JP3233275B2 - Sound image localization processing method and apparatus - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and apparatus for localizing a sound image used in a home A / V (audio / visual) system, a surround audio reproducing apparatus and the like.

[0002]

2. Description of the Related Art A home-use television set (TV set) capable of reproducing audio in a stereo system is provided with a pair of left and right speakers. In some cases, it may be difficult to install a surround speaker when playing back audio in a surround system such as movie software.

In such a case, the audio signal is
After performing sound image localization processing using a head related transfer function, etc., the sound image (virtual speaker) is localized at a position other than the actual speaker by supplying it to each speaker, and the distance between the left and right speakers is wider than the actual distance. Surround sound can be heard from the left and right or left and right rear of the listener using only the two front speakers.

[0004] By the way, although the sound image varies depending on the listener, it is easy to localize the sound image at the target position when the sound image is moved. However, when the sound image is stopped, particularly when the sound image is stopped at a position extending from the side to the rear of the listener, there is a problem that the sound image is likely to be flipped back and forth, and it is difficult to localize the sound image at the target position.

In order to solve the above-mentioned problem, in order to make the listener recognize a sound image at a target position which is separated from the front of the listener position by an angle θ in one direction in the circumferential direction, the sound position is determined to be near the target position and A first processing signal for localizing the sound image at a first localization position separated by an angle θ1 (where θ1 <θ) in the circumferential direction from the front of the position, and a first processing signal that is located near the target position and is in the circumferential direction from the front of the listener position. A second processing signal for localizing a sound image at a second localization position separated by an angle θ2 (where θ2> θ) is generated, and each processing signal is alternately and regularly supplied to a speaker, so that the sound image and the first localization position are separated. A method has been proposed in which a listener is allowed to recognize a sound image at a target position by alternately and regularly localizing the sound at a second localization position.

[0006]

However, each of the processed signals is alternately and regularly supplied to the loudspeaker, and the sound image is converted to the first image.
When the listener is localized at the localization position and the second localization position alternately and regularly, the listener feels a regular pattern, which may give an extremely unnatural impression on hearing.

It is an object of the present invention to provide a sound image localization processing method and an apparatus therefor which can solve the above-mentioned problem.

[0008]

In order to achieve the above object, the sound image localization processing method of the present invention is characterized in that:
A pair of left and right speakers is arranged in front of the listener position, and a sound signal localization process is performed by applying a sound image localization process to the audio signal to each of the speakers, thereby localizing the sound image in an arbitrary direction of the listener position. In the processing method, in order to recognize a sound image at a target position separated from the front of the listener position by an angle θ in one direction with respect to the circumferential direction,
A first processing signal for localizing the sound image at a first localization position that is located near the target position and is separated from the front of the listener position by an angle θ1 (here, θ1 <θ) in the circumferential direction, and a vicinity of the target position; And generating a second processing signal for localizing the sound image at a second localization position separated from the front of the listener position by an angle θ2 (here, θ2> θ) in the circumferential direction. One of the processing signals includes 0 or more and 1 Multiply by a coefficient k that varies randomly between:
k), the two processed signals multiplied by each coefficient are added, and the sum is supplied to each speaker. Note that the spectrum of the coefficient k may have a characteristic of approximately 1 / f.
Further, a random signal generator outputs a rectangular pulse-shaped random signal having a pulse height of 1 and a random pulse width and pulse pitch, and integrates the signal by an integrating circuit to obtain a coefficient k. Sometimes. Further, the audio signal may be squared by a squaring circuit as a coefficient generation signal and then passed through a low-pass filter to obtain a coefficient k. Also, the audio signal may be a two-channel stereo signal, and a signal of one channel, a sum signal of both channels, and a difference signal of both channels may be used as a coefficient generation signal. Further, a feature of the sound image localization processing device of the present invention is that a pair of speakers is disposed in front of the listener position, and a processing signal obtained by performing sound image localization processing on an audio signal is supplied to each of the speakers. so,
In order to localize the sound image in an arbitrary direction of the listener position, in order to recognize the sound image at a target position separated from the front of the listener position by one angle in the circumferential direction in the circumferential direction, the sound image is determined to be near the target position and the listener position A first angle θ1 (where θ1 <θ) separated from the front in the circumferential direction.
Means for generating a first processing signal for localizing the sound image at the localization position; and a sound image at a second localization position which is close to the target position and is separated from the front of the listener position by an angle θ2 (where θ2> θ) in the circumferential direction. Means for generating a second processing signal for localizing the signal; means for generating a coefficient k which randomly changes between 0 and 1; and a coefficient multiplier for multiplying one of the processing signals by the coefficient k. It has a coefficient multiplier for multiplying a processed signal by a coefficient (1-k), and a means for adding two processed signals multiplied by each coefficient and supplying the sum to each speaker.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to FIGS. 1 to 9. FIG. 2 shows an arrangement of speakers. In FIG. A speaker is shown and is arranged in front of a listener (listener position) M.

FIG. 1 shows a sound image localization processing device (virtual speaker processing device). Reference numerals 1 and 2 denote first and second input terminals to which an audio signal is input, and reference numerals 3 and 4 denote first and second output terminals.
The first output terminal 3 is connected to the left speaker SPL, and the second output terminal 4 is connected to the right speaker SPR. Although a two-channel stereo signal is input as an audio signal in the illustrated example, the audio signal may be a monaural signal.

As shown in the left half of FIG. 9, the sound image localization processing device has a target position P separated from the front face F of the listener M by an angle θ in one circumferential direction (counterclockwise direction in the example). When the sound image is to be recognized by the listener M, the sound image (left virtual speaker) is placed at the first localization position P1 which is near the target position P and is separated from the front face F of the listener M by the angle θ1 (where θ1 <θ) in the circumferential direction. ), And a sound image (right virtual speaker) is located at a second localization position P2 which is located near the target position P and is separated from the front face F of the listener M by the angle θ2 (here, θ2> θ) in the circumferential direction. Things. As shown in the right half of FIG. 9, when the listener M recognizes the sound image at the target position P separated from the front face F of the listener M position by an angle −θ in the clockwise direction, as shown in the right half of FIG. A sound image (right virtual speaker) is localized at a first localization position P1 separated from the front face F of the listener M by an angle -θ1 in the above direction, and the listener M
The sound image (the left virtual speaker) is localized at the second localization position P2 separated from the front F by an angle -θ2 in the above direction. still,
The difference between the angles θ and θ1 and the difference between the angles θ and θ2 are (substantially)
There are cases where they are the same and cases where they are not the same. The angle difference is, for example, about 40 ° or less.

The sound image localization processing device includes a first signal processing means (circuit) (first virtual speaker processing means (circuit)) S1
And a second signal processing means (circuit) (second virtual speaker processing means (circuit)) S2. The respective means S1 and S2 are connected to both input terminals 1 and 2. First signal processing means S1
Is used to localize a sound image at the first localization position P1, performs sound image localization processing on an audio signal in the same manner as in the related art, and outputs a first L processed signal for the left speaker SPL and a first R processed signal for the right speaker SPR. The second signal processing means S2 includes:
The sound image is localized at the second localization position P2. The audio signal is subjected to the sound image localization processing as in the related art, and the left speaker S
The second L processing signal for the PL and the second L signal for the right speaker SPR
An R processing signal is output.

The first and second signal processing means S1,
As S2, for example, a lattice type (localization processing) filter, a shuffler type (localization processing) filter, or the like is used. That is, the sound image localization processing device includes, for example, a pair (two sets) of a lattice type filter or a shuffler type filter. A method of using these filters to expand the stereo sound field only with the two front speakers SPL and SPR, or to localize the sound image to the side or behind the listener M to give a sense of surround has already been proposed. Have been.

The lattice filter is connected to the first input terminal 1 as shown in FIG.
A first L filter unit (signal processing unit) F1L for outputting an output signal for PL, and a first input terminal 1
A first R filter unit (signal processing unit) F1R that outputs an output signal for the right speaker SPR; a second input terminal 2
A second L filter unit (signal processing unit) F2L that is connected to the second speaker and outputs an output signal for the left speaker SPL;
A second R filter unit (signal processing unit) connected to the second input terminal 2 and outputting an output signal for the right speaker SPR
F2R and first and second L filter units F1L and F2
An adder M8 that adds the two output signals of L and forms a first L-processed signal or a second L-processed signal;
By adding both output signals of the filter units F1R and F2R,
Adder M as first R processed signal or second R processed signal
9 The 1L · second 1R · second 2L · second 2R filter unit f1L, F1R, F2L, each transfer function (transfer characteristic) of F2R are each a _{H 11, H 12, H 21} , H 22.

By the way, as shown in FIG.
When the sound images ZL and ZR are localized to the left, right, diagonally backward, etc., the head-related transfer functions from the left speaker SPL to the left and right ears of the listener M are _hLL , _hLR , and the right speaker SPL is the left and right of the listener M. H _RL and h _RR are the head-related transfer functions to each ear, and h _{L′ L} and h _L are the head-related transfer functions from the loudspeaker to the left and right ears of the listener M when a (virtual) loudspeaker is arranged in the left sound image ZL. _'R , the listener M from the speakers SPL, SPR when the (virtual) speakers SPL, SPR are arranged in the right sound image ZR.
The head-related transfer functions to each of the left and right ears are h _R'L and h _R'R ,
Next, these head-related transfer functions _hLL , _hLR , _hRL , _hRR , h
_{L'L, h L'R, h} R'L, from h _R'R, each filter portion F1L lattice type filter, F1R, F2L, F2R
, Transfer functions H ₁₁ , H ₁₂ , H ₂₁ and H ₂₂ are obtained.

First, a matrix of the head-related transfer functions from the speakers SPL and SPR to the ears of the listener M is [h], and the (virtual) speakers are arranged in the sound images ZL and ZR. The matrix of the head-related transfer function up to [h '] and the matrix of the transfer function of the lattice filter are [H], and each matrix is displayed as follows.

[0017]

(Equation 1)

The following equations hold from FIG. 2 and FIG. [H '] = [h] [H] (4) Here, when | h | ≠ 0, the inverse matrix of [h] [h]
^{There is -1} and the above equation can be transformed into the following equation. [H] = [h] [h] ^-1 (5) From the above equation, the following equation is established, and the transfer functions H ₁₁ , H ₁₂ , H of the filter units F1L, F1R, F2L, F2R of the lattice type filter are obtained. it can be determined _21, H _{22. H 11 = (h RR h L'} L -h RL h L'R) / (h LL h RR -h LR h RL) (6) H 12 = (h LL h L'R -h LR h L'L _{) / (h LL h RR -h} LR h RL) (7) H 21 = (h RR h R'L -h RL h R'R) / (h LL h RR -h LR h RL) (8) H _{22 = (h LL h R'R -h} LR h R'L) / (h LL h RR -h LR h RL) (9)

As shown in FIG. 4, the shuffler type filter has a first filter section (signal processing section) F1.
A second filter unit (signal processing unit) F2;
An adder M that adds the audio signals input to the first and second input terminals 1 and 2 and inputs the audio signal to the first filter unit F1
1 and an audio signal input to the second input terminal 2 from the audio signal input to the first input terminal 1 and a subtracter M2 input to the second filter unit F2.
And an adder M10 that adds the two output signals of the first and second filter units F1 and F2 and outputs the added signal as a first L-processed signal or a second L-processed signal, and an output signal of the first filter unit F1. There is a subtractor M11 that subtracts the output signal of the second filter unit F2 and outputs the result as a first R processed signal or a second R processed signal.

The shuffler type filter is used when the left and right speakers SPL and SPR and the left and right sound images ZL and ZR are symmetric with respect to the front face F of the listener M. Head-related transfer function h _LL , h
_{LR, h RL, h RR,} h L'L, h L'R, h R'L, the transfer function H _SUM of the first and second filter portions from h _R'R, seek H _DIF.

First, since the left and right speakers SPL, SPR and the left and right sound images ZL, ZR are symmetrical with respect to the front face F of the listener M, h _LL = h in the equations (6) to (9).
_{RR, h LR = h RL,} h L'L = h R'R, h L'R = h R'L are satisfied, therefore, the _{H 11 = H 22, H 12} = H 21.

Next, h _LL = h _RR = h _a , h _LR = h _RL = h
_{b, h L'L = h R'R} = h a ', h L'R = h R'L = h b' putting the transfer function H _SUM, H _DIF are represented by the following formula. _{H SUM = (h a '+} h b') / (h a + h b) H DIF = (h a '-h b') / (h a -h b)

In FIG. 1, K1L and K1R are the first L.
A first R coefficient multiplier, which outputs the first signal from the first signal processing means S1;
The L · first R processed signal is multiplied by a coefficient k (0 ≦ k ≦ 1) that randomly changes between 0 and 1 inclusive. K2
L and K2R are second L / second R coefficient multipliers which add a coefficient (1) to the second L / second R processed signal from the second signal processing means S2.
-K).

The coefficient k may be of any type, but if the spectrum of the coefficient k has a (substantially) 1 / f characteristic which is said to be physiologically natural, unnaturalness can be eliminated, preferable. There are the following methods for generating the coefficient k having the 1 / f characteristic.

(1) As shown in FIGS. 5 and 6, from a random signal generator (for example, a digital signal processor (DSP) or the like) PR, as a random signal, for example, as shown in FIG. An M-sequence signal is output. The signal is a rectangular pulse, the pulse height is 1, the pulse width and the pulse pitch are random. The M-sequence signal is multiplied by the coefficient a ₀ in the scaling unit SC1 to reduce the possibility that the output value at the next stage will exceed 1, and then, as shown in FIG. Is integrated. An integrating circuit SK for multiplying the input signal by a reciprocal of an appropriate impedance value;
A coefficient multiplier K for multiplying the output signal of the circuit J by a coefficient b ₁
4 and an adder (mixer) M4 for adding the output signal of the multiplier K4 to the input signal to the integration circuit SK. The output signal from the integration circuit SK is passed through an overflow limiter L whose maximum limit value is 1, and is set as a coefficient k. Note that the scaling unit SC and the overflow limiter L may be omitted as appropriate.

(2) It is said that a power spectrum of a music signal originally has a 1 / f characteristic in many cases. Therefore, as shown in FIG. 7 and FIG.
As a channel stereo signal, a signal of one channel, a sum signal of both channels, and a difference signal of both channels are selected as a signal for coefficient generation by a signal selection or addition / subtraction circuit SE. Then, the selected signal (FIG. 8)
(A) is squared by the square circuit SQ (see FIG. 8).
(See (B)), the scaling section SC2 multiplies by an appropriate coefficient to reduce the possibility that the output value at the next stage will exceed 1. Next, the output signal from the scaling unit SC2 is passed through a low-pass filter LPF having a cut-off frequency of, for example, about 10 Hz (see FIG. 8C) to obtain a coefficient k.

In FIG. 1, M6 and M7 are adders (mixers), and the adder M6 is a first multiplied coefficient.
The L / second L processed signal is added and supplied to the left speaker SPL. The adder M7 outputs the first R multiplied by the coefficient.
-Add the second R processed signal and supply it to the right speaker SPR.

According to the above configuration example, as shown in the left half of FIG. 9, for example, at the target position P which is separated from the front face F of the listener M by an angle θ (eg, + 120 °) in the counterclockwise direction, the sound image is obtained. When the listener M recognizes the first localization, the first localization is determined by the first signal processing means S1 to be in the vicinity of the target position P and separated from the front face F of the listener M by an angle θ1 (for example, 90 °) in the counterclockwise direction. A first L / R processing signal for localizing the sound image at the position P1 is generated, and the angle θ2 (for example, 150) is set by the second signal processing means S2 in the vicinity of the target position P and from the front F of the listener M with respect to the circumferential direction. °) A second L / second R processing signal for localizing the sound image at the separated second localization position P2 is generated.

Next, the first L / R first processed signal is multiplied by a coefficient k which randomly changes between 0 and 1 and the coefficient (1-k) is multiplied by the second L / 2R processed signal. Multiply. The multiplied first L processing signal and second L processing signal are added by an adder M6 and output to the left speaker SPL, and the multiplied first R processing signal and second R processing signal are also added by an adder M7. Add
Output to the right speaker SPR.

Thus, the left and right speakers SPL, SP
A first L / second L processed signal or a first R / second R processed signal added at a random ratio is input to R, and sound is output from the left and right speakers SPL and SPR, and the sound image is placed at the first localization position. P1 and the second localization position P2, and the volume from the first localization position P1 and the second localization position P2
The volume from changes randomly.

Thus, even when the sound image is stopped at a position extending from the side to the rear of the listener M, the listener M can be surely recognized that the sound image is at the target position P, and the first localization position P1 Since the sound volume from the sound source and the sound volume from the second localization position P2 randomly change, there is no fear that the listener M will give an unnatural impression to the listener.

In particular, (approximately) 1 / f
With this characteristic, the change in volume from the first and second localization positions P1 and P2 becomes physiologically natural, and the listener M can be given a more natural impression in terms of hearing.

[0033]

As described in detail above, according to the present invention,
Even when the sound image is stopped at a location extending from the side to the rear of the listener, the listener can be surely recognized that the sound image is at the target position, and the listener may have an unnatural impression in terms of hearing. Nor.

If the spectrum of the coefficient k has a characteristic of approximately 1 / f as in claim 2, the listener can be given a more natural audible impression.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an example of an embodiment of the present invention.

FIG. 2 is an explanatory diagram illustrating an example of an embodiment of the present invention.

FIG. 3 is a block diagram showing an example of first and second signal processing means of FIG. 1;

FIG. 4 is a block diagram showing an example of first and second signal processing means of FIG. 1;

FIG. 5 is a block diagram illustrating an example of a method for generating a coefficient k in FIG. 1;

FIG. 6 is a waveform chart showing signals of respective units in FIG. 5;

FIG. 7 is a block diagram illustrating an example of a method for generating a coefficient k in FIG. 1;

FIG. 8 is a waveform chart showing signals of respective parts in FIG. 7;

FIG. 9 is an explanatory diagram illustrating an example of an embodiment of the present invention.

[Explanation of symbols]

 K1L, K1R First L / First R coefficient multiplier K2L, K2R Second L / second R coefficient multiplier L Overflow limiter LPF Low pass filter M Listener M6, M7 Adder P1, P2 First and second localization position PR random signal generator S1, S2 First. Second signal processing means SC1, SC2 Scaling unit SK Integrator SPL, SPR Speaker

──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Kazuyoshi Takemura 2-1 Nissincho, Neyagawa City, Osaka Prefecture Inside (72) Inventor Tetsuro Nakatake 2-1 Nissincho, Neyagawa City, Osaka Onkyo Yo (56) References JP-A-4-30700 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04S 1/00 H04S 5/02 H04S 7/00

Claims (6)

(57) [Claims] 1. A pair of left and right speakers is disposed in front of a listener position, and a processing signal obtained by subjecting an audio signal to sound image localization processing is supplied to each of the speakers, so that a sound image is moved in an arbitrary direction at the listener position. In the sound image localization processing method for localizing an object at an angle θ in one direction in the circumferential direction from the front of the listener position,
In order to recognize the sound image at the separated target position, the sound image is localized at a first localization position that is close to the target position and is separated from the front of the listener position by an angle θ1 (here, θ1 <θ) in the circumferential direction. A first processing signal to be located near the target position and a second processing signal to localize the sound image at a second localization position at an angle θ2 (where θ2> θ) in the circumferential direction from the front of the listener position. Generated, multiplied by one coefficient k that randomly changes between 0 and 1 and multiplied by the coefficient (1-k), and multiplied by each coefficient. A sound image localization processing method characterized in that two processing signals are added and supplied to each speaker. 2. The sound image localization processing method according to claim 1, wherein the spectrum of the coefficient k has a characteristic of approximately 1 / f. 3. A random signal generator outputs a rectangular pulse-shaped random signal having a pulse height of 1 and a random pulse width and pulse pitch, and integrates the signal by an integrating circuit to obtain a coefficient. 2. The sound image localization processing method according to claim 1, wherein k is k. 4. The sound image localization processing method according to claim 1, wherein the audio signal is squared by a squaring circuit as a coefficient generation signal and then passed through a low-pass filter to obtain a coefficient k. 5. The sound image localization processing method according to claim 4, wherein the audio signal is a two-channel stereo signal, and a signal of one channel, a sum signal of both channels, and a difference signal of both channels are signals for coefficient generation. 6. A pair of speakers are disposed in front of the listener position, and a processing signal obtained by subjecting an audio signal to a sound image localization process is supplied to each of the speakers, so that a sound image is shifted in an arbitrary direction of the listener position. At the target position, which is separated from the front of the listener position by an angle θ in one direction in the circumferential direction with respect to the circumferential direction, in order to recognize a sound image, the target position is set near the target position, and the angle θ1 ( However, a means for generating a first processing signal for localizing the sound image at the first localization position separated by θ1 <θ), and an angle θ2 (where θ2>) in the circumferential direction from the front of the listener position and near the target position θ) The separated second
Means for generating a second processing signal for localizing the sound image at the localization position; and a coefficient k randomly varying between 0 and 1 inclusive
, A coefficient multiplier for multiplying one processing signal by a coefficient k, a coefficient multiplier for multiplying the other processing signal by a coefficient (1-k), and two processing signals multiplied by each coefficient. And a means for adding the obtained signals to each speaker.