JPH07288900A - Sound field playback device - Google Patents

Abstract

(57) [Abstract] [Purpose] To realize a sound field reproducing device capable of smoothly moving a sound image position in real time. An input means for inputting an audio signal, a sound image position input means for inputting a sound image position, and a signal processing means for performing sound image localization are provided. When moving the sound image position, the signal processing means performs an operation so that the tap coefficient gradually changes based on the tap coefficient of the current sound image position and the tap coefficient of the new sound image position stored in advance. In this way, the sound image position of the surround sound as radiated from the rear virtual speaker 14 can be smoothly changed only by the output sound of the front left and right speakers 6a and 6b.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound field reproducing apparatus for reproducing a realistic sound in real time in AV (audio / visual) equipment.

[0002]

2. Description of the Related Art In recent years, in the field of video and audio, VT
Due to the spread of R, a large screen and realistic sound reproduction are desired in order to enjoy movies and the like at home, and the development of hardware corresponding thereto is desired. As an example of such an application, a sound field reproducing device has been developed that allows sound to be continuously heard from the side or the back without actually placing the speaker on the side or the back of the viewer.

6 and 7 are block diagrams showing the configuration of a conventional sound field reproducing apparatus. First, a method of localizing sound to the left rear of the viewer 12 by using the speakers 6a and 6b placed in front of the viewer (also referred to as listener) 12 in FIG. 7 will be described.

Of the signals from the signal input means 1 shown in FIG.
(Left) ch acoustic signal is output to the A / D converter 2a,
The acoustic signal of R (right) ch is output to the A / D converter 2b. The A / D converter 2 is a circuit that converts an analog acoustic signal into a digital signal for digital processing. The outputs of the A / D converters 2a and 2b are given to the signal processing means 3a and 3b, respectively. The Lch signal processing circuit 3a is a circuit having a transfer function hL (t) for controlling the sound field, and the Rch signal processing circuit 3b is a circuit having a transfer function hR (t) for controlling the sound field. . The signal processing circuit 3 is composed of, for example, an FIR filter, and its tap coefficient is given by a coefficient register 9 described later.

D / A converters 4a and 4b in FIG. 7 are circuits for inputting the digital signals of the signal processing circuits 3a and 3b and converting them into analog acoustic signals. The amplifiers 5a and 5b are circuits that input the signals of the D / A converters 4a and 4b, respectively, amplify the power, and output the amplified signals to the speakers 6a and 6b. The speakers 6a and 6b are speakers provided in the left and right front of the AV device or the listener 13.

As shown in FIG. 6, the sound field reproducing device is provided with a sound image position input means 11. The sound image position input means 11 is for inputting the source of the AV equipment or the position of the sound source of the acoustic signal included in the transmitted AV signal. When reproducing surround sound, the sound sources are rear left and right rear of the listener 12, according to the video displayed on the AV device.
And it may move to the back. At this time, the sound image position input means 11 outputs a signal to the tap coefficient control means 7,
The necessary tap coefficient is output from the coefficient storage means 10 to the coefficient register 9.

The coefficient storage means 10a is a circuit for storing a plurality of sets of tap coefficients of the signal processing means 3a according to the positions of all sound sources that may be set. When the tap coefficient is designated by the tap coefficient control means 7, the coefficient storage means 1
It is read from 0a and temporarily stored in the coefficient register 9a. Similarly, the coefficient storage means 10b is a circuit for storing a plurality of sets of tap coefficients of the signal processing means 3b according to the positions of all the sound sources that may be set. When the tap coefficient is designated by the tap coefficient control means 7, the coefficient storage means 10
It is read from b and temporarily stored in the coefficient register 9b. The tap coefficients output from the coefficient registers 9a and 9b are given to the signal processing circuits 3a and 3b, respectively, and the transfer function h
The values of L (t) and hR (t) are adaptively controlled by convolution.

The operation of the conventional sound field reproducing apparatus thus configured will be described using mathematical expressions. In FIG.
h1 (t) is the impulse response from the speaker 6a to the left ear of the listener 12. To be precise, it is the response to the eardrum of the right ear of the listener 12 when an impulse is input to the speaker 6a, but the measurement is performed at the position of the ear canal entrance. Hereinafter, this function will be referred to as a head-related transfer function, and will be referred to as an impulse response when described in the time domain. It should be noted that the same idea holds when the image is captured in the frequency domain.

Next, h2 (t) is the speaker 6a and the listener 12
Impulse response at the position of the right ear of the speaker, h3 (t) is the impulse response at the position of the speaker 6b and the left ear of the listener 12, and h4 (t) is the impulse response at the position of the speaker 6b and the right ear of the listener 12. Is. The signal generating means 13 is a circuit for generating an impulse signal S (t), and the speaker 14 is, for example, a virtual speaker arranged on the left rear side of the listener 12. H5 (t) is the impulse response at the position of the speaker 14 and the left ear of the listener 12, and h6 (t) is
It is an impulse response at the position of the speaker 14 and the right ear of the listener 12.

When the signal S (t) of the signal generating means 13 is radiated from the virtual speaker 14, the sound reaching the ear of the listener 12 is L (t) = S (t) * h5 (t in the left ear. ) (1) In the right ear, it is expressed as R (t) = S (t) * h6 (t) ... (2) (however, * represents a convolution operation).

Actually, the values of the equations (1) and (2) are multiplied by the transfer function of the speaker itself, but these are ignored. Further, it may be considered that the transfer function of the speaker or the like is included in h5 (t) and h6 (t).

Further, the impulse response and the signal S (t) are
Considered as digital signals that are discrete in time, they are expressed as follows. L (t) → L (n) R (t) → R (n) h5 (t) → h5 (n) h6 (t) → h6 (n) S (t) → S (n) where n is a natural number And actually means nT. Further, T represents the sampling time, but in general, T
Is omitted.

Equations (1) and (2) are represented by the following equation (3),
It becomes (4).

[Equation 1]

[Equation 2] Here, N is the length of the impulse responses h5 (n) and h6 (n).

Similarly, the signal S (t) is sent to the speaker 6a,
The sound radiated from 6b and reaching the listener 12 is In the right ear Is expressed as

Considering equations (5) and (6) as a temporally discrete digital signal, Becomes

If the head-related transfer functions are equal, it is assumed that the sound is heard from the same direction (this assumption is generally correct), and L (n) = L '(n) (9) ), And h5 (n) = hL (n) * h1 (n) + hR (n) * h3 (n) (10) holds.

Similarly, R (n) = R '(n) ... (11), and h6 (n) = hL (n) * h2 (n) + hR (n) * h4 (n) ... ( 12) is established.

In order to allow the listener 12 to hear the sound from the rear left, which is the position of the virtual speaker 14, using only the speakers 6a and 6b, the expressions (10) and (12) are satisfied. Thus, the values of hL (n) and hR (n) may be determined.

For example, if the equations (10) and (12) are rewritten in the frequency domain representation, the convolution operation becomes a transfer function obtained by performing a fast Fourier transform (FFT) on the impulse response instead of multiplication. Since the transfer function of the FIR filter is obtained by measurement, the transfer function of the FIR filter is expressed by these two formulas (1).
It can be obtained from 0) and (12).

HL (n), hR (n) thus determined
And the signals S (n) and hL (n) from the speaker 6a,
By radiating the convoluted signal S (n) and hR (n) from the speaker 6b, the listener 12 hears a sound from the rear even if the virtual speaker 14 in the rear is not actually sounded. You will feel that you are

In this way, the tap coefficient corresponding to the position input by the sound image position input means 11 is stored in the coefficient storage means 10a,
It is extracted from 10b and set in the coefficient registers 9a and 9b. Then, convolution processing is performed by the input signal and the signal processing means 3a and 3b. This allows the sound image to be localized at an arbitrary position with respect to the listener 12.

[0022]

However, in the above configuration, when it is attempted to continuously change the position of the sound image, the tap coefficient must be changed, and the time waveform of the acoustic signal is changed when the tap coefficient is changed. There is a problem that the noise becomes discontinuous and switching noise occurs.

The present invention has been made in view of the above conventional problems, and provides a sound field reproducing apparatus capable of obtaining continuous sound image movement without generating switching noise of acoustic signals. The purpose is to do.

[0024]

DISCLOSURE OF THE INVENTION The present invention comprises a signal input means for outputting a stereo acoustic signal, and a plurality of A / D converters for converting the analog signals of the respective channels input from the signal input means into digital signals. , A plurality of signal processing means configured by a digital filter in which a plurality of tap coefficients are set, for performing a convolution operation by inputting the signal of the A / D converter, a sound image position input means for inputting a sound image position, and a sound image position Coefficient storage means for storing a plurality of sets of tap coefficients of the signal processing means corresponding to all sound image positions that may be instructed by the input means, and when the sound image position is switched by the sound image position input means, move to the current position Tap coefficient calculation means for respectively reading the tap coefficient of the previous sound image position from the coefficient storage means, and calculating the tap coefficient of the sound image position that gradually transitions from the current position to the destination. A plurality of coefficient registers for temporarily holding the tap coefficients calculated by the up coefficient calculation means and setting them in the signal processing means, and a plurality of D / A conversions for converting the output signal of each channel from the signal processing means into an analog signal. A vessel,
A sound image instructed by the sound image position input means, comprising a plurality of amplifiers for power-amplifying the output of the D / A converter, and a speaker for each channel to which the signals of the amplifiers are input and which reproduces sound including surround sound. The position tap coefficient is read from the coefficient storage means and set in the signal processing means to output the surround sound from the speaker, and when the sound image position is switched by the sound image position input means, the tap coefficient of the sound image at the transition position is calculated. Then, the sound image position is changed by applying it to the signal processing means.

[0025]

According to the present invention having such a feature, a stereo acoustic signal is input from the signal input means, and a sound image is localized in an arbitrary direction by using a speaker of each channel at a reference listening position. Convolve the tap coefficients.
When the sound image position changes, the tap coefficient calculation means calculates the tap coefficient so that the tap coefficient gradually changes, and alternately stores the tap coefficient in the coefficient register. With this arrangement, the signal processing means can output a realistic acoustic signal without switching noise even if the sound image position is switched by the sound image position input means.

[0026]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A sound field reproducing apparatus according to a first embodiment of the present invention (claims 1, 2, and 3) will be described with reference to FIGS. 1 and 2 are block diagrams showing the configuration of the sound field reproducing apparatus in the first embodiment, and like the conventional example, the signal input means 1, the A / D converter 2, the signal processing means 3, and the D / A. Converter 4, amplifier 5, speaker 6, tap coefficient control means 7,
Coefficient storage means 10 and sound image position input means 11 are provided.

As shown in FIG. 1, in this embodiment, a tap coefficient calculating means 20 is provided in addition to the conventional example. When the control signal is input from the tap coefficient control means 7, the tap coefficient calculation means 20 is a circuit for extracting the tap coefficient at a different sound image position from the coefficient storage means 10 and calculating the tap coefficient at the intermediate (transition state) sound image position. is there. The tap coefficient calculated by the Lch tap coefficient calculation means 20a is given to the coefficient registers 21a and 22a. Similarly, Rch
The tap coefficient calculated by the tap coefficient calculating means 20b is given to the coefficient registers 21b and 22b. The coefficient registers 21 and 22 are circuits for temporarily holding tap coefficients of adjacent sound image positions, and the outputs thereof are alternately given to the signal processing means 3.

The operation of switching the present sound image position Q0 to the destination sound image position Qx in the sound field reproducing apparatus of this embodiment having the above-described structure will be described with reference to FIG. In this case, a plurality of sound image positions Qj are set between the sound image positions Q0 and Qx, and control for smoothly changing the sound image position is performed. Now, in FIG. 1, when an analog acoustic signal is input from the signal input means 1, A / D conversion circuits 2a and 2b are input.
Are converted into digital signals by the signal processing means 3a, 3
Input to b. On the other hand, the sound image position input means 11 inputs the sound image position Qx. A control signal is output from the tap coefficient control means 7 according to the relationship between the input sound image positions. Then, the tap coefficients at the current sound image position Q0 and the destination sound image position Qx are read from the coefficient storage means 10a and 10b, respectively, and transferred to the tap coefficient calculation means 20a and 20b.

Let p be the number of multipliers (total number of taps) of the FIR filter constituting the signal processing means 3, and i (0≤i≤p
Consider the tap coefficients T0i and Txi of the -1) th multiplier.
Here, in the i-th multiplier, T0i is the tap coefficient at the current sound image position, and Txi is the tap coefficient of the movement destination.

The tap coefficient calculating means 20a, 20b gradually change the current tap coefficient T0i to the tap coefficient (destination or target position) Txi to be changed. Assuming that the tap coefficient during the transition for this purpose is Tji (j is an integer 0 ≦ i <m), the tap coefficient Tji is expressed by the following equation (1)
Given in 3).

[Equation 3] Here, when moving from the sound image position Q0 of No. 0 (current position) to the sound image position Qx of No. x (moving destination) via the sound image position Qj in the middle of transition, the current position and the moving destination are divided into m pieces.

The method of changing the tap coefficient Tji in this embodiment is as follows.
This is to linearly change (linearly interpolate) from the current tap coefficient T0i toward the tap coefficient Txi at the destination sound image position. The tap coefficient Tji thus obtained
Is set in the coefficient register 21 or 22 on the side not currently used for the convolution operation. When the tap coefficient is set in the coefficient register 21 or 22, the coefficient register 21 or 22 used for the convolution calculation is switched to the coefficient register 21 or 22 in which the tap coefficient is newly set. Then, the tap coefficient of the coefficient register 21 or 22 holding the tap coefficient to be convolved is output to the signal processing means 3.

That is, the value of j is incremented from 0 to m, and the signal processing means 3 is operated with the tap coefficient Tji in the transition. In the signal processing means 3a, 3b, the coefficient registers 21a, 22a and the coefficient registers 21b, 2 are included.
The tap coefficient set to 2b and the input signal are convolved. Next, the acoustic signals respectively input to the signal processing means 3a and 3b and the tap coefficient for sound image localization are convoluted. This operation is repeated until the tap coefficient of the sound image position Qx of the moving destination input to the sound image input means 11 is reached.

FIG. 3 is a graph showing changes in the tap coefficient thus generated. In the figure, at a certain time of the sound image position Q0, the FIR filter (signal processing means 3) having p multipliers starts from the 0th multiplier on the horizontal axis to p.
The tap coefficients up to the -1st multiplier are currently set to the values shown by the solid line A. Further, the tap coefficient of the sound image position Qx, which is the movement destination, is set to a value as shown by the chain line C. The tap coefficient calculation means 20 uses the tap coefficient T in the transition state.
j will be changed as shown by the broken line B.

The output signals of the signal processing means 3a and 3b are converted into analog signals by the D / A converters 4a and 4b of FIG. 2 and then reproduced by the speakers 6a and 6b. As a result, the listener 12 can localize the sound image to the intended position without noise when switching the sound image position. In this way, the sound image can be moved in real time without noise.

Next, a sound field reproducing apparatus according to the second embodiment of the present invention will be described. The sound field reproducing apparatus of the second embodiment (claim 4) has the same overall structure as that shown in FIGS. 1 and 2, and the description of the structure is omitted.

Now, the operation of the sound field reproducing apparatus of this embodiment will be described. First, in the tap coefficient calculation means 20, when switching from the current tap coefficient T0i to the tap coefficient Txi at the sound image position of the moving destination, the tap coefficient Tji in the middle of transition is gradually changed by using exponential function interpolation. The tap coefficient Tji in this embodiment is expressed by the following equation (14).

[Equation 4] Various symbols, including subscripts, used in this formula are expressed in formula (1
It is the same as the case of 3).

The tap coefficient Tji in the equation (13) changes the tap coefficient Tji exponentially from the current tap coefficient T0i to the tap coefficient Txi at the sound image position of the moving destination. The tap coefficient thus obtained is set in the coefficient register 21 or 22 on the side not currently used for the convolution operation. When the tap coefficient is set in the coefficient register 21 or 22, the coefficient register used for the convolution operation is switched to the side in which the tap coefficient is newly set.
Similar to the case of the first embodiment, this operation is repeated until the tap coefficient of the input sound image position Qx is reached. The operation after the signal processing means 3 is similar to that of the first embodiment.

As a result, it is possible to prevent a rapid change in the tap coefficient, and it is possible to smoothly move the sound image without generating noise at the time of switching.

Next, a sound field reproducing apparatus according to a third embodiment (claim 5) of the present invention will be described. The overall structure of the sound field reproducing apparatus of this embodiment is almost the same as that shown in FIGS. 1 and 2, but the tap coefficient calculating means 30 is used instead of the tap coefficient calculating means 20. Other configurations are similar to those of the first and second embodiments. FIG. 4 is a block diagram of the tap coefficient calculation means 30 in the sound field reproducing apparatus of this embodiment.

In FIG. 4, the first memory 31 is a circuit for storing the tap coefficient at the current sound image position, and the second memory 32 is a circuit for storing the tap coefficient at the destination sound image position. The FFT processing unit 33 includes the memories 31, 3
Since the tap coefficient input from 2 is a value expressed on the time axis, this is a circuit that converts this into a signal on the frequency axis by fast Fourier transform (FFT). The calculation unit 34 is F
It is a circuit for performing an operation for converting to a desired frequency characteristic when the signal of the FT processing unit 33 is input. The inverse FFT processing unit 35 is a circuit that, when a signal on the frequency axis is input from the calculation unit 34, restores the signal on the time axis.

The operation of the sound field reproducing apparatus having such tap coefficient calculating means 30 will be described. First, as in the first embodiment, the Lch and Rch tap coefficient calculating means 30a, 30b of FIG. 1 store the tap coefficient of the sound image position input by the sound image position input means 11 in the coefficient storing means 10a, 10b.
Read from each. The FFT processing unit 33 shown in FIG. 4 performs FFT on the tap coefficients set in the memories 31 and 32 and transforms them into the frequency domain. Then, the calculation unit 34 obtains the frequency characteristic of the tap coefficient to be convolved from the converted characteristic.
Then, calculation is performed so as to linearly change from the current frequency characteristic of the tap coefficient to the frequency characteristic of the tap coefficient at the sound image position of the moving destination using linear interpolation.

The following equation (15) shows how to obtain it.

[Equation 5] Note that in equation (15), the tap coefficient T of equation (13) is
Is replaced by the letter F, and the meaning of each symbol including the subscript is the same except that the variables are displayed in the frequency domain.

Next, the inverse FFT processing unit 35 performs inverse FFT on the obtained frequency characteristic to obtain the tap coefficient converted into the time domain. This tap coefficient is set in the coefficient register 21 or 22. These operations are performed by the sound image position input means 11
It is repeated until the tap coefficient of the sound image position input in is reached.

As described above, in the present embodiment, the tap coefficient calculation means 30a, 30b convert the tap coefficient of the sound image position to be moved into the frequency axis, and after obtaining the movement amount on the frequency axis,
Return to the time axis and find the actual tap coefficient. By doing so, it is possible to prevent a rapid change in the tap coefficient and realize smooth sound image movement without switching noise. FIG. 5 is a graph showing changes in tap coefficient. Here, the frequency characteristic F0 of the current tap coefficient and the frequency characteristic Fx of the destination tap coefficient Fx
Are displayed on the frequency axis like a solid line A and a chain line C, respectively. The tap coefficient calculation means 30 outputs the frequency characteristic Fj of the tap coefficient during the transition as shown by the broken line B.

Next, a sound field reproducing apparatus according to a fourth embodiment (claim 6) of the present invention will be described. Since the configuration of the sound field reproducing apparatus of this embodiment is the same as that of the third embodiment, the description thereof will be omitted and the operation will be described.

As described in the third embodiment, the tap coefficient calculating means 30a, 30b stores the tap coefficient of the sound image position input by the sound image position input means 11 in the coefficient storing means 10a,
It is read from 10b and stored in the memories 31 and 32. The FFT processing unit 33 in FIG. 4 transforms the tap coefficients set in the memories 31 and 32 into the frequency domain. Then, the calculation unit 34 obtains the frequency characteristic of the tap coefficient to be convolved from the converted characteristic. Equation (16) shows how to obtain it.

[Equation 6]

In the equation (16), the arithmetic unit 34 performs an arithmetic operation using exponential interpolation to change exponentially from the current frequency characteristic of the tap coefficient to the frequency characteristic of the tap coefficient at the destination sound image position. To do. Then the inverse FFT
The processing unit 35 performs inverse FFT on the obtained frequency characteristic and transforms it into the time domain to obtain the tap coefficient. This tap coefficient is set in the coefficient register 21 or 22. These operations are repeated until the tap coefficient of the input sound image position is reached.

In this way, the tap coefficient calculating means 30a,
The tap coefficient of the sound image position to be moved in 30b is converted to the frequency axis, the movement amount is obtained on the frequency axis, and then returned to the time axis to obtain the actual tap coefficient. This makes it possible to prevent a rapid change in the tap coefficient, and to smoothly move the sound image without switching noise.

In each of the above embodiments, all the data obtained by the tap coefficient computing means 20 or 30 are stored in the coefficient storage means 10a, 10b for storing the tap coefficient, and the current sound image position is gradually moved to the desired sound image position. It is also possible to switch the tap coefficient to, and in this case as well, the same effect as that of the sound field reproducing device described above can be obtained.

Further, the sound image moving method of the sound field control device of the present invention has been described by the method using the head related transfer function, but when the method for realizing the sound image localization by the method using the delay time and the level is used. However, it is possible to use the same tap coefficient switching method. Further, although the embodiment of the present invention describes the case of speaker reproduction, the same switching method can be used also when changing the sound image using headphones.

[0051]

As described above in detail, in the sound field reproducing apparatus of the present invention, when the sound image position is moved, the tap coefficient calculating means determines the tap coefficient of the current sound image position and the sound image position of the moving destination. The calculation is performed based on the tap coefficient so that the tap coefficient changes smoothly. Therefore, unnecessary noise does not occur when the sound image position is switched, and the listener can obtain the surround sound accompanied by the sound image movement only by the left and right channel speakers.

[Brief description of drawings]

FIG. 1 is a block diagram (No. 1) showing a configuration of a sound field reproducing apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram (No. 2) showing the configuration of the sound field reproducing apparatus in the first embodiment.

FIG. 3 is a graph showing changes in tap coefficients in the sound field reproducing devices of the first and second embodiments.

FIG. 4 is a block diagram showing a configuration of a tap coefficient calculating means provided in the sound field reproducing apparatus according to the third and fourth embodiments of the present invention.

FIG. 5 is a graph showing changes in tap coefficients in the sound field reproducing devices of the third and fourth examples.

FIG. 6 is a block diagram (No. 1) showing a configuration of a conventional sound field reproducing device.

FIG. 7 is a block diagram (No. 2) showing a configuration of a conventional sound field reproducing device.

[Explanation of symbols]

1 signal input means 2,2a, 2b A / D converter 3,3a, 3b signal processing means 4,4a, 4b D / A converter 5,5a, 5b amplifier 6,6a, 6b, 14 speaker 7,7a, 7b tap coefficient control means 10, 10a, 10b coefficient storage means 11 sound image position input means 12 listener 13 signal generation means 20, 20a, 20b, 30, 30a, 30b tap coefficient calculation means 21, 21a, 21b coefficient registers 31, 32 Memory 33 FFT processing unit 34 Operation unit 35 Inverse FFT processing unit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Toshihiko Date 1006 Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (6)

[Claims] 1. A signal input means for outputting a stereophonic acoustic signal, a plurality of A / D converters for converting an analog signal of each channel input from the signal input means into a digital signal, and a plurality of tap coefficients. A plurality of signal processing means configured by digital filters which are set and which perform a convolution operation by inputting the signal of the A / D converter, a sound image position input means for inputting a sound image position, and an instruction by the sound image position input means Coefficient storage means for storing a plurality of sets of tap coefficients of the signal processing means corresponding to all the sound image positions that may be generated, and when the sound image position is switched by the sound image position input means, Tap coefficient calculation means for reading out tap coefficients of sound image positions from the coefficient storage means, respectively, and calculating tap coefficients of sound image positions that gradually transition from the current position to the destination. A plurality of coefficient registers for temporarily holding the tap coefficients calculated by the tap coefficient calculation means and setting them in the signal processing means; and a plurality of D registers for converting the output signals of the respective channels from the signal processing means into analog signals. An A / A converter, a plurality of amplifiers that power-amplifies the output of the D / A converter, and a speaker of each channel that receives a signal of the amplifier and reproduces sound including surround sound, The tap coefficient of the sound image position designated by the sound image position input means is read from the coefficient storage means and set in the signal processing means to output the surround sound from the speaker, and the sound image position input means switches the sound image position. The tap coefficient of the sound image at the transition position is calculated and given to the signal processing means to change the sound image position. Sound field playback device. 2. The signal processing means, from the tap coefficient used in the convolution calculation at the current sound image position to the tap coefficient for the sound image position newly designated by the sound image position input means, 2. The sound field reproducing device according to claim 1, wherein the tap coefficient of is gradually changed. 3. The signal processing means from the tap coefficient used in the convolution calculation at the current sound image position to the tap coefficient for the sound image position newly designated by the sound image position input means. 3. The sound field reproducing apparatus according to claim 2, wherein the tap coefficient of is changed by linear interpolation. 4. The signal processing means from the tap coefficient used in the convolution calculation at the current sound image position to the tap coefficient for the sound image position newly designated by the sound image position input means. 3. The sound field reproducing apparatus according to claim 2, wherein the tap coefficient of is changed by exponential function interpolation. 5. The tap coefficient calculation means includes a first memory that holds a tap coefficient at a current sound image position, a second memory that holds a tap coefficient at a destination sound image position, and the first and the first memories. A Fourier transform processing unit that transforms the tap coefficient in the time domain held in the memory 2 into the frequency domain; an arithmetic unit that transforms the tap coefficient transformed by the Fourier transform processing unit into a predetermined frequency characteristic; An inverse Fourier transform processing unit for converting the output of the operation unit into a signal in the time domain, and a new instruction is given by the sound image position input means from the tap coefficient used for the convolution operation at the current sound image position. 3. The sound field reproducing apparatus according to claim 2, wherein the tap coefficient of the signal processing means is changed by linear interpolation up to the tap coefficient for the sound image position. 6. The tap coefficient calculation means includes a first memory that holds a tap coefficient at a current sound image position, a second memory that holds a tap coefficient at a destination sound image position, and the first and the first memories. A Fourier transform processing unit that transforms the tap coefficient in the time domain held in the memory 2 into the frequency domain; an arithmetic unit that transforms the tap coefficient transformed by the Fourier transform processing unit into a predetermined frequency characteristic; An inverse Fourier transform processing unit for converting the output of the operation unit into a signal in the time domain, and a new instruction is given by the sound image position input means from the tap coefficient used for the convolution operation at the current sound image position. 3. The sound field reproducing apparatus according to claim 2, wherein the tap coefficient of the signal processing means is changed by exponential interpolation up to the tap coefficient for the sound image position.