position according to various postures and sitting heights of the listener on the seat. Perform measurements as appropriate. That is, the head acoustic transfer function corresponding to each position of the head is measured while changing the position of the dummy head, and these are converted into a database as a head acoustic transfer function corresponding to the head position.
[0042]
Further, for the measurement of the head acoustic transfer function of (C), that is, for the position where the virtual sound source is to be realized and the head acoustic transfer function of the dummy head, a point is set as the virtual sound source position and the speakers 12L and 12R are set. A head-related sound transfer function is measured by arranging a corresponding speaker. In this case, the measurement is performed according to the head position according to various postures, sitting heights, and the like of the listener on the seat. Then, these are made into a database as a head acoustic transfer function corresponding to the head position.
[0043]
That is, according to the head position of the listener 8 sitting on the seat 14, the head acoustic transfer function from the speakers 6L and 6R of the real sound source to the listener 8 and the head sound transfer function from the speakers 12L and 12R of the virtual sound source to the listener 8 A database is formed so that the head acoustic transfer function can be selected, and such a database is stored in the storage unit 13.
[0044]
In the example of the first embodiment shown in FIG. 1, the virtual sound image 9 is fixed to the front of the listener 8, but in the second embodiment described later, the virtual sound image 9 is fixed. This enables the listener 8 to arbitrarily set the position. In this case, for example, as the head acoustic transfer function from the speakers 12L and 12R of the virtual sound source to the listener 8, the measurement of the above (C) is performed while changing the positions of the speakers 12L and 12R of the virtual sound source to various various positions. , Need to be a database.
[0045]
The correction process using the head acoustic transfer function is performed as follows.
First, it is assumed that the head-related acoustic transfer function measured and analyzed according to the above (A) to (C) is as follows, as shown in FIG.
[0046]
FLL (Z): head acoustic transfer function from the left channel speaker 12L of the virtual sound source to the left ear
FLR (Z): head acoustic transfer function from the left channel speaker 12L of the virtual sound source to the right ear
FRL (Z): head acoustic transfer function from the right channel speaker 12R of the virtual sound source to the left ear
FRR (Z): head acoustic transfer function from the right channel speaker 12R of the virtual sound source to the right ear
GLL (Z): head acoustic transfer function from the left channel speaker 6L of the real sound source to the left ear
GLR (Z): head-related transfer function from the speaker 6L of the left channel of the real sound source to the right ear
GRL (Z): head acoustic transfer function from the speaker 6R of the right channel of the real sound source to the left ear
GRR (Z): head-related transfer function from the speaker 6R of the right channel of the real sound source to the right ear
[0047]
However, as described above, the positions of the virtual sound sources 12L and 12R are the positions of the speakers that realize the ideal sound field or sound image, and the positions of the real sound sources 6L and 6R are the speakers that are actually installed. Position. Each head-related transfer function is represented by a complex number.
[0048]
further,
XL (Z): left channel sound input signal (sound signal before correction)
XR (Z): right channel audio input signal (audio signal before correction)
YL (Z): left channel sound output signal (corrected sound signal)
YR (Z): right channel sound output signal (corrected sound signal)
And
[0049]
Here, YL (Z), GLL (Z), and YL (Z) are applied to the left ear 7L of the listener 8 by the sound pressure output from the actual speakers 6L and 6R disposed behind the listener 8. A GRL (Z) sound pressure is obtained.
In addition, the sound pressure of YL (Z) and GRR (Z) and the sound pressure of YL (Z) and GLR (Z) are applied to the right ear 7R of the listener 8 by the sound pressure output from the actual speakers 6L and 6R. Is obtained.
[0050]
On the other hand, if the speakers 12L and 12R of the virtual sound source are actually arranged, the left ear 7L of the listener 8 has XL (Z) · FLL (Z) and XR (Z) · FRL (Z). Is obtained.
Similarly, sound pressures of XR (Z) and FRR (Z) and XL (Z) and FLR (Z) are obtained in the right ear 7R of the listener 8 by sound pressures output from the speakers 12L and 12R. Can be
[0051]
The sound pressure obtained at the left ear 7L and the right ear 7R by the actual speakers 6L and 6R is the same as the sound pressure obtained at the left ear 7L and the right ear 7R when the virtual sound source speakers 12L and 12R are actually arranged. In the same manner, the same state as when the virtual speakers 12L and 12R are actually arranged can be formed by the speakers 6L and 6R of the real sound sources.
That is,
YL (Z) · GLL (Z) + YR (Z) · GRL (Z) = XL (Z) · FLL (Z) + XR (Z) · FRL (Z) (3)
YR (Z) · GRR (Z) + YL (Z) · GLR (Z) = XR (Z) · FRR (Z) + XL (Z) · FLR (Z) (Formula 4)
May be satisfied.
That is, the correction processing in the correction filter unit 2 basically includes, for the left and right channel audio input signals XL (Z) and XR (Z), the left and right channel audio signals that satisfy the above (Equation 3) and (Equation 4). If the output signals YL (Z) and YR (Z) are obtained, the virtual speakers 12L and 12R of the virtual sound source can be provided to the left ear 7L and the right ear 7R of the listener 8 using the actual speakers 6L and 6R. A sound image 9 can be formed.
[0052]
Here, in order to reduce the amount of data processing in the correction filter unit 2, the head acoustic transfer function is “symmetric”, that is,
FLL (Z) = FRR (Z) (Equation 5)
FLR (Z) = FRL (Z) (Equation 6)
GLL (Z) = GRR (Z) (Equation 7)
GLR (Z) = GRL (Z) (Equation 8)
Is satisfied, the correction filter unit 2 is configured.
In order to assume the symmetry in this manner, the location of the dummy head when measuring the head-related acoustic transfer function is located at the center of the seat 14, that is, the middle of the actually installed left and right speakers 6L and 6R. Is desirable. In addition, by doing so, the correction error due to the sheet is reduced, and the correction effect can be expected for any sheet.
[0053]
Then, under the assumptions of (Equation 5), (Equation 6), (Equation 7), and (Equation 8), to correct the sound from the virtual sound sources 12L and 12R, the above (Equation 3) (Equation 4) ) May be modified to satisfy the following (Equation 9) and (Equation 10). That is, FRR (Z) is replaced with FLL (Z), FRL (Z) is replaced with FLR (Z), GRR (Z) is replaced with GLL (Z), and GRL (Z) is replaced with GLR (Z).
YL (Z) · GLL (Z) + YR (Z) · GLR (Z) = XL (Z) · FLL (Z) + XR (Z) · FLR (Z) (Equation 9)
YR (Z) · GLL (Z) + YL (Z) · GLR (Z) = XR (Z) · FLL (Z) + XL (Z) · FLR (Z) (Equation 10)
And
[0054]
Here, Hp (Z) and Hm (Z) are
Hp (Z) = (FLL (Z) + FLR (Z)) / (GLL (Z) + GLR (Z)) (Equation 11)
Hm (Z) = (FLL (Z) -FLR (Z)) / (GLL (Z) -GLR (Z)) (Equation 12)
YL (Z) and YR (Z) in the above (Equation 9) and (Equation 10) are defined as
YL (Z) = Hp (Z) · (XL (Z) + XR (Z)) / 2 + Hm (Z) · (XL (Z) −XR (Z)) / 2 (Expression 13)
YR (Z) = Hp (Z) · (XL (Z) + XR (Z)) / 2−Hm (Z) · (XL (Z) −XR (Z)) / 2 (Equation 14)
It becomes.
Here, Hp (Z) and Hm (Z) represent transfer characteristics of the first digital filter and the second digital filter incorporated in the correction filter unit 2.
[0055]
It is known that the difference component of the two-channel stereo sound signal strongly affects the sense of stereo and the sense of spread. “Hm (Z) · (XL (Z) −XR (Z))” in (Equation 13) and (Equation 14) is a difference component of the stereo sound signal.
Therefore, by controlling the level of the “Hm (Z) · (XL (Z) −XR (Z))” term, it is possible to control the sense of spatial expansion.
[0056]
Then, multiplying the “Hm (Z) · (XL (Z) −XR (Z))” term of (Equation 13) and (Equation 14) by a coefficient k as a parameter for controlling the feeling of spread gives (Equation 13) (Equation 14) is as shown in the following (Equation 15) (Equation 16).
YL (Z) = Hp (Z) · (XL (Z) + XR (Z)) / 2 + k · Hm (Z) · (XL (Z) −XR (Z)) / 2 (Equation 15)
YR (Z) = Hp (Z) · (XL (Z) + XR (Z)) / 2−k · Hm (Z) · (XL (Z) −XR (Z)) / 2 (Formula 16)
[0057]
In Equations (15) and (16), when the coefficient k is increased, the difference component of the term “Hm (Z) · (XL (Z) −XR (Z))” is emphasized. Spaciousness is enhanced.
[0058]
Then, the correction filter unit 2 of the present example performs the correction processing shown in the above (Equation 15) and (Equation 16) on the input audio input signals XL (Z) and XR (Z) of the left and right channels to input the left and right channels. The sound output signals YL (Z) and YR (Z) may be obtained. Thereby, the above (Equation 9) and (Equation 10) are satisfied, and the virtual sound image 9 can be formed by the speakers 6L and 6R. Furthermore, the spread of the reproduction sound field can be controlled. Therefore, the correction filter unit 2 performs addition and subtraction in the above-described (Equation 15) and (Equation 16) with the filters having the characteristics Hp (Z) and Hm (Z) whose characteristics are expressed by (Equation 11) and (Equation 12). , And a level control circuit for multiplying by the coefficient k.
[0059]
The correction filter unit 2 is composed of, for example, a DSP, and performs correction processing by its arithmetic processing. FIG. 7 shows the processing in hardware. The sound input signals XL (Z) and XR (Z) are supplied to a subtraction circuit 22 and an addition circuit 21 to form a difference signal and a sum signal. Then, the difference signal output from the subtraction circuit 22 is supplied to the level control circuit 24, and the level control corresponding to the coefficient k in (Equation 15) and (Equation 16) is performed. Then, the difference signal is supplied to the FIR filter circuit 25 having the transfer characteristic Hm (Z) of (Equation 12). The sum signal output from the adding circuit 21 is supplied to the FIR filter circuit 23 having the transfer characteristic Hp (Z) of (Equation 11).
Then, the output signals of these two FIR filter circuits 23 and 25 are multiplied by a predetermined ratio (1/2 multiplication) by the level control circuits 26 and 27 and supplied to the addition circuit 28 and the subtraction circuit 29, and the sound output signal YL ( Z) and YR (Z) are formed.
The sound output signals YL (Z) and YR (Z) are supplied to the D / A converter 4 of the next stage through the sound digital signal output circuit 3 as shown in FIG.
[0060]
In particular, in the present embodiment, as described above, by detecting the position of the listener's head from the infrared sensor 17 in real time, the CPU 11 always selects the optimal head acoustic transfer function and sets it in the correction filter unit 2. Is done.
In other words, a certain set of head acoustic transfer functions FLL (Z), FRR (Z), FLR (Z), FRL (Z), GLL (Z) according to the detection result of the head position of the listener 8 , GRR (Z), GLR (Z), GRL (Z) are selected from the database of the storage unit 13.
However, FLL (Z), FLR (Z), GLL (Z), and GLR (Z) are actually selected as one set of head acoustic transfer functions because of the assumptions of the above (Equation 5) to (Equation 8). It should be done.
[0061]
As understood from this, adopting the assumptions of the above (Equation 5) to (Equation 8) not only simplifies the operation of the correction filter unit 2, but also reduces the database size of the storage unit 13. Also, there is an effect that the measurement of the head acoustic transfer function can be simplified.
[0062]
The head acoustic transfer functions FLL (Z), FLR (Z), GLL (Z), and GLR (Z) are, as shown in (Equation 11) and (Equation 12), the characteristics Hp (Z), Hm (Z) will be specified.
In other words, the CPU 11 selects each of the FIR filters (FIR filter circuits 23 and 25 in FIG. 7) of the characteristics Hp (Z) and Hm (Z) in the correction filter unit 2 according to the detection result of the infrared sensor 17. Control is performed so as to set characteristics Hp (Z) and Hm (Z) derived from the partial acoustic transfer functions FLL (Z), FLR (Z), GLL (Z), and GLR (Z).
[0063]
As described above, in the sound reproducing device of this example, even if the speakers 6L and 6R of the real sound source are mounted in the rear space (for example, the headrest of the seat 14), the positions of the speakers 12L and 12R as the virtual sound source (for example, the seats) 14 can be obtained in the same manner as when a speaker is arranged at the left and right positions of the display 15 installed in front of the display 14. Even when the listener's head fluctuates or moves, the virtual image is appropriately reproduced. Since the sound source can be arranged and perceived as if the reproduced sound is being output from this virtual sound source, an ideal sound field and sound image can be created without any discomfort for the listener.
Furthermore, since such a sound field can be realized by the speakers 6L and 6R in the rear space, the speaker arrangement in the front space is not required, and the ease and flexibility of speaker installation are improved. This is suitable for installation of the sound reproducing device in various spaces such as indoors and automobiles.
[0064]
Further, it is possible to prevent the sound image from being localized in the listener's head or in the rear space, and to localize the sound image in an arbitrary space.
In addition, it is possible to eliminate the inconvenience that occurs when a small satellite speaker for high-frequency reproduction is provided, that is, inconsistency in which sound is heard separately, so that sound is output from one speaker. Can be perceived.
[0065]
Further, as described above, the difference component between the left and right channels of the audio signal strongly affects the stereo feeling and the spaciousness of the reproduced sound, but the level control circuit 24 is provided in the correction filter unit 2 as shown in FIG. The multiplication of the level control circuit 24 by the coefficient k can emphasize the spatial spread of the reproduced sound.
Of course, for example, by variably setting the multiplication coefficient k of the level control circuit 24 by the CPU 11, the level of the difference component supplied from the subtraction circuit 23 to the subsequent FIR filter circuit 25 can be variably controlled, and the spatial It is also possible to adjust and correct the feeling of spaciousness, and to perform optimum correction according to the volume level.
[0066]
In the above description, the head-related transfer function is stored in a database in the storage unit 13. However, the measured head-related transfer function itself does not always have to be stored. That is, the characteristics Hp (Z) and Hm (Z) according to various head position states of the listener 8 are calculated from the head acoustic transfer function measured in advance, and the transfer characteristic control is performed by the FIR filter circuits 23 and 25. The values may be stored in a database.
[0067]
Also, by taking the average of a plurality of head-related transfer functions, an effective correction filter unit 2 can be created even when three or more speaker systems are used, and therefore, a correction filter unit that does not limit the number of speakers is used. 2 (sound reproduction device).
[0068]
<Second embodiment>
FIG. 8 shows a sound reproducing apparatus according to the second embodiment. The configuration of FIG. 8 is obtained by adding an operation unit 18 to the configuration of FIG. Other configurations are basically the same as those in FIG.
[0069]
In FIG. 8, the operation unit 18 can be arbitrarily operated by the listener 8 and the like, and is for arbitrarily changing the position of the virtual sound image 9.
For example, the listener 8 can change the position of the virtual sound image 9 in the front space by operating the operation unit 18 as, for example, virtual sound images 9a and 9b.
For this purpose, the head sound transfer function from the speakers 12L and 12R of the virtual sound source that realizes the position of the virtual sound image 9 specified by the operation unit 18 may be variably selected.
[0070]
In this case, when the head acoustic transfer function is measured in advance, the positions of the speakers 12L and 12R of the virtual sound source are changed to various positions as the head acoustic transfer function from the speakers 12L and 12R of the virtual sound source to the listener 8. While the above-mentioned measurement (C) is performed, a database is stored in the storage unit 13.
[0071]
Then, the CPU 11 responds to the operation of the operation section 18 by measuring the head-related acoustic transfer functions (FLL (Z), FLR (Z) measured in the arrangement state of the speakers 12L and 12R for positioning the virtual sound image 9 instructed. ).
Note that there are a plurality of sets of head acoustic transfer functions (FLL (Z), FLR (Z)) selected according to the operation. This is because a plurality of head-related acoustic transfer functions are measured while changing the head position of the listener 8 in a certain speaker 12L, 12R arrangement state.
For example, when the position of the virtual sound image 9a is specified, the head acoustics measured a plurality of times by changing the dummy head position in the arrangement state of the speakers 12L and 12R (the positions of 12La and 12Ra) for realizing the virtual sound image 9a. The transfer function is chosen.
[0072]
In this case, the CPU 11 further selects the head acoustic transfer functions FLL (Z), FLR (Z), GLL (Z), and GLR (Z) according to the detection information from the infrared sensor 17.
That is, the current head position of the listener 8 is obtained from a plurality of FLL (Z) and FLR (Z) as head acoustic transfer functions corresponding to the position of the virtual sound image 9 selected in accordance with the operation of the operation unit 18. , And one FLL (Z) and GLR (Z) are selected in the same manner as in the first embodiment.
When the head-related acoustic transfer functions FLL (Z), FLR (Z), GLL (Z), and GLR (Z) are selected in this manner, the selected head-related acoustic transfer functions FLL (Z), FLR (Z) , GLL (Z) and GLR (Z) controlling the transfer characteristics Hp (Z) and Hm (Z) of the FIR filter circuits 23 and 25 in the correction filter unit 2 are the same as in the first embodiment. .
[0073]
That is, in the case of the second embodiment, the position of the virtual sound image 9 can be arbitrarily changed, and the position of the virtual sound image 9 can be moved according to the taste of the listener 8, or the sound reproduction device such as a room or an automobile can be used. A suitable position of the virtual sound image 9 can be adjusted according to the listening environment including the installation space, the arrangement position of the display device 15, and the like.
[0074]
While the embodiments of the present invention have been described above, various other modifications are possible as the present invention.
In the above example, the head position of the listener 8 corresponds to the movement in the front-back direction and the difference in the height direction (seat height difference) in real time. It is also conceivable to enable the detection and to select the head-related transfer function corresponding to the detection.
[0075]
The arrangement of the speakers 6L and 6R is, of course, not limited to the seat 14, and may be in any form as long as it is arranged in the space behind the listener 8.
In addition, in the case of an example in which the seat is arranged on the seat 14, not only the seat for the room, the seat in the car, but also the seat of a train, a ship, an airplane, a theater, and the like, a realistic sound experience in various environments can be provided. Can be realized.
[0076]
【The invention's effect】
As understood from the above description, according to the present invention, an acoustic signal output from a speaker installed in a space behind a listener is an acoustic signal subjected to a correction process based on a head acoustic transfer function. Then, the sound image is localized in a predetermined direction position such as a space in front of the listening position, and further, for the correction processing, an appropriate head acoustic transfer function is selected while detecting the position of the listener's head. I have to.
For this reason, even when the listener's head fluctuates or moves, the virtual sound source can be appropriately arranged, and the virtual sound source can be perceived as outputting a reproduced sound. It is possible to create an ideal and stable sound field and sound image without a sense of incongruity.
[0077]
In addition, since the speaker is installed in the space behind the listener, the sound image can be freely located in any space from the listener's position without installing the speaker in an unintended location in front of the listener. Can be done.
This eliminates the difficulty of arranging the speakers indoors. In addition, in the case of displaying an image, there is an advantage that there is no problem in the positional relationship of the speaker with the display device or the screen.
Also, by reproducing a virtual sound image in an arbitrary space by using a rear speaker, for example, a speaker disposed on a seat, an effective acoustic experience can be achieved, so that it can be used for automobiles, trains, ships, airplanes, and indoor viewing applications. The effect is that it is possible to realize an effective music appreciation as if listening to a live performance, and a movie appreciation as if it were in a movie theater.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of a sound reproducing device according to a first embodiment of the present invention.
FIG. 2 is an explanatory diagram of a speaker and a sensor provided on a seat in the embodiment.
FIG. 3 is a top view illustrating a horizontal direction of a rear space in the sound reproducing device according to the embodiment;
FIG. 4 is a side view illustrating a vertical direction of a rear space in the sound reproducing device according to the embodiment;
FIG. 5 is an explanatory diagram of a listener, a real sound source, a virtual sound source, and a head acoustic transfer function in the sound reproduction device according to the embodiment;
FIG. 6 is an explanatory diagram of a positional relationship between a speaker and a listener's head in the sound reproducing device according to the embodiment;
FIG. 7 is an explanatory diagram of a correction process of the sound reproducing device according to the embodiment;
FIG. 8 is a block diagram illustrating a configuration example of a sound reproducing device according to a second embodiment of the present invention.
FIG. 9 is a block diagram illustrating a configuration example of a conventional sound reproducing device.
FIG. 10 is an explanatory diagram of radiation of an acoustic signal E from a virtual sound image.
[Explanation of symbols]
Reference Signs List 1 audio digital signal input unit, 2 correction filter unit, 3 audio digital signal output unit, 4 D / A converter, 5 output amplifier, 6L left channel speaker, 6R
Right channel speaker, 17 infrared sensor, 18 operation unit

Claims (5)

As a sound output device for supplying a sound output signal of a plurality of channels to a speaker of a plurality of channels installed in a space behind a listening position of a listener,
For each acoustic digital input signal of a plurality of channels, so as to localize the sound image in a predetermined direction space with respect to the listening position, a correction filter means for performing a correction process based on the head acoustic transfer function,
Sound signal output means for supplying signals of a plurality of channels corrected by the correction filter means as sound output signals to the speakers of the plurality of channels;
Sensor means for detecting the positional relationship between the position of the speaker and the head position of the listener;
Control means for selecting a head acoustic transfer function according to the detection result of the sensor means, and controlling a correction process of the correction filter means;
A sound output device comprising: The sound output device according to claim 1, wherein the space in the predetermined direction in which the sound image is localized is a space in front of the listening position. Operating means for variably instructing a predetermined direction space for sound image localization is further provided,
2. The apparatus according to claim 1, wherein the control unit selects a head acoustic transfer function according to a detection result of the sensor unit and operation information of the operation unit, and controls a correction process of the correction filter unit. 3. The sound output device according to claim 1. A multi-channel speaker installed in the space behind the listening position of the listener,
For each acoustic digital input signal of a plurality of channels, so as to localize the sound image in a predetermined direction space with respect to the listening position, a correction filter means for performing a correction process based on the head acoustic transfer function,
Sound signal output means for supplying the signals of the plurality of channels corrected by the correction filter means as sound output signals to the speakers of the plurality of channels;
Sensor means for detecting the positional relationship between the position of the speaker and the head position of the listener;
Control means for selecting a head acoustic transfer function according to the detection result of the sensor means, and controlling a correction process of the correction filter means;
A sound output system comprising: When supplying a sound output signal of a plurality of channels to a speaker of a plurality of channels installed in a space behind a listening position of a listener,
Detecting the positional relationship between the position of the speaker and the position of the listener's head,
A head-related sound transfer function is selected according to the detection result,
For each audio digital input signal of a plurality of channels, so as to localize a sound image in a predetermined direction space with respect to the listening position, perform a correction process based on the selected head acoustic transfer function,
A sound output method, wherein the corrected signals of a plurality of channels are supplied as sound output signals to the speakers of the plurality of channels.

Images