JPH09327099A - Acoustic reproduction device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the cost and to decrease the circuit scale. SOLUTION: A couple of FIR filters (1R(1), 1L(1)) and (1R(2), 1L(2)) are provided to two acoustic signals for different virtual sound images S1 , S2 . Each of the FIR filters has a filter characteristic where a transfer function to cancel a crosstalk component reaching a right ear or a left ear of a listener is convoluted on a transfer function of a virtual sound image corresponding to an input signal to be reached to the right ear or the left ear of the listener. Then a 3-dimensional sound field space where two virtual sound images are localized at a prescribed position is realized by adding outputs of the FIR filters 1R(1) and 1R(2) and supplying the sum to a right speaker SP(R) and adding outputs of the FIR filters 1L(1) and 1L(2) and supplying the sum to a left speaker SP(L).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound reproducing device, and is suitable for application to a sound reproducing device for reproducing a three-dimensional sound space, for example.

[0002]

2. Description of the Related Art Conventionally, as a sound reproduction system, a system in which a stereo sound is output by a pair of left and right speakers for listening or a headphone or the like is used is widely known. By the way, when a headphone is used as a sound reproducing system for reproducing the normal stereo sound as described above, the reproduced sound image is heard as if it is localized in the listener's head. It is called inward localization. That is, it is not possible to obtain a three-dimensional stereophonic image that can be heard in reality. Also,
When stereo sound is reproduced by a pair of left and right speakers, a spatial expansion of the sound field can be obtained as compared with the case of reproduction using the above headphones. However, the reproduced sound image is basically a two-dimensional sound image localized only in the space between the left and right speakers, and a three-dimensional sound image localization cannot be obtained.

Therefore, the binaural system is known as one of the techniques capable of reproducing a three-dimensional acoustic space. For example, in this binaural system,
For example, a sound source collects sound by a pair of left and right microphones provided on both ears of the dummy head. Then, in the case of reproducing using a headphone, if the acoustic signal of the collected sound source is supplied to the headphone, it is possible to reproduce the same three-dimensional sound localization as at the time of collecting the sound source. .

Alternatively, instead of collecting the sound using the dummy head as described above, the characteristics of the reproduced sound field set so that the desired sound source is localized three-dimensionally with respect to a general stereo sound signal. By convolving with, a sound signal equivalent to that picked up by the dummy head as described above is obtained by the binaural method. By reproducing this acoustic signal with headphones, a three-dimensional virtual sound image localization can be obtained, and the same audibility as that of a reproduced sound source by the binaural method can be obtained. The signal processing circuit for obtaining such a virtual sound image can be configured by, for example, a digital filter.

On the other hand, in order to properly reproduce the acoustic signal based on the binaural system by the pair of left and right speakers, crosstalk until the sound of the reproduced sound source reaches the ears of the listener from the left and right speakers. Ingredients must be considered. That is, unlike when listening to the playback sound through the headphones, when listening to the playback sound through the speaker, the component of the playback sound that reaches the listener's left ear from the right channel speaker and the right channel of the listener from the left channel speaker. There is always a component of the reproduced sound that reaches to the extent that it cannot be ignored. Therefore, in order to properly obtain a three-dimensional sound field by reproduction using a speaker, for example, the crosstalk component is removed from the acoustic signals of the left and right channels, and as a result, as if to the listener. It is necessary to hear the sound without crosstalk from each of the left and right speakers. In recent years, a so-called crosstalk cancellation filter configured to perform processing on an acoustic signal so that the sound reaching the left and right ears of the listener has the same characteristics as the original input signal to the speaker. Is also proposed. Such a crosstalk cancel filter can also be configured by a digital filter.

[0006]

By the way, a digital filter for convolving a three-dimensional sound field characteristic into an acoustic signal in order to obtain a virtual sound image having the above-mentioned three-dimensional localization,
The digital filter for forming the crosstalk cancel filter has a large circuit scale and is expensive as hardware. Therefore, it is preferable to reduce the circuit scale and cost of the sound reproducing device by reducing the number of these digital filters. In addition, when considering a sound reproduction system that realizes three-dimensional sound field reproduction by using a digital filter for convolving a three-dimensional sound field characteristic in an acoustic signal, conventionally, a headphone and a speaker are used. The case differs in the configuration of the digital filter for forming the sound reproducing device. For this reason, it is difficult to support both the system using headphones and the system using speakers by common hardware.

[0007]

Therefore, in order to solve the above-mentioned problems, the present invention relates to the transfer function of the path which is supposed to reach the right ear of the listener from the localization position of one virtual sound image. First digital filter means having a transfer function obtained as a filter characteristic by convolving the transfer function obtained so as to cancel the crosstalk component of the virtual sound image that is supposed to reach the right ear of the With respect to the transfer function of the path that is supposed to reach the left ear of the listener from the localization position of the one virtual sound image, the crosstalk component of the virtual sound image that is reached by crosstalking to the left ear of the listener is canceled. Second digital filter means having a transfer function obtained by convolving the transfer function obtained to perform It was possible to configure the sound reproducing apparatus includes a. Then, the output signals of the first digital filter means and the second digital filter means are respectively configured to be output as audio by a speaker, or the transfer function obtained to cancel the crosstalk component is invalidated. So that the filter coefficients of the first digital filter and the second digital filter means are set, and the output signals of the first digital filter means and the second digital filter means are output as audio by headphones. I decided to configure it.

A set of the first digital filter means and the second digital filter means is provided for each of a plurality of virtual sound images, and a first set is provided for each of the plurality of virtual sound images. Second adding means for adding and outputting the outputs of the digital filter means and the output of the second digital filter means provided corresponding to each of the plurality of virtual sound images. And a sound reproducing device. And
The output signals of the first adding means and the second adding means are
Each of them is configured to output sound by a speaker,
Alternatively, the filter coefficients of the first digital filter and the second digital filter means are set so as to invalidate the transfer function obtained so as to cancel the crosstalk component, and the first addition means and The output signal of the second adding means is configured to be output as sound by headphones.

According to the above configuration, the sound reproducing device is 1
The number of digital filters required to localize one virtual sound image is two. Also, as the characteristics of the digital filter, for example, by changing the setting of the coefficient,
By setting the characteristics that make the transfer function required to cancel the crosstalk component invalid, the system with headphones that does not need to take the crosstalk component into consideration will be equipped with a common digital filter. This can be dealt with by the configuration of the playback device.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of a sound reproducing device of the present invention will be described below with reference to FIGS. The following description will be given in the following order. 1. FIR filter configuration 2. Background to the present invention 3. Sound reproduction device of the present embodiment

1. FIR Filter Configuration In order to configure the sound reproducing apparatus of the present embodiment, for example, an FIR (Finite Impulse Responce) filter whose impulse response is represented by a finite time length is used as a digital filter. The configuration of will be described.

FIG. 14 shows an FIR filter (direct type configuration).
FIG. 3 is a block diagram showing a configuration of the F-th order in this case.
A general configuration as an IR filter is shown. As shown in this figure, the FIR filter has n delay elements D (1) to D (n) connected in series having a delay time T.
And the input signal x (nT) and each delay element D (1) to D (D).
By the n + 1 multipliers M (0) to M (n) that perform multiplication on the delay output of (n) and the connection configuration shown in the figure,
N adders AD (1) for adding and outputting the multiplication result for the current input signal and the multiplication result for the past input signal
~ AD (n). In this case, the adder AD
The output of (n) becomes the output signal y (nT) of the FIR filter. The delay time T of the delay elements D (1) to D (n) corresponds to the sampling period, and the multipliers M (0) to M (M).
The filter coefficients of (n) are h (0) to h (N-
1). The relationship between the inputs and outputs of such a FIR filter is

[Equation 1] It is represented by In the present embodiment, as FIR filters forming an audio reproducing device as described later, filter coefficients h (0) to h (N-1) of multipliers are provided so that required frequency characteristics (filter characteristics) can be obtained. ), A sound reproducing device capable of reproducing a three-dimensional sound space is configured.

2. Background to the Present Invention Next, the background to the present invention will be described with reference to FIGS. 8 to 13 as a premise for describing the present embodiment.

A structure shown in FIG. 8 is used as a crosstalk canceling filter for processing the acoustic signal so that the sound reaching the left and right ears of the listener has the same characteristics as the original input signal to the speaker. Are known. Actually, in such a system, it is necessary to perform filtering for flattening the transfer characteristic peculiar to the left and right channel speakers SP (L) (R), but in this case, for convenience, Speaker SP
It is assumed that the transfer characteristics of (L) and (R) are 1, respectively.

Here, the transfer characteristic (transfer function) of the audio path reaching the right ear of the listener LS from the right speaker SP (R) that outputs the audio signal as sound is H _RR, and the right speaker is the crosstalk component. The transfer characteristic (transfer function) of the voice path that reaches the left ear of the listener LS from SP (R) is H _RL . Also, from the left speaker SP (L), the listener L
The transfer characteristic of the voice path reaching the left ear of S is H _LL, and the transfer characteristic of the voice path reaching the right ear of the listener LS from the left speaker SP (L) is H _LR as a crosstalk component.

In the configuration of the crosstalk cancel filter shown in this figure, the acoustic signal S _R of the right channel is input to the adder AD _R and the FIR filter 2 _R. FI
The characteristic of the R filter 2 _R is represented by C _R ,

[Equation 2] Required by. The output of the FIR filter 2 _R is input to the adder AD _L. Also, the acoustic signal S of the left channel
_L is input to the adder AD _L and the FIR filter 2 _L. The characteristic of the FIR filter 2 _L is represented by C _L ,

(Equation 3) Required by. The output of the FIR filter 2 _R is input to the adder AD _R.

In the adder AD _R , the input right channel acoustic signal S _R and the left channel acoustic signal S _L via the FIR filter 2 _L (characteristic C _L ) are added and output,
Output to the FIR filter 3 _R. FIR filter 3 _R
Is

(Equation 4) The output of the right speaker SP (R)
Is supplied to. On the other hand, in the adder AD _L , the left channel acoustic signal S _L and the right channel acoustic signal S _R via the FIR filter 2 _R (characteristic C _R ) are added and output, and FI
Output to R filter 3 _L. The FIR filter 3 _L is

(Equation 5) And the output is the left speaker S.
Is supplied to P (L). Note that the left speaker SP is actually
(L) and the right speaker (R), the amplifier for outputting the acoustic signal as a voice includes FIR filters 3 _L and FI.
Although it is provided in the latter stage of the R filter 3 _R , its illustration is omitted here.

The acoustic signals S _R and S _L are input to the crosstalk canceling filter having the above-mentioned structure and the speaker SP is supplied.
By outputting from (R) and (L), the crosstalk component reaching the ears of the listener LS is canceled, and the left and right ears of the listener LS can hear sounds equivalent to the acoustic signals S _L and S _R , respectively. It will be. And the acoustic signal S _R ,
If S _L is for reproducing a three-dimensional sound image localization, a signal similar to the sound reproduced by the headphones based on the binaural system is obtained, and the listener L
S can also listen to the sound reproduced by the three-dimensional sound image localization through the two speakers SP (R) (L).

Here, if it is assumed that the left and right channel systems of the crosstalk cancel filter shown in FIG. 8 are symmetrical, the crosstalk cancel filter of FIG. 8 has the crosstalk shown in the block diagram of FIG. It can be regarded as equivalent to the configuration of the cancel filter. That is, as described below, the characteristics of the FIR filters 2 _R and 2 _L in FIG. 8 are replaced with the characteristics of the FIR filters FIR filters 4 _R and 4 _L in FIG. 9, and the characteristics of the FIR filters 3 _R and 3 _L in FIG. The FIR filter of FIG. 9 can be replaced with the characteristics of FIR filters 5 _R and 5 _L. In addition,
Since the connection form of each circuit block is the same in FIG. 8 and FIG. 9, description thereof will be omitted.

In FIG. 9, the transfer characteristic of the voice path reaching from the speaker SP (R) to the right ear of the listener LS and the transfer characteristic of the voice path reaching from the speaker SP (L) to the left ear of the listener LS. On the other hand, as S, on the other hand, as a crosstalk component, a transfer characteristic of a voice path reaching from the speaker SP (R) to the left ear of the listener LS and a voice reaching from the speaker SP (L) to the right ear of the listener LS. The transfer characteristics of the paths are both A. That is, for the transfer characteristics H _RR and H _LL shown in FIG. 8, H _RR = H _LL = S, and for the transfer characteristics H _RL and H _LR of the crosstalk component, H _RL = H _LR.
= A. According to this, the characteristics of the FIR filters 2 _R and 2 _L shown in FIG.

(Equation 6) This transfer characteristic C becomes the filter characteristic of the FIR filters 4 _R and 4 _L shown in FIG. Further, in the crosstalk cancel filter of FIG. 9, if the left and right channel systems of the crosstalk cancel filter are symmetrical, the FIR filter 3 shown in FIG.
_Since the terms C _L and C _R of the mathematical expression representing the characteristics of _R 3 _L are expressed as C _L = C _R = C, the FIR filters 5 _R 5 _L shown in FIG. The filter characteristics are

(Equation 7) Will be represented by

For the sake of convenience, the following sound reproducing device will be described as being constructed based on the system shown in FIG.

When an attempt is made to reproduce a three-dimensional sound field with two speakers using the crosstalk cancel filter having the structure shown in FIG. 9, a FIR filter as shown in FIG. 10 is provided. The sound can be reproduced by constructing the sound reproducing device. This figure shows a configuration in which acoustic signals S ₁ , S _{2 to} Sn of _n virtual sound images are input and sound image localization is set three-dimensionally with respect to these input signals. In this case, as shown in FIG. 11, as the sound signal S _1, S ₂ ~S _n to the virtual sound image corresponding respectively VS (1), VS (2 ) ~VS (n), the virtual sound VS (1), From the positions where VS (2) to VS (n) are localized, the transfer characteristics (transfer functions) of the paths of the voice components reaching the right ear of the listener LS are set to H _1R , H _2R , and H _nR , respectively. The transfer characteristic (transfer function) of the path of the voice component reaching the left ear of the person LS is H _1L ,
H _2L and H _nL .

In the sound reproducing device shown in FIG. 10, sound signals S ₁ and S _{2 to} S _n respectively corresponding to virtual sound images VS (1) and VS (2) to VS (n) are shown as input signals. There is. The acoustic signal S ₁ is the FIR filter 6 _R (1), 6 _L
The acoustic signal S ₂ input to (1) is input to the FIR filter 6 _R.
(2), 6 _L (2) is input and the acoustic signal S _n is FIR
It is input to the filters 6 _R (n) and 6 _L (n). These FIR filters 6 _R (1 to n), 6 _L (1 to n)
Is assumed to have a transfer function (see FIG. 11) that matches the localization of the virtual sound image corresponding to each input acoustic signal as a filter characteristic. That is, the acoustic signal S is generated by the FIR filters 6 _R (1 to n) and 6 _L (1 to n).
_1, the virtual sound VS (1) with respect to S ₂ to S _n, VS
(2) to VS (n) will be given such a characteristic as to be heard with localization (under the condition that there is no crosstalk). The outputs of the FIR filters 6 _R (1 to n) corresponding to the right channel system are input to the adder AD _1R and added, and the outputs of the FIR filters 6 _L (1 to n) corresponding to the left channel system. Is the adder AD
Input to _1L and added. Adder AD _1R and adder A
The output of D _1L is input to the crosstalk cancel crosstalk cancel filter unit 11.

The crosstalk canceling filter section 11 shown in this figure is constructed by rearranging and modifying the system of the crosstalk canceling filter shown in FIG. 9, and as a result, the following filter characteristics and The configuration of the connection form is adopted. The crosstalk cancellation filter unit 11 includes four FIR filters 7 _R , 7 _L ,
The FIR filter 7 includes 8 _R and 8 _L.
R and 8 _L

The FIR filters 7 _L and 8 _R have the characteristics shown by

(Equation 8) It has the characteristics shown by.

FIR filter 7 _R and FIR filter 7 _L
The output of the adder AD _1R is branched and input to the FIR
The output of the adder AD _1L is branched and input to the filter 8 _R and the FIR filter 8 _L. The outputs of the FIR filter 7 _R and the FIR filter 8 _R are input to the adder AD _2R of the right channel system, and the outputs of the FIR filter 7 _L and the FIR filter 8 _L are input to the adder AD _2L of the left channel system. To be done. The output of the adder AD _2R is supplied to the right speaker SP (R), and the output of the adder AD _2L is supplied to the left speaker SP (L) and is output as the sound of the left channel and the right channel, respectively.

According to this structure, the virtual sound image VS is generated by the FIR filters 6 _R (1 to n) and 6 _L (1 to n).
(1) to VS (n), the acoustic signals S _{1 to} S _n given the transfer characteristics of the sound image localization are canceled by the crosstalk cancellation filter unit 11 to eliminate the crosstalk voice components reaching the left and right ears. With the above characteristics, the sound is output from the right speaker SP (R) and the left speaker SP (L). Thus, the listener LS, acoustic signals S ₁ to S having a characteristic based on the transfer function shown in FIG. 11
_n (sound equivalent to the signal at the previous stage input to the crosstalk cancellation filter unit 11) is heard,
It is possible to realize the reproduction of a three-dimensional sound field.

In the above description, the case where the virtual sound image is reproduced by the two speakers to reproduce the three-dimensional sound field has been described. However, when the three-dimensional reproduced sound field is obtained using the headphones, for example, This can be realized by the configuration of the sound reproducing device shown in FIG. In this case,
As shown in FIG. 13, the paths of the audio components of the virtual sound images VS (L) and VS (R) of the sound signals S _L and S _R corresponding to the left and right channels are assumed to reach the right ear of the listener LS. The transfer characteristics are H _RR and H _LR , respectively, and the transfer characteristics of the path of the voice component reaching the left ear of the listener LS are H _RL and H _RL , respectively.
H _LL .

In this case, the headphones HP (R),
Since the sound is output in a situation where there is no indirect sound that crosstalks to both ears of the listener LS due to (L), it is not necessary to consider the crosstalk component. Therefore, for example, FIR filters 6 _R (1 to n) and FIR filters obtained by omitting the configuration of the crosstalk cancellation filter unit 11 from the sound reproducing apparatus shown in FIG.
A three-dimensional sound field can be reproduced only by applying the system composed of the filter 6 _L (1 to n) side.

That is, as shown in FIG. 12, the FIR filters 9 _R and FIR for inputting the sound signal S _R of the right channel
The filter 9 _L provided, FIR filter for inputting an acoustic signal S _L of the left channel 10 _R, provided FIR filter 10 _L. These FIR filters 9 _R , 9 _L , 10 _R , 1
0 _L is the transfer function H _RR , H _RL , H described in FIG.
Coefficients are set so that the characteristics of _LR and _HLL are obtained. Then, the outputs of the FIR filters 9 _R , 10 _R are added by the adder AD _R , and this output is supplied to the headphone HP (R) of the right channel, and the FIR filters 9 _L , 1
By adding the output of 0 _L by the adder AD _L and supplying the output to the headphone HP (L) of the left channel, the listener LS receives the three-dimensional sound image based on the transfer characteristic shown in FIG. You are hearing the sound.

Considering the configuration of the sound reproducing apparatus for reproducing the three-dimensional sound field described above, for example, in the sound reproducing apparatus having the structure shown in FIG. 10, the three-dimensional sound field is reproduced by the speaker. If you try, the number of virtual sound images that should be localized three-dimensionally (the number of input acoustic signals)
Where n is 2n FIR filters (6 _R (1 to
n) and 6 _L (1 to n)) and four FIR filters (7 _R , 7 _L , 8 _R , 8 _L ) forming the crosstalk canceling filter unit 11 are required. That is, a total of 2n + 4 FIR filters are required. For example, when such an FIR filter is configured by hardware such as an IC for digital signal processing, the number of FIR filters to be provided in the sound reproducing device is reduced as much as possible because it is relatively expensive and the circuit scale becomes large. It is preferable from the viewpoints of cost, reduction in size and weight of the device, and power saving.

Further, according to the sound reproducing device described so far, as can be seen by comparing FIG. 9 and FIG.
The configuration of the filter is different in terms of hardware.
For this reason, it is difficult to use a common configuration for speaker compatibility and headphone compatibility, and it is necessary to provide sound reproduction systems having different configurations, which is also disadvantageous in terms of cost, for example.

3. Sound reproduction device of the present embodiment With the background as described above, the present embodiment reduces the number of FIR filters constituting the sound reproduction device for reproducing a three-dimensional sound field as compared with the conventional one. It is configured such that a common hardware configuration can be used for speaker compatibility and headphone compatibility. Hereinafter, a sound reproducing device as an embodiment of the present invention will be described with reference to FIGS.

FIG. 1 shows the configuration of the sound reproducing apparatus of this embodiment in the case of localizing one virtual sound image. That is, it is the basic configuration of the sound reproducing device according to the present embodiment. In this case, as shown in FIG. 2, as the sound of the virtual sound image VS (1), the transfer characteristic of the path reaching the right ear of the listener LS is H _1R, and the transfer characteristic of the path reaching the left ear is H. _{It is 1L} .

Now, let us consider a case in which only the virtual sound image VS (1) shown in FIG. 2 is localized by the sound reproducing device having the configuration described with reference to FIG. Let S ₁ be the acoustic signal that is 1), and let L and R be the outputs of the speakers SP (R) and (L), respectively.

[Equation 9] Can be represented by And these expressions are

(Equation 10) And can be transformed.

Based on this, the sound reproducing device of the present embodiment can be configured as shown in FIG. That is, in this case, a pair of FIR filters 1 _R and 1 _L is provided in the left and right systems, and the FIR filter 1 _{R is}
Based on the above (Equation 10),

[Equation 11] The FIR filter 1 _L is similarly configured based on the above (Equation 10) so that the characteristics shown by

(Equation 12) It is configured so that the characteristics shown by can be obtained. As can be understood from the above description, the characteristics of the FIR filters 1 _R and 1 _L have a transfer function indicating a transfer characteristic of a path in which a virtual sound image reaches the listener's left ear / right ear. On the other hand, it can be considered that the characteristic for canceling the crosstalk component reaching the listener's ear is obtained by performing the convolution calculation. That is, F shown in FIG.
The IR filters 6 _R (1) and 6 _L (1) and the FIR filter 7 forming the crosstalk cancel filter 11
Acoustic signal S ₁ comprising _R , 7 _L , 8 _R , 8 _L
The configuration equivalent to the system corresponding to is formed by two FIR filters 1 _R and 1 _L.

Then, the acoustic signal S ₁ forming the virtual sound image VS (1) is branched and input to the FIR filters 1 _R and 1 _L ,
The output of the FIR filter 1 _R is input to the right speaker SP (R), and the output of the FIR filter 1 _L is the left speaker SP.
Enter in (L). As a result, the speaker SP (R)
As the voice heard from (L), the crosstalk component that intersects the left and right systems is canceled, and the acoustic signal S _{1 in} which the transfer characteristics H _1R and H _1L shown in FIG. 2 are convoluted.
Will be equivalent to. Therefore, the listener can hear the virtual sound image VS (1) by the localization determined based on the transfer characteristics H _1R and H _1L shown in FIG. As described above, the sound reproducing device according to the present embodiment includes:
Two FIR filters are used to localize one virtual sound image.

Next, FIG. 3 shows a configuration of an audio reproducing device for localizing two virtual sound images based on the basic configuration of the audio reproducing device of the present embodiment shown in FIG. Here, as shown in FIG. 4, two virtual sound images VS (1)
Of (2), the transfer characteristics of the virtual sound image VS (1) reaching the right ear and the left ear of the listener LS are H _1R and H _1L , respectively.
The transfer characteristics of the virtual sound image VS (2) reaching the right ear and the left ear of the listener LS are H _2R and H _2L , respectively.

Here, the FIR filter 1 shown in FIG.
Similar to the case of deriving the characteristics of _R and 1 _L , consider the case where the two virtual sound images VS (1) and (2) shown in FIG. 4 are localized by the sound reproducing device having the configuration shown in FIG. , The system formula is that the acoustic signals to be the virtual sound images VS (1) and (2) are S ₁ and S ₂ , respectively, and the speaker SP
Each output of (R) and (L) is L and R,

(Equation 13) Can be represented by And this formula

[Equation 14] And can be transformed.

Based on this, the sound reproducing apparatus for localizing two virtual sound images according to the present embodiment can be configured as shown in FIG. In this case, F corresponding to the acoustic signal S ₁ that becomes the virtual sound image VS (1)
IR filter 1 _R (1), 1 _L (1), and virtual sound image V
FIR filters 1 _R (2) and 1 _L (2) corresponding to the acoustic signal S ₂ of S (2) are provided. And FIR
The characteristics of the filters 1 _R (1) and 1 _L (1) are the same as those shown in (Equation 11) and (Equation 12) even when based on (Equation 14). Moreover, as the characteristics of the FIR filters 1 _R (2) and 1 _L (2),
Based on, the FIR filter 1 _R (2) is

(Equation 15) FIR filter 1 _L (2) is

(Equation 16) The filter coefficient is set so as to obtain the characteristic indicated by.

The FIR filter 1 _R (1), 1 _L
The acoustic signal S ₁ is input to (1) and the FIR filter 1 _R
(2) The acoustic signal S ₂ is input to 1 _L (2), and FIR filter 1 _R (1) and FIR filter 1 _R
The output of (2) is input to the adder AD _R to perform addition, and the output is supplied to the right speaker SP (R). Also, FI
The outputs of the R filter 1 _L (1) and the FIR filter 1 _L (2) are input to the adder AD _L to be added and supplied to the left speaker SP (L), thereby configuring the sound reproducing device.

The sound reproducing device having such a configuration is shown in FIG.
FIR filters 6 _R (1), 6 _L (1) and 6 in
FIR filters 7 _R , 7 _L and 8 _R forming the crosstalk cancel filter 11 and a system that convolves the transfer characteristics for localizing the acoustic signals S ₁ and S ₂ composed of _R (2) and 6 _L (2) , 8 _L is equivalent to the circuit portion configured with four FIR filters 1 _R
It can be considered that it is formed by including (1), 1 _L (1), 1 _R (2), and 1 _L (2).

As a result, as for the voice heard by the listener from the speakers SP (R) (L), the crosstalk components of the left and right systems are canceled, and the localization determined based on the transfer characteristics shown in FIG. 4 is performed. Virtual sound image VS (1), V
You will hear S (2).

Based on the sound reproducing apparatus of FIGS. 1 and 3 which have been described so far, the sound reproducing apparatus of the present embodiment is capable of setting the localization for n virtual sound images. A general configuration is as shown in FIG. Here, as shown in FIG. 6, of the n virtual sound images VS (1) (2) to (n), the virtual sound image V
The transfer characteristics of S (1) reaching the right and left ears of the listener LS are H _1R and H _1L , respectively, and the transfer characteristics of the virtual sound image VS (2) are H _2R and H _2L. Virtual image VS
The transfer characteristics of (n) are H _nR and H _nL .

In this case, as shown in FIG. 5, the acoustic signals S ₁ and S ₂ to become virtual sound images VS (1) (2) to (n).
So as to correspond to each S _n, the pair respectively [FI
R filter 1 _R (1), 1 _L (1)], [FIR filter 1 _R (2), 1 _L (2)] to [FIR filter 1 _R
(N), 1 _L (n)] are provided. Where [FIR
Each characteristic of the filter 1 _R (1), 1 _L (1)] is represented by (Equation 11) and (Equation 12) described above, respectively. In addition, [FIR filter 1 _R (2), 1 _L (2)]
The characteristics of are also expressed by (Equation 15) and (Equation 16) shown above, respectively. Then, [FIR filter 1 _R
(N), 1 _L (n)], the FIR filter 1
_R (n) is

[Equation 17] The FIR filter 1 _L (n) is

(Equation 18) Will be represented by

The acoustic signal S ₁ is the FIR filter 1 _R
(1), 1 _L (1) is input and the acoustic signal S ₂ is FIR
Input to filters 1 _R (2) and 1 _L (2). The acoustic signal S _n is input to the FIR filters 1 _R (n) and 1 _L (n). The right speaker SP (R) has FIR filters 1 _R (1), 1 _R (2) to 1 _R.
The output obtained by adding the output of (n) by the adder AD _R is supplied. On the other hand, the left speaker SP (L) is supplied with an output obtained by adding the outputs of the FIR filters 1 _L (1), 1 _L (2) to 1 _L (n) by the adder AD _L. In this way, the sound reproducing device of the present embodiment for localizing n virtual sound images is configured. Then, by reproducing the acoustic signals S _{1 to} S _n by this acoustic reproduction device, as the sound heard by the listener from the speakers SP (R) (L), the crosstalk components of the left and right systems are canceled. , Virtual sound images VS (1), VS (2) to VS (n) by localization determined based on the transfer characteristics shown in FIG.
Is something you can hear.

For example, in the sound reproducing apparatus having the configuration shown in FIG. 10, 2 × n + 4 FIR filters are required as described above in order to localize n virtual sound images. On the other hand, in the sound reproducing device of the present embodiment, FIG.
As can be understood from the configuration described with reference to FIG. 4, 2n FIR filters are required to localize n virtual sound images. Specifically, for example, in most cases, it is considered that the acoustic reproduction device that localizes two virtual sound images is required as the actual usage pattern.
In order to localize two virtual sound images in the sound reproducing device having the configuration shown in (2), 2 × 2 + 4 = 8 FIR filters are required. On the other hand, in order to localize two virtual sound images with the sound reproducing device of the present embodiment, 2 × 2 = 4 FIR filters may be required as shown in FIG.

Further, the configuration of the system of the sound reproducing apparatus shown in FIG. 3 as the present embodiment can be regarded as equivalent to the system of the sound reproducing apparatus using the headphones shown in FIG. it can. That is, the characteristics of the FIR filters 1 _R (1), 1 _L (1), 1 _R (2), 1 _L (2) shown in FIG.

[Equation 19] FIR filter 1 _R (1), 1 _L
By setting the filter coefficients of (1), 1 _R (2), 1 _L (2), the FIR filter 1 _R (1), 1
_L (1), 1 _R (2) and 1 _L (2) are shown in FIG. 12, respectively.
The characteristics of the FIR filters 9 _R , 9 _L , 10 _R and 10 _L shown in FIG. Therefore, in the present embodiment, the configuration common to the system of the sound reproducing device shown in FIG. 3 makes it possible to configure the sound reproducing device corresponding to the case where the headphones are used. It should be noted that when performing reproduction by using the speaker or headphones by the sound reproducing device described as the embodiment so far, the reverse characteristic for flattening the transfer characteristic peculiar to the actually used speaker or headphone is used. A filter can be provided for filtering. By doing so, it becomes possible to more faithfully reproduce the three-dimensional sound field.

FIG. 7 conceptually shows the configuration when the sound reproducing device of the present embodiment is applied to a Dolby surround system. The Dolby Surround system includes, for example, a Left channel reproduced by localization of the left speaker SP (L), a Center channel reproduced by localization of the position of the virtual sound image VS (C), and localization of the position of the right speaker SP (R). Right played by
Each sound of the channel and the virtual sound image VS (S
R channel surround sound reproduced by position localization of R) and virtual sound image VS (SL) surround sound reproduced by position localization of five channels, that is, sound of five channels is reproduced together with video. If a system including a system for localizing five virtual sound images is used as the sound reproducing device of the present embodiment, two speakers S
Only P (L) (R) makes it possible to realize surround reproduction that reproduces the localization of the audio of the above 5 channels. In this case, the listener L located at the predetermined listening point based on the acoustic characteristics in the viewing room 101.
The transfer characteristics and crosstalk components of the above-mentioned audio signals of the five channels that are supposed to reach the S ears are obtained by measurement, and based on the measurement results of these transfer characteristics and crosstalk components, appropriate sound image localization for the above-mentioned five channels is performed. As obtained, the filter coefficient of the FIR filter which constitutes the system of the sound reproducing device is set. An image is projected on the screen 100.

Further, the virtual sound image VS (1) in the configurations of FIGS. 5 and 6 is taken as an example, and the virtual sound image VS is obtained.
(1) is the transfer characteristic H of the voice that reaches the right ear of the listener LS.
_{For 1R} , the characteristic that is given to the sound radiated by the speaker SP (R) is H _SP, and the transfer characteristic including the acoustic characteristics of the listening room until reaching the right ear of the listener LS is the component of H _room. Considering it separately, H _1R = H _SP · H
It can be represented as a _room . Then, in the sound reproducing device of the present embodiment, the transfer characteristic H _room including the characteristic H _SP of any speaker and the acoustic characteristic of the listening _room is set, and F is set based on the set characteristics H _SP and H _room.
By setting the filter coefficient of the IR filter, it is possible to obtain a feeling of hearing that is caused by the characteristics of the signal radiated by an arbitrary speaker, and a feeling of hearing when listening in a listening room with arbitrary acoustic characteristics. It is also possible to perform reproduction.

[0050]

As described above, the sound reproducing apparatus of the present invention uses 2n FIRs to localize n virtual sound images.
Since a filter is required, it is possible to obtain a sound reproducing device for obtaining a three-dimensional sound field with a smaller number of FIR filters than the conventional one, and in particular FIR
When the filter is configured by hardware, it has the effects of significantly reducing the cost, reducing the size and weight of the device, and reducing the power consumption. Also,
Since the FIR filter of the present invention has a filter characteristic in which the transfer characteristic for crosstalk cancellation is convoluted with the transfer characteristic of the path where the virtual sound image reaches the listener, It is possible to easily change the filter characteristics so that the transfer characteristics for crosstalk cancellation are invalidated by changing the filter coefficient, and only the transfer characteristics of the path where the virtual sound image reaches the listener are valid. However, due to this, even for a sound reproduction device corresponding to headphones that does not need to consider crosstalk as in the case of speaker output,
Hardware with a common FIR filter can be utilized, which also facilitates cost savings, for example. Further, in the case where one sound reproduction system is configured to support both a speaker and a headphone, the size and weight of the device can be further reduced.

[Brief description of drawings]

FIG. 1 shows a sound reproducing device according to an embodiment of the present invention:
It is a block diagram showing the composition of the system for locating two virtual sound images.

FIG. 2 is an explanatory diagram showing a transfer characteristic that a sound of one virtual sound image reaches a listener.

FIG. 3 is a block diagram showing a configuration of a system for localizing two virtual sound images as the present embodiment.

FIG. 4 is an explanatory diagram showing transfer characteristics in which sounds of two virtual sound images reach a listener.

FIG. 5 is a block diagram showing a configuration of a system for localizing n virtual sound images according to the present embodiment.

FIG. 6 is an explanatory diagram showing transfer characteristics in which sounds of n virtual sound images reach a listener.

FIG. 7 is an explanatory diagram showing a configuration when the sound reproducing device of the present embodiment is applied to a Dolby surround system.

FIG. 8 is a block diagram showing a configuration of a conventional crosstalk cancellation filter.

9 is a block diagram showing a configuration of a crosstalk cancellation filter that is equivalent to FIG. 8 when left and right systems are targeted.

10 is a block diagram showing a configuration of a system for locating n virtual sound images based on the configuration of the crosstalk cancellation filter shown in FIG.

FIG. 11 is an explanatory diagram showing transfer characteristics in which sounds of n virtual sound images reach a listener.

FIG. 12 is a block diagram showing a configuration of an audio reproducing device compatible with the use of headphones.

FIG. 13 is an explanatory diagram showing a transfer characteristic that a sound of a virtual sound image reaches a listener when headphones are used.

FIG. 14 is a block diagram showing a configuration of an FIR filter.

[Explanation of symbols]

1 _R , 1 _L , 1 _R (1), 1 _L (1), 1 _R (2), 1
_L (2), 1 _R (n), 1 _L (n) FIR filter,
VS (1)-(n) virtual sound image, LS listener, AD
_R , AD _L adder, SP (R) right speaker, SP
(L) Left speaker

Claims (8)

[Claims] 1. A cross of a virtual sound image which is crosstalked to the right ear of the listener with respect to a transfer function of a path which is supposed to reach the right ear of the listener from a localization position of one virtual sound image. First digital filter means having, as a filter characteristic, a transfer function obtained by convolving a transfer function obtained so as to cancel the talk component, and reaching the left ear of the listener from the localization position of the one virtual sound image. The transfer function obtained by convolving the transfer function obtained so as to cancel the crosstalk component of the virtual sound image that is supposed to reach the listener's left ear by crosstalk with the transfer function of the path. A second digital filter means having a function as a filter characteristic, and a sound reproducing device. 2. The sound reproducing apparatus according to claim 1, wherein the output signals of the first digital filter means and the second digital filter means are output as audio by a speaker, respectively. 3. The filter coefficients of the first digital filter and the second digital filter means are set so as to invalidate the transfer function obtained so as to cancel the crosstalk component, and at the same time, The sound reproducing device according to claim 1, wherein the output signals of the first digital filter means and the second digital filter means are output as sound by headphones. 4. The sound reproducing device according to claim 1, wherein the first digital filter means and the second digital filter means are FIR filters. 5. A cross of a virtual sound image that is supposed to reach the right ear of the listener by crosstalk with respect to a transfer function of a path that reaches the right ear of the listener from a localization position of one virtual sound image. First digital filter means having as a filter characteristic a transfer function obtained by convoluting a transfer function obtained so as to cancel the talk component, and reaching the left ear of the listener from the localization position of the one virtual sound image. The transfer function obtained by convolving the transfer function obtained so as to cancel the crosstalk component of the virtual sound image that is supposed to reach the left ear of the listener with the transfer function of the path. A pair of second digital filter means having a function as a filter characteristic is provided corresponding to each of a plurality of virtual sound images, and A first addition means for adding and outputting the outputs of the first digital filter means provided for each of the plurality of virtual sound images, and a second addition means provided for each of the plurality of virtual sound images A second adding means for adding and outputting the outputs of the digital filter means, and the sound reproducing device. 6. The sound reproducing apparatus according to claim 5, wherein the output signals of the first adding means and the second adding means are output as voices from a speaker, respectively. 7. The filter coefficients of the first digital filter and the second digital filter means are set so as to invalidate the transfer function obtained so as to cancel the crosstalk component, and the first digital filter and the second digital filter means are set. 1
6. The sound reproducing device according to claim 5, wherein the output signals of the adding means and the second adding means are output by headphones. 8. The sound reproducing apparatus according to claim 5, wherein the first digital filter means and the second digital filter means are FIR filters.