JP3255580B2 - Stereo sound image enlargement device and sound image control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a stereo sound image enlarging device for enlarging a sound image based on a stereo signal and a sound image control device for localizing a sound image based on a monaural signal at an arbitrary three-dimensional position in two-channel speaker reproduction. .

[0002]

2. Description of the Related Art Conventionally, there has been known a technique for generating a stereo signal for both left and right channels, supplying the stereo signals to left and right speakers, and generating sound simultaneously to localize a sound image. This sound image localization technique localizes the sound image mainly by changing the balance of the left and right sound volumes, and was able to localize the sound image only between the left and right speakers.

However, in recent years, several techniques for localizing a sound image outside the left and right speakers have been developed. First
In the two-channel speaker reproduction, the left channel sound is mixed with the opposite phase sound of the right channel sound, and the right channel sound is mixed with the opposite phase sound of the left channel sound. The sound image is localized outside the left and right speakers. As a technique using this kind of technology, for example, WO94 / 16538 (PCT / US93)
No./12688), “Sound image operating device and sound image enlarging method.
(SOUND IMAGE MANIPULATION APPARATUS AND METHOD FOR
SOUND IMAGE ENHANCEMENT) ". Hereinafter, this is referred to as “first prior art”. According to the first prior art, the sound image can be localized outside the left and right speakers, and the stereo sound image can be enlarged.

[0004] As a second technique, a sound image localization technique called a Schrader method is widely known. This Schrader system creates a situation like headphone listening 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 called "crosstalk sounds"). . This allows
The sound image can be localized not only between the two speakers but also at any other position, for example, right beside the listener. Hereinafter, this is referred to as “second prior art”.

Further, as a third technique, a technique is known in which the addition of a head-related acoustic transfer function and the crosstalk canceling process are performed by a convolution operation, thereby localizing a sound image at an arbitrary position (for example, "RSS"). About ”, Roland Corp., The Acoustical Society of Japan, 48, 9
issue). Hereinafter, this is referred to as “third prior art”.

[0006]

However, in the first prior art described above, when listening at a position distant from the speakers, the listener may feel that the sound image is localized outside the left and right speakers. However, when listening at a position close to the speaker, there is a problem that the listener cannot clearly feel the position where the sound image is localized.

In the first prior art, when the mixing ratio of the opposite phase sound of the other channel to the sound of the channel is increased, the monaural component of the stereo input forms a comb filter, and the sound field is enlarged. However, there is a problem that sound quality is deteriorated. This inferior sound quality mainly appears as a phenomenon in which the sound is felt like a sound without a low frequency region.

In the second prior art, there is a problem that the sound image is felt to be localized at the listener's ear and unnaturalness remains.

In the third prior art, when the addition of the head acoustic transfer function and the crosstalk cancellation processing are realized by a convolution operation, an extremely long number of convolution steps is required, and the scale of hardware is increased. There's a problem. If the number of convolution steps is reduced in order to solve such a problem, the accuracy of the head acoustic transfer function in the low range is reduced, and the accuracy of crosstalk cancellation is significantly reduced.

[0010] Therefore, a first object of the present invention is to provide a two-channel loudspeaker which allows the user to clearly sense the position where the sound image is localized even when listening at a position close to the loudspeaker. An object of the present invention is to provide a stereo sound image enlarging device that can remove unnaturalness by localizing a sound image at a position away from a listener and can obtain a richer feeling of spread.

It is a second object of the present invention to provide a stereo sound image enlarging apparatus which does not cause sound quality deterioration even if the stereo sound image enlarging effect is enhanced.

A third object of the present invention is to provide a sound image control apparatus capable of adding a head acoustic transfer function and improving the accuracy of crosstalk cancellation with a small-scale hardware without performing a convolution operation. Is to do.

[0013]

According to a first aspect of the present invention, a stereo sound image enlarging apparatus according to a first aspect of the present invention, as shown in FIG. A first left signal Lm1 from which the left crosstalk signal has been removed and a first left signal Lm1 from which the right crosstalk signal has been removed from the right input signal Rin.
And a second left signal Lm2 obtained by mixing the left input signal Lin with a signal having the opposite phase of the right input signal Rin, and a right input signal Rin having the opposite phase of the left input signal Lm1. A negative-phase signal generating means for generating a second right signal Rm2 by mixing the signals; a first left signal Lm1 from the crosstalk canceling means; and a second left signal Lm2 from the negative-phase signal generating means. Mixing to produce a left output signal Lout, and a first right signal Rm1 from the crosstalk canceling means and a second right signal Rm2 from the antiphase signal generating means.
And a mixing means for generating the right output signal Rout by mixing.

Here, the left crosstalk signal refers to a signal corresponding to a left crosstalk sound reaching the listener's right ear from the left speaker SPL as shown in FIG. 2, and the right crosstalk signal is A signal corresponding to the right crosstalk sound reaching the left ear of the listener from the right speaker SPR. The sound that reaches the listener's left ear from the left speaker SPL is called a left direct sound, and the sound that reaches the listener's right ear from the right speaker SPR is called a right direct sound.

The crosstalk canceling means acts to localize a sound image formed by the left input signal Lin and the right input signal Rin to the ear. Further, the negative-phase signal generating means acts to spread the sound image formed by the left input signal Lin and the right input signal Rin to the outside of the left and right speakers. Accordingly, in the mixing means, the left output signal Lout generated by mixing the first left signal Lm1 and the second left signal Lm2 at an appropriate ratio, and the first right signal Rm1 and the second right signal Lm1 are mixed. If the sound is generated based on the right output signal Rout generated by mixing Rm2 and Rm2 at an appropriate ratio, the sound image can be located at any position in a wide range from the side of the listener to the front of the listener, And it can be localized at a position away from the listener.

The above-mentioned crosstalk canceling means is shown in FIG.
, A left crosstalk signal generating means for generating a left crosstalk signal based on the left input signal Lin, and a right crosstalk signal generating means for generating a right crosstalk signal based on the right input signal Rin Means, a left channel (hereinafter abbreviated as “Lch”) cross computing unit for subtracting a right crosstalk signal from the left input signal Lin, and a first signal obtained by filtering a signal from the Lch cross computing device. , A right channel (hereinafter abbreviated as “Rch”) cross calculator for subtracting the left crosstalk signal from the right input signal Rin, An Rch cross filter 2 for filtering a signal from the arithmetic unit to generate a first right signal Rm1.

The above-mentioned left crosstalk signal generating means is provided with Lc
It can be constituted by an attenuator for h cross, a delay unit for Lch cross and a filter 1 for Lch cross. The left crosstalk signal generation means attenuates the left input signal Lin with the Lch cross attenuator, delays the attenuated signal with the Lch cross delay, and converts the delayed signal into the Lch cross delay.
A left crosstalk signal is generated by filtering with the cross filter 1.

Similarly, the right crosstalk signal generating means can be composed of an Rch cross attenuator, an Rch cross delay and an Rch cross filter 1. This right crosstalk signal generation means converts the right input signal Rin to R
The signal is attenuated by the channel cross attenuator, the attenuated signal is delayed by the Rch cross delay unit, and the delayed signal is filtered by the Rch cross filter 1 to generate a right crosstalk signal.

The gains of the Lch cross attenuator and the Rch cross attenuator can be made variable within a range of 0 to 1. Further, the Lch cross filter 1 and the Rch cross filter 1 can each be constituted by a primary IIR type filter. Further, the respective delay amounts of the Lch cross delay device and the Rch cross delay device are as follows:
This can be set to about 8 sampling points when an analog signal is sampled at 48 KHz to generate a digital signal.

The filter coefficients of the Lch cross filter 2 and the Rch cross filter 2 are Lch cross filters.
A gain of the cross attenuator or the Rch cross attenuator,
It can be generated based on the filter coefficient of the Lch cross filter 1 or the Rch cross filter 1.

As shown in FIG. 4, the negative phase signal generating means includes an Lch negative phase attenuator for attenuating the left input signal Lin, and an Lch negative phase attenuator for delaying the output signal from the Lch negative phase attenuator. And a delay R for attenuating the right input signal Rin
a channel negative phase attenuator, an Rch negative phase delay that delays an output signal from the Rch negative phase attenuator, and an Rch negative phase delay from the output signal from the Lch negative phase attenuator. From the output signal from the Lch anti-phase attenuator to generate an output signal from the Lch anti-phase attenuator from the output signal from the Rch anti-phase attenuator. And an Rch anti-phase operation unit that generates the second right signal Rm2 by subtraction.

The Lch anti-phase attenuator attenuates the left input signal Lin and supplies it to the Lch anti-phase operation unit and the Lch anti-phase delay. The Lch anti-phase delay device delays the attenuated left input signal Lin for a predetermined time and supplies the delayed input signal Lin to the Rch anti-phase operation unit. Similarly, the Rch anti-phase attenuator outputs the right input signal R
a is attenuated and supplied to the Rch reverse phase operation unit and the Rch reverse phase delay unit. The Rch anti-phase delay unit delays the attenuated right input signal Rin for a predetermined time and supplies the delayed input signal to the Lch anti-phase operation unit.

The Lch antiphase arithmetic unit subtracts the output signal from the Rch antiphase delay from the output signal from the Lch antiphase attenuator. The signal obtained by this subtraction is a signal obtained by attenuating and delaying the right input signal Rin with respect to the left input signal Lin and further mixing the inverted signal. The output signal from the Lch antiphase arithmetic unit is supplied to the above-described mixing means as a second left signal Lm2. Similarly, the Rch antiphase arithmetic unit calculates the Lch from the output signal from the Rch antiphase attenuator.
The output signal from the anti-phase delay device is subtracted. The signal obtained by the subtraction is a signal obtained by mixing the right input signal Rin with the left input signal Lin, which is attenuated and delayed, and further in phase. The output signal from this Rch antiphase arithmetic unit is
Is supplied to the above-mentioned mixing means as the right signal Rm2 of

Each gain of the Lch negative phase attenuator and the Rch negative phase attenuator can be made variable in the range of 0-1. Further, each delay amount of the Lch anti-phase delay device and the Rch anti-phase delay device is 8 in a case where an analog signal is sampled at 48 KHz to generate a digital signal.
It can be about the sampling point.

The mixing means, as shown in FIG.
It can be composed of an h mixer and an Rch mixer. Lc
The h mixer includes a first left signal Lm1 from the crosstalk canceling unit and a second left signal Lm2 from the negative phase signal generating unit.
And mixing. An adder can be used as the Lch mixer. The output signal from the Lch mixer is output to the outside as a left output signal Lout. Similarly, the Rch mixer mixes the first right signal Rm1 from the crosstalk canceling unit and the second right signal Rm2 from the negative-phase signal generating unit. An adder can be used as the Rch mixer. The output signal from the Rch mixer is output to the outside as a right output signal Rout.

The crosstalk canceling means, the negative-phase signal generating means and the mixing means can be constituted by, for example, processing of a digital signal processor (DSP). In this case, it is preferable to perform time-division multiplexing in order to ensure real-time signal processing.

According to the stereo sound image enlarging device having the above-described configuration, the sound generated based on the left input signal Lin is converted to the sound generated based on the right input signal Rin only by the left ear by the crosstalk canceling means. It can be incident only on the right ear. Therefore, the panned sound image can be felt near both ears.

Also, the left-channel input sound is localized further to the left of the left speaker by the reverse-phase signal generating means,
The right channel input sound can be localized further to the right of the right speaker. Therefore, the panned sound image can be felt outside the left and right speakers.

Then, the localization formed near both ears by the crosstalk canceling means and the localization formed outside both speakers by the opposite-phase signal generation means are mixed by the mixing means, so that the reception has never been achieved before. A sound field expanded to the side of the listener is obtained. Also,
By changing the mixing ratio, the magnitude of the stereo sound image enlargement effect can be changed.

In order to achieve the second object, the stereo sound image enlarging device according to the second aspect of the present invention performs a predetermined process on the left input signal Lin and the right input signal Rin to respectively execute the left output signal Lout A stereo sound image enlarging device for outputting as a right output signal Rout and the left input signal Lin;
Lch mixing means for mixing a signal obtained by multiplying the left input signal Lin by a gain e, a signal obtained by multiplying the right input signal Rin by a gain e, delaying the signal, and further making the signal out of phase, and the right input signal Rin; Rch mixing means for mixing a signal obtained by multiplying the right input signal Rin by a gain e and a signal obtained by multiplying and delaying the left input signal Lin by a gain e and further out of phase, and an output from the Lch mixing means. A low-pass filter is applied to the signal obtained by multiplying the signal by the gain (1-e) and the output signal of the Lch mixing means, and the gain e
And a signal obtained by multiplying the left output signal Lou
a signal obtained by multiplying an output signal from the Rch mixing means by a gain (1-e) and a signal obtained by applying a low-pass filter to the output signal of the Rch mixing means and multiplying the output signal by a gain e And Rch sound quality correction means for outputting the addition result as a right output signal Rout.

The second stereo sound image enlarging apparatus according to the present invention further comprises a cross signal for generating a left signal and a right signal corresponding to the sound from which the crosstalk sound has been removed based on the left input signal Lin and the right input signal Rin, respectively. The Lch mixing unit further includes a talk cancellation unit, wherein the Lch mixing unit includes a left signal from the crosstalk cancellation unit, a signal obtained by multiplying the left input signal Lin by a gain e, and a gain e added to the right input signal Rin.
, And the signals having the opposite phases are mixed. The Rch mixing means includes: a right signal from the crosstalk cancel means; a signal obtained by multiplying the right input signal Rin by a gain e; The input signal Lin can be configured to be multiplied by a gain e, delayed, and mixed with a signal having a reversed phase.

In order to achieve the third object, the sound image control apparatus according to the third aspect of the present invention transmits the head input for the left channel to the monaural input signal Min as shown in FIG. A left-channel head-related transfer function providing means for providing a function and outputting as a left input signal Lin, and a right-channel head-related transfer function to the monaural input signal Min;
A right channel head-related transfer function providing means for outputting as a right input signal Rin, a first left signal Lm1 obtained by removing a left crosstalk signal from the left input signal Lin from the left channel head-related transfer function providing means, and the right channel Crosstalk canceling means for generating a first right signal Rm1 by removing a right crosstalk signal from the right input signal Rin from the head-related transfer function providing means, and a signal having an opposite phase to the right input signal Rin as the left input signal Lin. A second right signal R obtained by mixing a signal having a phase opposite to that of the left input signal Lm1 with the mixed second left signal Lm2 and right input signal Rin.
m2, and a left output signal Lou by mixing the first left signal Lm1 from the crosstalk canceling means and the second left signal Lm2 from the negative phase signal generating means.
mixing means for generating a right output signal Rout by generating a first right signal Rm1 from the crosstalk canceling means and a second right signal Rm2 from the negative phase signal generating means; Having.

The present sound image control apparatus is the first embodiment of the present invention described above.
A left-channel head-related transfer function assigning means for assigning a left-channel head-related transfer function to the monaural input signal Min, and a right-channel head to the monaural input signal Min. Right channel head-related transfer function providing means for providing a transfer function. Therefore, the crosstalk canceling means, the anti-phase signal generating means, and the mixing means can be configured in the same manner as in the case of the stereo sound image enlarging device according to the first aspect of the present invention described above.

The left channel head-related transfer function assigning means and the right channel head-related transfer function assigning means are respectively shown in FIG.
As shown in (1), a filter for a direct unit for filtering a monaural input signal Min, a binaural time difference delay unit for delaying an output signal from the filter for a direct unit, and a filter for a reflector for filtering the monaural input signal Min A plurality of delay units D _{1 to} D _n for delaying an output signal from the reflection unit filter; and a plurality of multipliers G _{1 to} G _n for amplifying output signals from the delay units D _{1 to} D _n. A reflector adder for adding output signals from the plurality of multipliers G _{1 to} G _n ;
An adder for adding the output signal from the binaural time difference delay unit and the output signal from the reflector adder can be used.

The left channel head-related transfer function providing means outputs a left input signal Lin, and the right channel head-related transfer function providing means outputs a right input signal Rin.

The direct part filter simulates a frequency characteristic of a head-related transfer function of a sound that directly reaches a listener's ear from a sound source (for example, a speaker). The binaural time difference delay device simulates a time difference in the binaural sound from a sound source that directly reaches a listener's ear. The reflection unit filter simulates a frequency change due to reflection in the room. Delay devices D _{1 to} D _n , multipliers G ₁ to
The _Gn and the reflector adder reproduce the arrival time of the reflected sound in the room to both ears and the loudness of the sound at the time of arrival. Then, the adder adds the signal corresponding to the direct sound from the binaural time difference delay device and the signal corresponding to the reflected sound from the reflector adder, and adds the addition result to the left input signal Lin or the right input signal. Output as a signal Rin.

The left channel head-related transfer function providing means and the right channel head-related transfer function providing means are, for example, a DSP
Can be configured. In this case, it is preferable to perform time-division multiplexing in order to ensure real-time signal processing.

As described above, since the head-related transfer function is provided to the monaural input signal Min by the left-channel head-related transfer function providing means and the right-channel head-related transfer function providing means, the sound image is localized at an arbitrary position when listening to the headphones. be able to.

Further, by having the crosstalk canceling means, it is possible to make the signal input to the Lch enter only the left ear and the signal input to the Rch enter only the right ear. Therefore, the listener is in a situation equivalent to headphone listening, so that a sound image can be localized at an arbitrary position in speaker reproduction.

Further, the presence of the anti-phase signal generating means makes it possible to make the sound image appear at a position away from the listener, as compared with a sound image control apparatus comprising only the head-related transfer function applying means and the crosstalk canceling means. it can.

Further, each signal generated by the crosstalk canceling means and the antiphase signal generating means can be weighted by the mixing means. By mixing the signal from the crosstalk canceling means and the signal from the negative-phase signal generating means at an appropriate ratio, it is possible to localize the sound image at a long distance in the two-channel speaker reproduction. When the weight of the signal from the opposite-phase signal generating means is set to zero, theoretically accurate sound image control processing can be performed. Further, when the weight of each signal from the crosstalk canceling means and the negative-phase signal generating means is set to zero, sound image control can be performed by listening to headphones.

The sound image control apparatus according to the third aspect of the present invention further comprises direction instructing means for instructing a direction in which the sound image is to be moved, and the left channel head-related transfer function providing means is provided in the first direction. First left-channel head-related transfer function providing means for providing a head-related transfer function for the corresponding left channel, and first right channel for providing a head-related transfer function for the right channel corresponding to the first direction Head-related transfer function providing means, second left-channel head-related transfer function providing means for providing a head-related transfer function for the left channel corresponding to the second direction, and right-hand channel transfer function providing means for the right channel corresponding to the second direction Second head-related transfer function providing means for providing a head-related transfer function, and the first left-channel head-related transfer function providing means and the first right channel in response to an instruction from the direction indicating means. With head related transfer function First weighting means for multiplying each signal from the means by a weighting factor α, the second left channel head-related transfer function providing means and the second right channel head-related transmission in response to an instruction from the direction indicating means A second weighting means for multiplying each signal from the function giving means by a weighting factor (1−α), and a first weighting means weighted by the first weighting means
Is mixed with the signal from the second left-channel head-related transfer function assigning means weighted by the second weighting means, thereby generating a left input signal. Left mixing means, a signal from the first right channel head-related transfer function weighting means weighted by the first weighting means, and a second right channel head-related transfer function weighted by the second weighting means And right mixing means for mixing the signal from the means to generate a right input signal. Where 0 ≦ α ≦
It is one.

As the direction indicating means, it is possible to use a device capable of inputting continuous values, such as a joystick, a slide volume, and a rotary volume. The instruction by the direction instructing means includes a weight coefficient α
Is reflected in

Now, it is assumed that the sound image based on the signal to which the head-related transfer function is provided by the first left-channel head-related transfer function providing means and the first right-channel head-related transfer function providing means is localized in a predetermined direction. Assume. In this case, the weighting coefficient α of the first weighting means is “1”, and the weighting coefficient α of the second weighting means is “0”. Here, when a new direction is instructed by the direction instructing means, the second left-channel head-related transfer function providing means and the second right-channel head-related transfer function providing means cause the head-related transfer function corresponding to the new direction. Is set to be given. Then, the weighting coefficient α sequentially changes from “0” to “1”. Thus, each signal from the first left channel head-related transfer function providing means and the first right channel head-related transfer function providing means, and the second left channel head-related transfer function providing means and the second right channel head-related transfer function providing means. Each signal from the partial transfer function providing means is cross-fade mixed and output. Then, when the weighting coefficient α becomes “1”, only each signal from the second left-channel head-related transfer function providing unit and the second right-channel head-related transfer function providing unit is output, and a new direction is obtained. The sound image is localized.

As described above, since the sound image is moved by the cross-fade mix, the sound image moves smoothly,
Generation of noise is suppressed.

[0046]

Embodiments of the present invention will be described below in detail with reference to the drawings.

(Embodiment 1) Before describing an embodiment of a stereo sound image enlarging apparatus according to the first aspect of the present invention, in order to facilitate understanding of the present invention, first, a known Schroeder system is used. Will be described.

FIG. 6A shows a path from both the left and right speakers SPL and SPR to both ears of the listener. In the figure, S is a transfer function from a speaker to an ear on the same side as the speaker. A is a transfer function from the speaker to the ear opposite to the speaker. As described above, in the two-channel speaker reproduction, characteristics of S and A are added before the sound generated by the speaker reaches both ears. For example, even if a sound based on the left input signal Lin is output only from the Lch speaker SPL, SLin is incident on the left ear and ALin is incident on the right ear.

FIG. 6B shows the configuration of a Schroeder type crosstalk canceling device. It is assumed that the left input signal Lin is input to Lch and the right input signal Rin is input to Rch.

When the signal of the sound reaching the left ear (left ear signal) is calculated according to the configuration shown in FIG. 6B, the following equation (1) is obtained.
Is obtained.

(Equation 1)

Similarly, when the signal of the sound reaching the right ear (right ear signal) is calculated according to the configuration shown in FIG. 6B, the result of the following equation (2) is obtained.

(Equation 2)

From the above equations (1) and (2), according to the Schrader method, the sounds corresponding to the left input signal Lin and the right input signal Rin are left and right irrespective of the transfer functions S and A, respectively. It can be understood that the light is incident on each of the right ears and the crosstalk sound is cancelled.

Next, a first example of the crosstalk canceling means of the present invention will be described with reference to the drawings.

FIG. 7A shows a path from both the left and right speakers SPL and SPR to the listener's both ears. It is assumed that the transfer function S in the Schrader method is “1” in the first example. Similarly, it is assumed that the transfer function A in the Schroeder method is “aZ ⁻ⁿ L ₀ ” in the first example. Here, a is the gain, n is the delay time of the crosstalk sound with respect to the direct sound, and L ₀ is a low-pass filter that simulates the diffraction of the crosstalk sound by the listener's head.

According to the above assumption, S = 1 and A = aZ ^-n L ₀
Therefore, C can be represented by the following equation (3).

(Equation 3)

When this is applied to a Schroeder type crosstalk canceling device (see FIG. 6B), the configuration of the crosstalk canceling means of the present invention is shown in FIG. 7B.
It becomes as shown in.

If the low-pass filter is constituted by a primary IIR filter, the low-pass filter L ₀ can be expressed by the following equation (4).

(Equation 4)

When this equation (4) is applied to FIG. 7 (b),
The crosstalk canceling means of the present invention can be represented as shown in FIG. This is shown in a block diagram in FIG. 9, which can be realized by DSP processing.

In FIG. 9, the Lch cross attenuator, the Lch cross delay and the Lch cross filter 1 constituting the left crosstalk signal generating means of the present invention are a multiplier 10, a delay unit 11 and an IIR type filter, respectively. It comprises a first-order low-pass filter 12. Similarly, the Rch cross attenuator, the Rch cross delay, and the Rch cross filter 1 constituting the right crosstalk signal generating means of the present invention are a multiplier 20, a delay 21, and an IIR primary low-pass filter 22, respectively. It is composed of Since the primary low-pass filter 22 has the same configuration as the primary low-pass filter 12, it is not shown.

In FIG. 9, a is the gain of the multipliers 10 and 20, b to d are the filter coefficients of the primary low-pass filters 12 and 22, and n is the delay amount of the delay units 11 and 21.

The arithmetic unit for Lch cross and the arithmetic unit for Rch cross of the crosstalk canceling means can be constituted by subtractors 13 and 23, respectively. Subtractor 13
Subtracts the right crosstalk signal from the left input signal Lin. The output signal of the subtractor 13 is supplied to the filter 14. Similarly, the subtractor 23 subtracts the left crosstalk signal from the right input signal Rin. The output signal of the subtracter 23 is supplied to a filter 24. In addition, the filter 24
Since the configuration is the same as that of the filter 14, the illustration is omitted.

The filter coefficients and the delay amounts of the filters 14 and 24 are created based on the above a to d and n.

When the low-pass filter is constituted by a second-order IIR type filter, assuming that S = 1 and A = aZ ^-n L ₀ , the low-pass filter L ₀ is constructed as in the following equation (5). By doing so, a crosstalk canceling means can be configured.

(Equation 5)

This case can also be realized by DSP processing. The same applies to the case where the low-pass filter is constituted by a third-order or higher-order IIR filter.

As described above, the crosstalk canceling means sets the characteristic from the speaker to the ear on the same side as “1” and the characteristic from the speaker to the opposite ear on the basis of the gain a, the delay amount n, and the low-pass filter L _0. As a result, the Schroeder crosstalk canceling device can be simplified, so that highly accurate crosstalk cancellation can be performed with a simple configuration.

Next, the second of the crosstalk canceling means
Will be described with reference to FIG.

FIG. 10A shows both left and right speakers SPL.
And the path from the SPR to the listener's both ears. In the figure, Hm is the frequency characteristic of the ear on the same side as the speaker. “AZ ⁻ⁿ Hs” is the frequency characteristic of the ear opposite to the speaker. At this time. H for Schrader type S
And A can be represented by aZ ⁻ⁿ Hs. Therefore,
C can be represented by the following equation (6).

(Equation 6)

Here, Hs / Hm can be considered as the characteristic of the ear on the opposite side when the frequency characteristic of the ear on the same side as the speaker is set to “1”. Can be represented by the following equation (7). Therefore, 1 / (1-C
The item ² ) is the same as in FIG.

(Equation 7)

In the Schroeder system, since the term of 1 / S exists in the last stage, in the second example, 1 / Hm
Will enter. Therefore, the crosstalk canceling means of the second example is as shown in FIG. Here, if the characteristic of Hm is realized by a primary filter,
Hm can be represented by the following equation (8).

(Equation 8) Therefore,

(Equation 9) Becomes

The filter L ₀ can be expressed by the following equation (10), as in the first example described above.

(Equation 10)

When this equation (10) is substituted into FIG. 10B, the present crosstalk canceling means can be represented as shown in FIG. This is shown in a block diagram in FIG. 12, which can be realized by DSP processing. The filter L ₀ of the second example can also be constituted by a second or higher order filter.

In the crosstalk canceling means of the second example shown in FIG. 12, a filter 15 is further provided on the output side of the filter 14 shown in FIG. Is different from the crosstalk canceling means of the first example.

As described above, the crosstalk canceling means assumes that the characteristic from the speaker to the ear on the same side is Hm, and the characteristic from the speaker to the opposite ear is gain a, delay amount n and Hs. This makes it possible to simplify the Schroeder-type crosstalk canceling device, so that high-quality and high-accuracy crosstalk cancellation that can be realized with a simple configuration can be performed.

Next, the reverse phase signal generating means will be described. FIG. 13 shows the configuration of the anti-phase signal generation means. In FIG. 13, the Lch anti-phase attenuator, the Lch anti-phase delay, and the Lch anti-phase operation unit are respectively a multiplier 30 and a delay unit 3.
1 and an adder 32. Similarly, the Rch anti-phase attenuator, the Rch anti-phase delay, and the Rch anti-phase operation unit are respectively a multiplier 40, a delay unit 41, and an adder 42.
It is composed of

In FIG. 13, e is the gain of the multipliers 30 and 40, and m is the delay amount of the delay units 31 and 41. m
Is the number of sampling points when an analog signal is sampled at 48 KHz to generate a digital signal, and for example, m can be set to about 6.

Now, consider the case where only the left input signal Lin is given. At this time, the second left signal Lm2 is "eLin-e
Z ^-m Rin ". Similarly, the second right signal Rm2 can be represented by "eRin-eZ- ^m Lin".

Assuming that m is the time difference between the time of reaching the ear on the speaker side from the speaker and the time of reaching the ear on the opposite side (the delay time of the crosstalk sound with respect to the direct sound), for example, the signal output from the left speaker SPL is output. Among the sounds, the crosstalk sound is incident on the right ear with a delay of m. Therefore, if the crosstalk sound is delayed by m from the Rch speaker and inverted and output, the sounds cancel each other around the right ear. This processing causes a phenomenon similar to the above-described crosstalk cancellation processing.

However, in this anti-phase signal generation means, the gain,
Since the frequency change due to diffraction is not taken into account, the crosstalk cancellation is not performed accurately.
When the crosstalk cancellation is performed correctly, the sound image seems to cling to the ear. On the other hand, in the case of imperfect crosstalk cancellation by the anti-phase signal generation means having the above configuration, the sound image is felt not at the ear but around the outside of the speaker. However, only the configuration shown in FIG. 13 has a problem that the clarity of the sound image is poor.

This problem can be solved by configuring the anti-phase signal generation means as shown in FIG. The anti-phase signal generation means shown in FIG. 14 further mixes the left input signal Lin with the second left signal Lm2 in the anti-phase signal generation means shown in FIG. 13, and further inputs the right input signal into the second right signal Rm2. The signal Rin is mixed. Thus, the sound image can be clearly localized outside the left and right speakers.

As described above, the present anti-phase signal generating means can localize the sound image outside both speakers due to imperfect crosstalk cancellation.

Next, an embodiment of the mixing means is shown in FIG.
This will be described with reference to FIG. The mixing means is composed of the Lch mixer and the Rch mixer as described above.
The Lch mixer can be configured by the adder 50.
The adder 50 receives the first signal from the crosstalk canceling unit.
Left signal Lm1 and the second left signal L
m2, and outputs the result as a left output signal Lout. Similarly, the Rch mixer can be configured by the adder 51. The adder 51 adds the first right signal Rm1 from the crosstalk canceling unit and the second right signal Rm2 from the negative phase signal generating unit, and outputs the result as a right output signal Rout.

The first left signal Lm1 from the crosstalk canceling means and the second left signal Lm1 from the negative phase signal generating means
The mixing ratio of m2 (also the mixing ratio of the first right signal Rm1 from the crosstalk canceling means and the second right signal Rm2 from the negative phase signal generating means) is the same as that of the multipliers 10 and 20 of the crosstalk canceling means. The gain is adjusted according to the magnitudes of the gain a and the gain e of the multipliers 30 and 40 of the negative phase signal generation means. That is, the amount of crosstalk cancellation is determined by the magnitude of the gain a, and the amount of mixing of the opposite-phase signals is determined by the gain e.

Depending on the values of the gains a and e, the following sound field can be formed. When a = e = 0, the left input signal Lin becomes the left output signal Lout as it is, and the right input signal Rin becomes the right output signal Rout as it is. The filters 14 and 24 of the crosstalk canceling means (see FIGS. 9 and 12)
Since it is 0, the coefficients of the filters 14 and 24 are symmetrical and operate as a filter that outputs the input signal as it is. When a ≠ 0 and e = 0, only the crosstalk canceling means is effective, and the sound field is such that the sound image can be heard at the ear. When a = 0 and e ≠ 0, only the anti-phase signal generating means is valid, and the sound image is a sound field localized outside the two speakers. In the case of a ≠ 0 and e 音 0, the sound field where the sound image can be heard at the ear due to the crosstalk cancellation and the sound field where the sound image is localized outside the two speakers by mixing the anti-phase signals overlap each other. The sound field is a sound field that can be localized in a wide direction from the side of the listener to the front of the listener and at a position away from the listener.

In the above case, in the crosstalk cancellation, the amount of delay changes depending on the opening angle of the speaker.
There is a problem that the clarity of sound image localization differs between when listening at a position close to the speaker and when listening at a position away from the speaker. Therefore, by setting the delay amount m in the opposite phase signal generation means to a value different from n, it is possible to maintain the clarity of the sound image localization due to the opening angle of the speaker. Therefore, the listening point can be selected in a wide range. These n and m can be, for example, n = 8 and m = 6. Note that n and m may have the same value. In this case, the best listening point is determined at a predetermined speaker opening angle.

Next, an example of a stereo sound image enlarging system using the above-described stereo sound image enlarging apparatus will be described with reference to FIG.
This will be described with reference to FIG.

FIG. 16 is an external view of the stereo sound image enlargement system viewed from above. This stereo sound image enlargement system inputs a left input signal Lin and a right input signal Rin from a line input terminal, performs a crosstalk canceling process and a reverse phase signal mixing process on the input signal, and outputs a left output signal Lout and a right output signal from a line output terminal. Output as a signal Rout. The volume 72a is used to specify the amount of crosstalk cancellation. The signal from the volume 72a is
Used to determine the gain a. The volume 72b is used to specify the mixing amount of the reverse phase signal. The signal from the volume 72b is used to determine the gain e.

FIG. 17 is a block diagram showing a configuration of an electric circuit of the stereo sound image enlarging system. In the figure, A
The / D converter 60 outputs the left input signal Lin which is an analog signal.
And the right input signal Rin into digital data. This digital data is supplied to the DSP 61. DSP6
1 performs the above-described crosstalk cancel processing and the reverse-phase signal mixing processing according to the control data from the CPU 70. Then, the processed digital data is sent to the D / A converter 62. The D / A converter 62 converts the received digital data into analog data, and outputs a left output signal Lout
And a right output signal Rout.

The CPU 70 has a memory 71 and an input device 72.
And the display device 73 are connected. In the memory 71,
In addition to the control program for operating the CPU 70,
Various coefficients are stored. The CPU 70 has the memory 7
The filter coefficient and the like are read from 1 and sent to the DSP 61 as control data.

The input device 72 includes the above-mentioned volume 72a and volume 72b. C
The PU 70 stores the volume 72a and the volume 7
2b, and converts them into values indicating the gain a and the gain e.
Supply to SP61. The DSP 61 performs the above-described crosstalk cancellation processing and antiphase signal mixing processing based on the filter coefficients and the values indicating the gains a and e.

The display device 73 is, for example, an LED display, L
It can be constituted by a CD display or the like. By looking at the display 73, the user can know the state of the stereo sound image enlarging system and other information.

Although the stereo sound image enlarging device of the present invention can be used alone, it may be incorporated in another device to form one operation mode such as a multi-effector and a reverb. Other devices include, for example, a sound source module, an electronic piano, an electronic keyboard, a game machine, an audio device, and the like.

(Embodiment 2) Next, a second embodiment of the stereo sound image enlarging apparatus according to the second aspect of the present invention will be described. FIG. 18 is a block diagram showing a stereo sound image enlarging apparatus according to the second embodiment.

In FIG. 18, the multiplier 100 amplifies the left input signal Lin with a gain e. The output signal from the multiplier 100 is supplied to the delay unit 101 and the adder 102. The delay unit 101 converts the output signal from the multiplier 100 into m
Just delay. The output signal from the delay unit 101 is supplied to an adder 112. Similarly, the multiplier 110 amplifies the right input signal Rin with a gain e. This multiplier 110
Is supplied to the delay unit 111 and the adder 112. Delay device 111 delays the output signal from multiplier 110 by m. The output signal from the delay unit 111 is supplied to the adder 102.

The adder 102 corresponds to the Lch mixing means of the present invention. This adder 102 outputs the left input signal Lin
And the output signal from the multiplier 100 (the signal obtained by multiplying the left input signal Lin by the gain e), and subtracts the output signal from the delay unit 111 from the addition result. This realizes a function of mixing the left input signal Lin, a signal obtained by multiplying the left input signal Lin by the gain e, and a signal obtained by multiplying the right input signal Rin by the gain e, delaying the signal, and further making the signal out of phase. I have.

Similarly, the adder 112 corresponds to the Rch mixing means of the present invention. The adder 112 outputs the right input signal Rin and the output signal from the multiplier 110 (the right input signal Rin).
and a signal obtained by multiplying in by a gain e), and subtracting the output signal from the delay unit 101 from the addition result. This realizes a function of mixing the right input signal Rin, the signal obtained by multiplying the right input signal Rin by the gain e, and the signal obtained by multiplying the left input signal Lin by the gain e, delaying the signal, and further making the signal out of phase. I have.

The Lch sound quality correcting means of the present invention comprises a multiplier 1
03, a low-pass filter 104, a multiplier 105, and an adder 106. The output signal from the Lch mixing means (adder 102) is amplified by the multiplier 103 with a gain (1-e) and supplied to the adder 106. At the same time, the data is filtered by the low-pass filter 104 and supplied to the multiplier 105. The signal amplified by the gain e in the multiplier 105 is supplied to the adder 106. In the adder 106, the output signal from the multiplier 103 and the output signal from the multiplier 105 are added, and the addition result is output to the outside as a left output signal Lout.

Similarly, the Rch sound quality correcting means of the present invention
Multiplier 113, low-pass filter 114, multiplier 115
And an adder 116. The output signal from the Rch mixing means (adder 112)
, And is supplied to the adder 116. At the same time, it is filtered by the low-pass filter 114 and supplied to the multiplier 115. Multiplier 11
5, the signal amplified by the gain e is added to the adder 116.
Supplied to In the adder 116, the multiplier 113
And the output signal from the multiplier 115 are added, and the addition result is output to the outside as a right output signal Rout.

The above e is a gain for adjusting the magnitude of the stereo sound image enlargement effect. H to j are filter coefficients of the low-pass filters 104 and 114. Further, m is a delay amount corresponding to the time difference between the time of reaching the ear on the speaker side from the speaker and the time of reaching the ear on the opposite side. The delay amount m is represented by the number of sampling points when an analog signal is sampled at 48 KHz to generate a digital signal.

FIG. 20 shows the low-pass filters 104 and 11
4 shows the disturbance of the frequency characteristics of the monaural component due to the incorporation of the opposite phase signal when there is no 4. As shown in the figure, the monaural component forms a comb filter, and a reduction in the low frequency band is particularly noticeable. Therefore, as the stereo sound image enlargement effect is increased, (e
The sound quality seems to be lighter (as the value of becomes larger). The stereo sound image enlarging device includes a low-pass filter in the output stage, and suppresses the above phenomenon by increasing the degree of application of the low-pass filter and increasing the sound quality as the stereo sound image enlarging effect is increased.

The gain e is variable in the range of 0 to 1. When e = 0, the left input signal Lin becomes the left output signal Lout as it is, and the right input signal Rin becomes the right output signal Rout as it is. If e = 1, the low-pass filter 1
04 and 114 have the maximum effect.

FIG. 19 shows a modification of the second embodiment.
In this modification, a crosstalk canceling unit 130 is further provided, and the adder 102 includes a left signal from the crosstalk canceling unit 130 and an output signal from the multiplier 100 (a signal obtained by multiplying the left input signal Lin by the gain e). ), And subtracts the output signal from the delay unit 111 (the signal obtained by multiplying the right input signal Rin by the gain e and further delaying by m) from the addition result. Further, the adder 112 adds the right signal from the crosstalk canceling means 130 and the output signal from the multiplier 110 (a signal obtained by multiplying the right input signal Rin by the gain e), and determines a delay from the addition result. Vessel 10
The output signal from 1 (the signal obtained by multiplying the left input signal Lin by the gain e and further delaying by m) is subtracted.

As the crosstalk canceling means 130, those described in the first and second examples in the first embodiment described above can be used. Of course, a conventional Schroeder type crosstalk canceling device can also be used.

According to the stereo sound image enlarging device according to this modification, since the sound quality is corrected in accordance with the magnitude of the stereo sound image enlarging effect, the sound quality does not deteriorate even if the stereo sound image enlarging effect is increased. It is possible to clearly feel that the sound image is localized at a position distant from the listener, and it is possible to obtain a richer sense of spreading.

Further, the left input signal Lin and the right input signal Rin
Is a binaural signal to which a head-related transfer function is added, sound image control can be performed by two-channel speaker reproduction. Also in this case, the sound quality is reduced as the stereo sound image enlargement effect is increased. Therefore, if the output stage is provided with the Lch sound quality correction means and the Rch sound quality correction means according to the present invention so as to perform sound quality correction, it is possible to eliminate the disadvantage that the sound quality is reduced.

(Embodiment 3) Next, Embodiment 3 of the sound image control apparatus according to the third aspect of the present invention will be described.

The head-related transfer function (head-related impulse response) in the sound image control device according to the third embodiment is obtained by measurement. The measurement is performed not in an anechoic room, but in a room with some reflection as shown in FIG. This is because the localization of the sound image becomes clearer in a room with reflection. The measurement is performed by collecting an impulse sound from a sound source (speaker) using a dummy head.

The signals directly reaching the both ears of the dummy head from the speaker are direct sounds, and are indicated by solid lines in the figure. A signal from the speaker to both ears via a wall (including a floor and a ceiling) is a reflected sound, and is represented by a broken line in the figure. By performing the above measurement while changing the sound source position, data of the head related transfer function (head impulse response) in each direction can be obtained. Note that the measurement was performed using T
It may be performed by the SP (time stretch pulse) method.

FIG. 24 shows an example of the measured head impulse response. FIG. 3A shows the impulse response of the head of the left ear when the sound source is placed at an angle of 60 degrees to the left front, and FIG.
Is the impulse response of the head of the right ear when the sound source is placed at an angle of 60 degrees to the left. When this head impulse response is convoluted with the input sound and is heard with headphones, the listener can feel the position of the sound source at a position in the left diagonal direction of 60 degrees.

However, the number of points in the impulse response shown in FIG.
8 KHz), and large-scale hardware is required to perform the convolution processing. Therefore, the Lch and Rch head-related transfer function imparting means of the present invention are configured by a part that simulates a direct sound and a part that simulates a reflected sound as described below, thereby simplifying the hardware. I'm trying.

That is, as shown in FIG. 22, the Lch and Rch head-related transfer function assigning means of the present invention includes a part simulating a direct sound, a part simulating a reflected sound, and an adding unit for adding these parts. It is configured.

The part simulating the direct sound of the Lch head-related transfer function providing means includes a direct part filter 200 for filtering the monaural input signal Min and a direct part filter 2.
Binaural time difference delay device 2 for delaying the output signal from 00
01 and 01.

The portion that simulates the reflected sound includes a filter 202 for a reflector for filtering an output signal from the monaural input signal Min, and a plurality of delay units D ₁ for delaying an output signal from the filter 202 for the reflector. a delay unit 203 comprised in to D _n, the amplifying part 204 consisting of a plurality of multipliers G ₁ ~G _n for amplifying the output signals from the respective delay units D ₁ to D _n, a plurality of multipliers G ₁ ~G _n And a reflector adder 205 for adding the output signals from.

The adder includes a binaural time difference delay unit 201.
And an adder 206 for adding the output signal from the adder 205 to the output signal from the reflector adder 205. The output signal from the adder 206 is supplied to the crosstalk canceling means and the negative-phase signal generating means as the left input signal Lin.

Similarly, the portion of the Rch head-related transfer function providing means that simulates the direct sound is a direct filter 210 for filtering the monaural input signal Min and a binaural filter for delaying the output signal from the direct filter 210. And a time difference delay unit 211.

The portion that simulates the reflected sound includes a filter 212 for the reflector for filtering the output signal from the monaural input signal Min, and a plurality of delay units D ₁ for delaying the output signal from the filter 212 for the reflector. delay unit 213 made in to D _n, and the amplification unit 214 consisting of a plurality of multipliers G ₁ ~G _n for amplifying the output signals from the respective delay units D ₁ to D _n, a plurality of multipliers G ₁ ~G _n It is composed of a reflecting portion adder 215 for adding the output signal of the _mosquito et al.

The adder includes a binaural time difference delay unit 211.
And an adder 216 that adds the output signal from the adder 215 to the output signal from the reflector adder 215. The output signal from the adder 216 is supplied to the crosstalk canceling means and the negative phase signal generating means as the right input signal Rin.

The direct part filters 200 and 210
Simulates a frequency characteristic of a head-related transfer function of a sound that reaches a listener's ear directly from a sound source (for example, a speaker). The binaural time difference delay units 201 and 211 simulate the time difference between the two ears of the sound that directly reaches the listener's ear from the sound source. The reflection unit filters 202 and 212 simulate a frequency change due to room reflection. Delay unit 2 composed of delay units D _{1 to} D _n
03 and 213, the magnitude of the multiplier G ₁ ~G _n amplifier 204 and 214 and reflective portions adder 205 and 215 made of the sound at the time of arrival and arrival of the plurality of reflected sound binaural in a room To reproduce. And adders 206 and 2
16 is a signal corresponding to the direct sound from the binaural time difference delay units 201 and 211 and the reflection unit adders 205 and 215.
And a signal corresponding to the reflected sound from the input signal is added, and the addition result is output as a left input signal Lin and a right input signal Rin.

FIG. 25 shows an impulse response generated by the head related transfer function providing means of the present invention. Direct part II
This is equivalent to an R filter and a delay unit, and the reflection unit is equivalent to an IIR filter, a multiplier and a delay unit. FIG. 25 shows an example in which the direct portion filter is constituted by a 6th-order IIR type filter, the reflection portion filter is constituted by a 3rd-order IIR type filter, and the tap of the reflection portion is set to 8 taps. The Lch and Rch head transfer function providing means of the present invention is not limited to this.

As described above, since the head transfer function is added to the monaural input signal Min by the Lch head transfer function applying means and the Rch head transfer function applying means, it is possible to localize the sound image at an arbitrary position when listening to the headphones. I can do it.

Next, a method of moving a sound image in the sound image control device according to the third embodiment will be described.

As shown in FIG. 26, when the front is set to 0 degrees and the rear is set to 180 degrees, a case is considered where the sound image is moved using the head-related transfer function data prepared every 10 degrees in the horizontal plane. For example, in the head related transfer function providing means having the configuration shown in FIG.
When changing from data in the 0-degree direction to data in the 70-degree direction, if input signals are continuous, noise is generated due to waveform distortion. In order to suppress this, the movement of the sound image is switched so that the sound image generated based on the data in the 60-degree direction and the sound image generated based on the data in the 70-degree direction are cross-fade with each other. For this purpose, the head-related transfer function assigning means always performs sound image localization processing in two directions,
Based on the direction change information, a sound image in a different direction is replaced while cross-fading. The direction change information may be configured to be input by a user from the outside, or may be configured to be automatically performed by a processing device such as a CPU.

FIG. 27 is a block diagram of the present sound image control apparatus having a structure for smoothly moving sound images in adjacent directions by cross-fading. In FIG.
The first Lch head-related transfer function assigning means 300 is used to assign an Lch head-related transfer function corresponding to the first direction. First Rch head-related transfer function providing means 301
Is used to provide a head-related transfer function for Rch corresponding to the first direction. Similarly, the second Lch head-related transfer function assigning means 310 is used to assign a head-related transfer function for the left channel corresponding to the second direction. The second Rch head-related transfer function assigning means 311 is used to assign an Rch head-related transfer function corresponding to the second direction.

The first sequence gain adjustment circuit 302 corresponds to the first weighting means of the sound image control device according to the third aspect of the present invention. The first sequence gain adjustment circuit 302 multiplies each signal from the first Lch head-related transfer function providing means and the first Rch head-related transfer function providing means by a weight α in accordance with an instruction from the direction instructing means. . Here, 0 ≦ α ≦ 1. The first series gain adjustment circuit 302 is provided with a first Lc
It can be configured by a multiplier for multiplying each signal from the h-head transfer function providing means 300 and the first Rch head-related transfer function providing means 301 by the weight coefficient α.

Similarly, the second series gain adjustment circuit 312
Corresponds to a second weighting unit of the sound image control device according to the third aspect of the present invention. This second series gain adjustment circuit 3
12 is a first Lch according to an instruction from the direction instruction means.
Each signal from the head-related transfer function providing means and the first Rch head-related transfer function providing means is multiplied by a weight (1-α). The second sequence gain adjustment circuit 312 multiplies each signal from the second Lch head-related transfer function providing means 310 and the second Rch head-related transfer function providing means 311 by a weight coefficient (1−α). It can be composed of a container.

The adder 303 corresponds to the left mixing means of the sound image control device according to the third aspect of the present invention. This adder 3
03 is a signal from the first Lch head-related transfer function assigning means weighted by the first series gain adjusting circuit 302 and a signal from the second Lch head-related transfer function assigning means weighted by the second series gain adjusting circuit 312. And the left input signal Lin is generated.

Similarly, the adder 313 corresponds to the right mixing means of the sound image control device according to the third embodiment of the present invention. The adder 313 includes a signal from the first Rch head-related transfer function assigning unit weighted by the first series gain adjustment circuit 302 and a second Rch head-related transfer function weighted by the second series gain adjustment circuit 312. The right input signal Rin is generated by mixing the signal from the applying means.

In the above configuration, 60-degree direction data is supplied to the first Lch head-related transfer function providing means 300 and the first Rch head-related transfer function providing means 301, and the second Lch head-related transfer function providing means 310 and The data in the 70-degree direction is set in the second Rch head-related transfer function providing unit 311,
The sound image is moved by changing the weight coefficient α. When replacing the data with a new direction, the gain of the series is set to “0”, and the gain accompanying the data replacement can be eliminated by gradually increasing the gain over time. .

The direction and distance at which the sound image is localized may be other than the horizontal plane, and may be any direction and distance from up to down, left and right, short distance to long distance. If the head-related transfer functions in all directions and distances are measured and stored as data, the sound image can be localized in all directions and distances.

As described above, the head-related transfer function imparting means can smoothly move the sound image to any place in the three-dimensional space when listening to headphones.

Next, an example of a sound image control system using the above-described sound image control device will be described with reference to FIG. 28 and FIG. The electric circuit is substantially the same as the electric circuit (FIG. 17) of the stereo sound image enlarging apparatus according to the first embodiment of the present invention, except for the following points. That is, the input device 72 is different in that a volume 400c and a joystick 401 are added.

FIG. 28 is an external view of the sound image control system viewed from above. This sound image control system inputs a monaural input signal Min from a line input terminal, performs a process of adding a head-related transfer function thereto, a crosstalk cancellation process, and a reverse phase signal mixing process, and outputs a left output signal Lout from a line output terminal. And a right output signal Rout.

The volume 400a of the input device 72 is used to specify the amount of crosstalk cancellation.
The gain a is determined by the signal from the volume 400a. The volume 400b is used to specify the amount of mixing of the opposite-phase signals. This volume 40
The gain e is determined by the signal from 0b. The volume 400c is used to specify the vertical direction of the sound image. The joystick 401 of the input device 72 is used to determine the horizontal angle (0 to 360 degrees) of the sound image and the distance from the listener. This volume 400
The weight coefficient α is determined based on the signal from the signal c and the signal from the joystick 401.

FIG. 17 is a block diagram showing a configuration of an electric circuit of the sound image control system. DSP 61 is CPU 70
The stereo sound image according to the first aspect of the present invention is characterized in that, in addition to performing the crosstalk canceling process and the anti-phase signal mixing process in accordance with the control data from Different from magnifying device.

The CPU 70 reads out filter coefficients and the like from the memory 71 and sends them to the DSP 61 as control data. In addition, the CPU 70 controls the volumes 400a to 400
c and the signal from the joystick 401,
These are converted into data values indicating the weight coefficient α, the gain a and the gain e, and supplied to the DSP 61 as control data. The DSP 61 performs the process of assigning the head related transfer function, the crosstalk canceling process, and the anti-phase signal mixing process based on the weight coefficient α, the data indicating the gain a and the gain e, the filter coefficient, and the like.

Although the stereo sound image enlarging device of the present invention can be used alone, it may be incorporated in another device as a device formed in one operation mode such as a multi-effector and a reverb. Other devices include, for example, a sound source module, an electronic piano, an electronic keyboard, a game machine, an audio device, and the like.

[0136]

As described above in detail, according to the stereo sound image enlarging apparatus according to the first aspect of the present invention, the left and right input signals are subjected to the crosstalk canceling processing and the reverse phase signal mixing processing, respectively. By mixing and outputting the processing results, it is possible to form a sound field having an unprecedented richness.

By changing the mixing ratio, both the left and right input signals can be output as they are, only the crosstalk cancellation processing can be performed and output, only the reverse phase signal mixing processing can be performed and output, or Both crosstalk cancellation processing and antiphase signal mixing processing can be performed and output. Further, the degree of crosstalk cancellation and the degree of mixing of the opposite-phase signals can be adjusted.

Furthermore, the listening point can be set in a wide range by making the delay amount in the crosstalk canceling process and the delay amount in the reverse phase signal mixing process variable.

Further, according to the stereo sound image enlarging apparatus according to the second aspect of the present invention, the sound quality is degraded because the degree of application of the low-pass filter is increased as the mixing ratio of the negative-phase signal increases. The effect of enlarging the stereo sound image can be increased without the need.

Further, according to the sound image control device according to the third aspect of the present invention, in the two-channel speaker reproduction,
The input signal is subjected to a process of assigning a head-related transfer function, and the signal to which the head-related transfer function is assigned is further subjected to a crosstalk canceling process and a reverse phase process. Regardless, the sound image can be localized at any position in the three-dimensional space.

[Brief description of the drawings]

FIG. 1 is a diagram showing in principle a stereo sound image enlarging device according to a first embodiment of the present invention.

FIG. 2 is a diagram for explaining sound transmission in the stereo sound image enlarging device of the present invention.

FIG. 3 is a diagram showing a configuration of a crosstalk canceling means of the stereo sound image enlarging device of the present invention.

FIG. 4 is a diagram showing a configuration of an anti-phase signal generating means of the stereo sound image enlarging device of the present invention.

FIG. 5 is a diagram showing a configuration of a mixing unit of the stereo sound image enlarging device of the present invention.

FIG. 6 is a diagram for explaining transmission of sound according to a conventional Schrader method.

FIG. 7 is a block diagram illustrating a first example of a crosstalk canceling unit of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 8 is a diagram for explaining a first example of crosstalk canceling means of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 9 is a block diagram illustrating a configuration of a first example of a crosstalk canceling unit of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 10 is a diagram for explaining a second example of the crosstalk canceling means of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 11 is a diagram for explaining a second example of the crosstalk canceling means of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 12 is a block diagram showing a configuration of a second example of the crosstalk canceling means of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 13 is a block diagram illustrating a configuration of an anti-phase signal generation unit of the stereo sound image enlarging device according to the first embodiment of the present invention.

FIG. 14 is a block diagram showing another configuration of the anti-phase signal generating means of the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 15 is a block diagram illustrating a configuration of a mixing unit of the stereo sound image enlarging device according to the first embodiment of the present invention.

FIG. 16 is an external view of a stereo sound image enlarging system to which the stereo sound image enlarging device according to the first embodiment of the present invention is applied, as viewed from above.

FIG. 17 is a block diagram illustrating a configuration of an electric circuit of a stereo sound image enlarging system to which the stereo sound image enlarging device of Embodiment 1 is applied and a sound image control system to which the sound image control device of Embodiment 3 is applied.

FIG. 18 is a block diagram illustrating a configuration of a stereo sound image enlarging device according to a second embodiment of the present invention, in a second embodiment;

FIG. 19 is a block diagram showing a modified example of Embodiment 2 of the stereoscopic sound image enlarging device according to the second aspect of the present invention.

FIG. 20 is a diagram showing disturbance in frequency characteristics of monaural components due to mixing of opposite-phase signals in the stereo sound image enlarging apparatus according to the first embodiment of the present invention.

FIG. 21 is a diagram showing in principle a sound image control device according to a third embodiment of the present invention.

FIG. 22 illustrates an L of the sound image control device according to the third embodiment of the present invention.
It is a figure which shows the structure of ch head transfer function provision means and Rch head transfer function provision means.

FIG. 23 is a view for explaining measurement of a head-related transfer function (head-related impulse response) in the sound image control device according to the third embodiment of the present invention.

FIG. 24 is a diagram showing an example of a measured impulse response in the sound image control device according to the third embodiment of the present invention.

FIG. 25 is a diagram showing an example of an impulse response generated by the head-related transfer function providing means of the sound image control device according to the third embodiment of the present invention.

FIG. 26 is a diagram for explaining movement of a sound image in the sound image control device according to the third embodiment of the present invention.

FIG. 27 is a block diagram showing a configuration for moving a sound image in the sound image control device according to the third embodiment of the present invention.

FIG. 28 is an external view of a sound image control system to which the sound image control device according to the third embodiment of the present invention is applied, as viewed from above.

[Explanation of symbols]

 10, 20 Multiplier 11, 21 Delay unit 12, 22 Primary low-pass filter 13, 23 Subtractor 14, 24 Filter 15, 25 Filter 30, 40 Multiplier 31, 41 Delay unit 32, 42 Adder 32 ', 42' Addition Devices 50, 51 adder 60 A / D converter 61 DSP 62 D / A converter 70 CPU 71 memory 72 input device 72a, 72b volume 73 display device 100, 110 multiplier 101, 111 delay device 102, 112 adder 103 , 113 Multiplier 104, 114 Low-pass filter 105, 115 Multiplier 106, 116 Multiplier 130 Crosstalk canceling means 200, 210 Filter for direct unit 201, 211 Binaural time difference delay unit 202, 212 Filter for reflective unit 203, 213 Delay unit 204, 214 Multiplier 2 05, 215 Reflector adder 206, 216 Adder 300, 310 First Lch head transfer function assigning means 301, 301 First Rch head transfer function assigning means 310, 311 Second Lch head transfer function Giving means 303, 313 Second Rch head-related transfer function giving means 302 First series gain adjustment circuit 312 Second series gain adjustment circuit 303, 313 Adders 400a, 400b, 400c Volume 401 Joystick 402 Display

(Equation 10)

──────────────────────────────────────────────────続 き Continuation of the front page (51) Int.Cl. ⁷ Identification symbol FI H04S 7/00 H04S ^{7 /} 00F (56) References JP-A-8-146974 (JP, A) JP-A-2-92200 ( JP, A) JP-A-4-30700 (JP, A) JP-A-2-188098 (JP, A) JP-A-4-10499 (JP, U) (58) Fields investigated (Int. Cl. ⁷ , (DB name) H04S 1/00 H04S 5/00 H04S 7/00

Claims (5)

(57) [Claims] (1)Perform predetermined processing on left input signal and right input signal
And output them as a left output signal and a right output signal, respectively.
Stereo sound image magnifying device, The left input signal and a signal obtained by multiplying the left input signal by a gain e
And the right input signal is multiplied by a gain e, delayed, and
Left channel for mixing the phased signals
Mixing means; The right input signal and a signal obtained by multiplying the right input signal by a gain e
Multiply the left input signal by a gain e, delay it, and
Right channel for mixing phased signals
Mixing means, Gain is applied to the output signal from the left channel mixing means.
(1-e) and the left channel mixing
Apply a low-pass filter to the output signal of the
The multiplied signal is added, and the addition result is used as a left output signal.
A left channel sound quality correcting means for outputting, Gain is applied to the output signal from the right channel mixing means.
(1-e) and the right channel mixing
Apply a low-pass filter to the output signal of the
The multiplied signal is added, and the addition result is output as a right output signal.
Output right channel sound quality correction means,  And a stereo sound image enlarging device having: (2)Based on the left input signal and the right input signal
Left signal corresponding to the sound from which the crosstalk sound has been removed
Signal and right signal are output
Have more, The left channel mixing means includes the crosstalk key.
A left signal from the cancel means and a gain e
, And the right input signal is multiplied by a gain e.
And then mix the signals with opposite phases
And The right channel mixing means includes the crosstalk key.
A right signal from the cancel means and a gain e
, And the left input signal is multiplied by a gain e.
And then mix the signals with opposite phases.
, Characterized by The stereo sound image enlarging device according to claim 1.
Place. (3)The crosstalk canceling means, Generate a left crosstalk signal based on the left input signal
Means for generating a left crosstalk signal for Generate a right crosstalk signal based on the right input signal
Means for generating a right crosstalk signal for To subtract the right crosstalk signal from the left input signal
And the left channel cross calculator Filter the signal from the left channel cross computing unit.
Channel channel for generating the first left signal
A loss filter, To subtract the left crosstalk signal from the right input signal
And the right channel cross calculator Filter the signal from the right channel cross computing unit
Right channel clock to generate a first right signal
Loss filter, 3. The method according to claim 2, wherein Stereo sound
Image magnifier. (4)Head for left channel on monaural input signal
A left channel to which a transfer function is assigned and output as a left input signal
Flannel head transfer function providing means, A head-related transfer function for the right channel is added to the monaural input signal.
Right channel head signal that is added and output as a right input signal.
Delivery function providing means, The method according to claim 3, further comprising: Stele
A sound image magnifier. (5)Direction indication to indicate the direction to move the sound image
Further comprising means, The left channel head-related transfer function providing means and the right channel
The channel head related transfer function providing means is: The head-related transfer function for the left channel corresponding to the first direction is
First left channel head related transfer function providing means for providing; Head-related transfer function for the right channel corresponding to the first direction
Right channel head related transfer function providing means for providing
When, Adds a head-related transfer function for the left channel corresponding to the second direction.
Second left channel head related transfer function providing means for providing; Head-related transfer function for the right channel corresponding to the second direction
Right channel head related transfer function assigning means for assigning
When, The first left channel according to an instruction from the direction instruction means.
Head transfer function providing means and the first right channel head
First to apply weight α to each signal from the transfer function assigning means
Weighting means, The second left channel according to an instruction from the direction instruction means.
Head transfer function providing means and the second right channel head
Weighting each signal from the transfer function assigning means (1-α)
Second weighting means for applying; The first weighted α is given by the first weighting means.
From the left channel head-related transfer function providing means and the
Weighting (1-α) is performed by weighting means 2
A signal from the second left-channel head-related transfer function providing means;
And a left mixing means for generating a left input signal. The first weighted α is given by the first weighting means.
From the right channel head-related transfer function applying means of
Weighting (1-α) is performed by weighting means 2
A signal from the second right-channel head-related transfer function providing means;
And a right mixing means for generating a right input signal, The stereo sound according to claim 4, characterized by:
Image magnifier. However, 0 ≦ α ≦ 1.