JP4922211B2 - Acoustic signal converter, method and program thereof - Google Patents

Description

  The present invention relates to an acoustic signal conversion device that converts an acoustic signal corresponding to a certain number of channels into an acoustic signal corresponding to a number of channels different from the number of channels, a method thereof, and a program thereof.

Conventionally, a downmix method for converting an audio signal for a 5.1 channel surround system into an audio signal for a 2 channel stereo system has been standardized by the ITU-R (International Telecommunication Union Radiocommunication Division) ( For example, refer nonpatent literature 1).
In addition to this standard, an invention for downmixing is disclosed (for example, see Patent Document 1). The downmix device described in Patent Document 1 generates an acoustic signal after downmix that accurately reproduces a three-dimensional localization from an acoustic signal before downmix.

An invention for panning an acoustic signal three-dimensionally is also disclosed (see, for example, Patent Documents 2 and 3). The three-dimensional panning device described in Patent Document 2 pans an acoustic signal three-dimensionally regardless of the arrangement of speakers. In addition, the three-dimensional panning apparatus described in Patent Document 3 reproduces the three-dimensional panning of a sound source as an acoustic panning formed by acoustic signals radiated from a plurality of speakers.
JP 2007-53624 A JP 2007-194900 A JP 2007-329746 A RECOMMENDATION ITU-R BS. 775-2 “MULTICHANNEL STREOPHONIC SOUND SYSTEM WITH AND WITHOUT ACCOMPANYING PICTURE”

  However, it is possible to arbitrarily convert the number of channels of an acoustic signal, for example, an acoustic signal for a 5.1 channel surround system, an acoustic signal for a 22.2 channel acoustic system for Super Hi-Vision, or WFS (Wave Field Synthesis). It is not common to convert to an acoustic signal for a wavefront synthesis acoustic system. Therefore, in order to promote the distribution of audio content, there is a demand to arbitrarily convert the number of channels of the audio signal without limiting the number of channels for transmitting or recording the audio signal and the number of channels for reproducing the audio signal. Yes.

  In addition, the inventions described in Patent Documents 1 to 3 convert an acoustic signal so as to provide a good listening environment at a sound receiving point (usually near the user's head) indicating one point. For this reason, in the inventions described in Patent Documents 1 to 3, the user moves the body a little and the head moves away from the sound receiving point. Therefore, the movement of the user and a plurality of users are the same in the same space. There is a problem that it is not possible to respond to listening to music and a good listening environment cannot be provided.

  An object of the present invention is to provide an acoustic signal converter, a method thereof, and a program thereof that can arbitrarily convert the number of channels of an acoustic signal and provide a good listening environment when reproducing the converted acoustic signal. .

  In order to solve the above-described problem, the acoustic signal conversion device according to claim 1 converts an original acoustic signal corresponding to one or more channels into a reproduced acoustic signal corresponding to a number of channels different from the number of channels of the original acoustic signal. An acoustic signal conversion device that includes Fourier transform means, original sound field vector generation means, acoustic signal conversion means, reproduction sound field vector generation means, conversion matrix setting means, and Fourier inverse transform means It was.

  In such a configuration, the acoustic signal conversion device Fourier-transforms the input original acoustic signal into an original acoustic physical quantity signal indicating an acoustic physical quantity in the original sound field for reproducing the original acoustic signal by the Fourier transform means. Thus, the acoustic signal converter can handle the acoustic physical quantity in the original sound field by Fourier transforming the original acoustic signal. Here, the acoustic physical quantity includes at least one of sound pressure and particle velocity. In the acoustic signal conversion device, the original sound field vector generation means generates an original sound field vector whose component is an acoustic physical quantity in the original sound field, based on the original acoustic physical quantity signal Fourier-transformed by the Fourier transform means.

  The acoustic signal conversion device indicates an acoustic physical quantity in a reproduction sound field for reproducing a reproduction acoustic signal based on a predetermined transformation matrix from the original acoustic physical quantity signal Fourier-transformed by the Fourier transformation means by the acoustic signal conversion means. Converts to a reproduced acoustic physical quantity signal. In this way, by converting the original acoustic physical quantity signal, the acoustic signal conversion device can handle the acoustic physical quantity in the reproduction sound field. Here, the acoustic signal converter uses a transformation matrix as an initial value in order to obtain a reproduced acoustic physical quantity signal.

  In the acoustic signal conversion device, the reproduction sound field vector generation unit generates a reproduction sound field vector having an acoustic physical quantity in the reproduction sound field as a component based on the reproduction acoustic physical quantity signal converted by the acoustic signal conversion unit. Further, the acoustic signal conversion apparatus sets a conversion matrix that minimizes the square error between the original sound field vector and the reproduced sound field vector in a predetermined sound receiving region by the conversion matrix setting unit. Here, the sound receiving area is arbitrarily set in shape and size according to the number of users, the user's head position, and other listening environments, for example, a square of 0.4 m in length and 0.4 m in width. It is set as a space which is a certain plane or a cube of 0.4 m length × 0.4 m width × 0.4 height.

  The acoustic signal conversion apparatus reconverts the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal based on the transformation matrix set by the transformation matrix setting means by the acoustic signal conversion means. Thus, by reconverting the original acoustic physical quantity signal, the acoustic signal conversion device can minimize the square error between the acoustic physical quantity in the original sound field and the acoustic physical quantity in the reproduced sound field. Further, the acoustic signal conversion device performs inverse Fourier transform of the reproduced acoustic physical quantity signal converted by the acoustic signal conversion unit into a reproduced acoustic signal by the inverse Fourier transform unit. Thereby, the acoustic signal converter can obtain a reproduced acoustic signal that is substantially the same as the original acoustic signal or as close as possible. Furthermore, since the acoustic signal converter corresponds to a sound receiving area wider than the sound receiving point indicating one point, it can cope with movements of users and when a plurality of users want to listen to the same music in the same space. .

The acoustic signal conversion device according to claim 2 further includes storage means for storing the position of the original speaker corresponding to the channel of the original acoustic signal and the position of the reproduction speaker corresponding to the channel of the reproduced acoustic signal, and a conversion matrix. If the reproduction speaker is arranged three-dimensionally as in the 22.2 channel sound system, the setting means is adjacent in the reproduction sound field based on the position of the original speaker and the position of the reproduction speaker stored by the storage means. A divided area surrounded by three or more reproduction speakers is obtained, and for an original speaker in which a virtual line to the sound receiving area passes through the divided area, the channel of the original sound signal corresponding to the original speaker is set to three or more reproduction speakers. It is characterized in that a conversion matrix assigned to the channel of the corresponding reproduced sound signal is set.
On the other hand, when the reproduction speakers are two-dimensionally arranged as in the 5.1 channel surround system, the transformation matrix setting unit may use a divided area surrounded by two or more reproduction speakers.

  In such a configuration, the acoustic signal conversion device can prevent a situation in which the direction in which the sound of a channel in the original sound field is heard and the direction in which the sound of the channel is heard in the reproduction sound field are completely different.

  In the acoustic signal conversion device according to claim 3, the Fourier transform means uses the original sound signal as an acoustic physical quantity in the original sound field, the sound pressure in the original sound field as a scalar quantity, and the original sound field as a three-dimensional vector quantity. The original sound field vector generation means generates an original sound field vector that is a four-dimensional vector having the sound pressure in the original sound field and the particle velocity in the original sound field as components. The acoustic signal conversion means includes the original acoustic physical quantity signal as the acoustic physical quantity in the reproduced sound field, and includes the sound pressure in the reproduced sound field as a scalar quantity and the particle velocity in the reproduced sound field as a three-dimensional vector quantity. It is converted into a reproduction acoustic physical quantity signal, and the reproduction sound field vector generating means generates a reproduction sound field vector that is a four-dimensional vector having the sound pressure in the reproduction sound field and the particle velocity in the reproduction sound field as components. Features.

  In this configuration, the acoustic signal converter can handle a four-dimensional vector having sound pressure and particle velocity as components as the original sound field vector and the reproduced sound field vector. Thus, the acoustic signal conversion device does not need to convert the sound pressure, which is a scalar amount, into a three-dimensional vector.

  In addition, in order to solve the above-described problem, the acoustic signal conversion method according to claim 4 uses an original acoustic signal corresponding to one or more channels as a reproduced acoustic signal corresponding to a number of channels different from the number of channels of the original acoustic signal. An acoustic signal conversion method for converting to a Fourier transform step, an original sound field vector generation step, an acoustic signal conversion step, a reproduction sound field vector generation step, a conversion matrix setting step, an acoustic signal reconversion step, And a Fourier inverse transform step.

  In order to solve the above-described problem, an acoustic signal conversion program according to claim 5 is configured to reproduce an original sound signal corresponding to one or more channels with a reproduced sound corresponding to a number of channels different from the number of channels of the original sound signal. In order to convert the signal, the computer is configured to function as a Fourier transform unit, an original sound field vector generation unit, an acoustic signal conversion unit, a reproduction sound field vector generation unit, a transformation matrix setting unit, and a Fourier inverse transform unit.

According to the acoustic signal conversion device, the method and the program thereof according to the present invention, the following excellent effects can be obtained.
According to the inventions according to claims 1, 4, and 5, it is possible to obtain a reproduced sound signal that is substantially the same as the original sound signal or as close as possible, and the movement of the user and a plurality of users are the same in the same space. Since it is possible to cope with listening to music, it is possible to arbitrarily convert the number of channels of the acoustic signal, and to provide a good listening environment when reproducing the converted reproduced acoustic signal.
According to the second aspect of the present invention, it is possible to prevent a situation in which the direction in which the sound of the channel that is the original sound field is heard from the direction in which the sound of the channel is heard in the reproduction sound field is completely different. can do.
According to the invention of claim 3, since the scalar quantity can be handled as it is, it is not necessary to convert the sound pressure into a three-dimensional vector.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings as appropriate. In each embodiment, means having the same function and the same member are denoted by the same reference numerals, and description thereof is omitted.

[Outline of acoustic signal converter]
First, the outline of the acoustic signal converter according to the present invention will be briefly described.
The sound pressure is a minute change in air pressure when a sound wave arrives, and is given as a scalar amount. In a space where the sound reverberation is not so large, the Fourier transform of the sound pressure at the sound receiving point that is separated from a certain sound source by a predetermined distance is expressed by Expression (1).

  The particle velocity is a velocity at which the fluid (air) minutely vibrates when a sound wave arrives, and is given as a three-dimensional vector quantity. Further, under the same condition as that of the sound wave, the Fourier transform of the particle velocity is expressed by the equation (2).

  Therefore, when the sound signal is converted by controlling both the sound pressure and the particle velocity, the sound pressure and the four-dimensional components having the respective components of the particle velocity in the x, y, and z-axis directions (three-dimensional directions) as elements. Is generated so that the square error of the sound field vector is minimized. That is, the acoustic signal conversion apparatus according to the present invention obtains a conversion matrix that minimizes the square error of the sound field vector. When there are a plurality of speakers, the sound field vectors for reproducing the sound with all the speakers can be obtained by adding the sound field vectors of the speakers at the sound receiving point.

[Configuration of acoustic signal converter]
Next, the configuration of the acoustic signal conversion device will be described in detail with reference to FIG. FIG. 1 is a block diagram of an acoustic signal converter according to an embodiment of the present invention. The acoustic signal conversion apparatus 1 is configured to implement a Fourier transform unit 10, an original sound field vector generation unit 20, an acoustic signal conversion unit 30, a reproduction sound field vector generation unit 40, and a conversion matrix setting in order to realize each function described later. Means 50, inverse Fourier transform means 60, and storage means 70 are provided.

In FIG. 1, the original sound signal is indicated by s (t), the original sound physical quantity signal is indicated by s (ω), the original sound field vector is indicated by Φ, and the transformation matrix is indicated by W. Also, the playback sound signal q (t), showed playback sound physical quantity signal q (omega) and the reproduction sound field vector Φ in ^~. Here, the original sound signal s (t) is input by 6 channels (not shown) if it corresponds to the number of channels of the original sound signal, for example, a 5.1 channel surround system. The reproduced sound signal q (t) is output in 24 channels (not shown) if the number corresponds to the number of channels of the reproduced sound signal, for example, 22.2 channel sound system.

  The Fourier transform means 10 performs a Fourier transform on the input original sound signal to an original sound physical quantity signal indicating an acoustic physical quantity in an original sound field for reproducing the original sound signal. Further, the Fourier transform means 10 uses the original acoustic signal as an acoustic physical quantity in the original sound field, and includes an original acoustic physical quantity including a sound pressure in the original sound field as a scalar quantity and a particle velocity in the original sound field as a three-dimensional vector quantity. Fourier transform the signal.

  Here, the Fourier transform means 10 Fourier transforms the original sound signal to the sound pressure in the original sound field using Expression (3).

  Further, the Fourier transform means 10 performs Fourier transform of the original sound signal to the particle velocity in the original sound field using Expression (4).

  The original sound field vector generation means 20 generates an original sound field vector whose component is an acoustic physical quantity in the original sound field, based on the original acoustic physical quantity signal Fourier-transformed by the Fourier transform means 10. Further, the original sound field vector generation means 20 generates an original sound field vector which is a four-dimensional vector having the sound pressure in the original sound field and the particle velocity in the original sound field as components.

  Here, the original sound field vector generation means 20 generates an original sound field vector represented by Expression (5).

  The acoustic signal conversion means 30 converts the original acoustic physical quantity signal Fourier-transformed by the Fourier transformation means 10 into a reproduced acoustic physical quantity signal indicating an acoustic physical quantity in a reproduction sound field for reproducing the reproduction acoustic signal based on a predetermined transformation matrix. To do. The acoustic signal conversion means 30 uses the original acoustic physical quantity signal as the acoustic physical quantity in the reproduced sound field, the sound pressure in the reproduced sound field as a scalar quantity, and the particle velocity in the reproduced sound field as a three-dimensional vector quantity. Is converted to a reproduced acoustic physical quantity signal.

  At this time, a conversion matrix is required to convert the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal. However, when the acoustic signal conversion device 1 is used for the first time, or when the acoustic signal conversion device 1 is made to correspond to a new acoustic system, there are cases where no conversion matrix is set. For this reason, a conversion matrix as an initial value is stored in the storage means 70, and the acoustic signal conversion means 30 converts the original acoustic physical quantity signal into a reproduced acoustic physical quantity signal using the conversion matrix as an initial value. Then, the conversion matrix setting means 50 described later sets a conversion matrix that minimizes the square error, and the acoustic signal conversion means 30 reconverts the original acoustic physical quantity signal into a reproduced acoustic physical quantity signal based on this transformation matrix. .

  Here, the acoustic signal conversion means 30 converts the original acoustic physical quantity signal into a sound pressure in the reproduction sound field using Expression (6).

  The acoustic signal conversion means 30 converts the original acoustic physical quantity signal into a particle velocity in the reproduction sound field using Expression (7).

  From the above, the acoustic signal conversion means 30 converts the original acoustic physical quantity signal into a reproduced acoustic physical quantity signal using Expression (8).

  The reproduction sound field vector generation unit 40 is configured to generate a reproduction sound field vector having an acoustic physical quantity in the reproduction sound field as a component based on the reproduction acoustic physical quantity signal converted by the acoustic signal conversion unit 30. The reproduction sound field vector generation means 40 generates a reproduction sound field vector that is a four-dimensional vector having the sound pressure in the reproduction sound field and the particle velocity in the reproduction sound field as components. Here, the reproduction sound field vector generation means 40 generates a reproduction sound field vector represented by Expression (9). At this time, the definition represented by Formula (10) is used.

Here, when only the sound pressure that is a scalar quantity is handled as the acoustic physical quantity, for example, as in Patent Documents 1 to 3, it is necessary to convert the sound pressure that is the scalar quantity into a three-dimensional vector.
On the other hand, the acoustic signal conversion device 1 can handle the sound pressure as a scalar amount as it is by using the particle velocity in addition to the sound pressure, and makes such conversion unnecessary.

  The conversion matrix setting means 50 sets a conversion matrix that minimizes the square error between the original sound field vector and the reproduced sound field vector in a predetermined sound receiving area. Moreover, the conversion matrix setting means 50 can divide the area and set the conversion matrix. A specific example will be described later.

  Here, the transformation matrix setting means 50 obtains a transformation matrix that minimizes the square error represented by the equation (11).

  Here, since equation (11) depends on the original sound signal, an error function for calculating the square error is defined by equation (12). From Equation (12), the transformation matrix that minimizes the error function is represented by Equation (13).

  Furthermore, Expression (14) is defined as an error function for the sound receiving area. That is, in Expression (14), the error function is obtained while moving a certain point (sound receiving point) within the sound receiving region. From the above, the transformation matrix setting means 50 sets the transformation matrix that minimizes the error function, using Equation (15). Then, as described above, the acoustic signal converting means 30 reconverts the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal based on this conversion matrix.

  The inverse Fourier transform means 60 performs inverse Fourier transform of the reproduced acoustic physical quantity signal converted by the acoustic signal conversion means 30 into a reproduced acoustic signal. Here, the inverse Fourier transform means 60 performs inverse Fourier transform on the equations (3) and (4).

  The storage means 70 includes speaker coordinate storage means 71 for storing the position of the original speaker corresponding to the channel of the original acoustic signal and the position of the reproduction speaker corresponding to the channel of the reproduced acoustic signal. The storage means 70 includes a speaker number storage means 72 for storing the number of original speakers corresponding to the channel of the original acoustic signal and the number of reproduction speakers corresponding to the channel of the reproduced acoustic signal, the shape of the sound receiving area, You may provide the sound receiving area memory | storage means 73 which memorize | stores the magnitude | size and the position of the sound receiving area in a sound field.

  Here, the position of the original speaker and the position of the reproduction speaker can be specified by an arbitrary method. For example, these positions may be three-dimensional coordinates relative to a certain point in the space. When applied to an outdoor concert venue, the position of the speaker may be measured by GPS and used as the coordinates.

  The acoustic signal conversion device 1 may include input means such as a keyboard and a mouse for inputting the position of the original speaker, the position of the reproduction speaker, the number of original speakers, the reproduction speaker, the sound receiving area, and other necessary information ( Not shown). In addition, the acoustic signal conversion device 1 may include display means such as a display for displaying information such as the position of the original speaker.

  The Fourier transform unit 10, the original sound field vector generation unit 20, the acoustic signal conversion unit 30, the reproduction sound field vector generation unit 40, the transformation matrix setting unit 50, and the Fourier inverse transform unit 60 are, for example, a CPU (Central Processing Unit), A ROM (Read Only Memory) and a RAM (Random Access Memory) can be used. In addition, the storage unit 70 can be configured by, for example, an HDD (Hard Disk Drive).

[Specific examples of area division]
Hereinafter, a specific example of area division will be described in detail (see FIG. 1 as appropriate). FIG. 2 is an explanatory view in which a plane on which a speaker is arranged is viewed from the top in order to explain region division in the present invention. In FIG. 2, the original speaker is indicated by a white circle, and the reproduction speaker is indicated by a black circle. In FIG. 2, the original speaker is indicated by reference signs ξ (1) to ξ (4), the reproduction speaker is indicated by reference signs ζ (1) to ζ (7), the sound receiving area is indicated by reference sign L, and the divided areas are indicated. The virtual line which shows with the code | symbol R and connects the sound receiving area | region L and an original speaker was shown with the dashed-dotted line. Further, the divided region R is indicated by hatching. In FIG. 2, the sound receiving area L is set in a range including the user's head.

  Here, an example is shown in which the acoustic signal conversion device 1 divides an original acoustic signal corresponding to 10 channels (10 speakers) and converts it into a reproduced acoustic signal corresponding to 7 channels (7 speakers). For illustration, only four of the original speakers are shown in FIG.

Here, based on the position of the reproduction speaker stored in the storage unit 70, the transformation matrix setting unit 50 obtains the divided region R surrounded by the reproduction speakers adjacent in the reproduction sound field. Here, as shown in FIG. 2, the divided region R is a triangular range surrounded by the reproduction speakers ζ (1) to ζ (3) (reproduction acoustic signals q ₁ (ω) to q ₃ (ω)). Corresponding).

Further, the transformation matrix setting unit 50 determines the original speaker through which the virtual line to the sound receiving region L passes through the divided region R based on the position of the original speaker stored in the storage unit 70. Here, as shown in FIG. 2, the imaginary lines of the four native speakers ξ (1) to ξ (4) pass through the divided region R (the original acoustic signals s ₁ (ω) to s ₄ (ω)). Correspondence).

  Then, the transformation matrix setting means 50 reproduces the channel of the original acoustic signal corresponding to the original speaker corresponding to the three or more reproduction speakers for the original speaker through which the virtual line up to the sound receiving region L passes through the divided region R. Sets the transformation matrix to be assigned to the acoustic signal channel. Here, the transformation matrix setting means 50 corresponds to the channels of the original acoustic signal corresponding to the four original speakers ξ (1) to ξ (4) to the three reproduction speakers ζ (1) to ζ (3). A conversion matrix assigned to the channel of the reproduced sound signal to be set is set.

Such area division is expressed by Expression (16). In Expression (16), the first row of the transformation matrix is an element w ₁₁ , w ₂₁ , w indicating the correspondence between the original acoustic signal s ₁ (ω) and the reproduced acoustic signals q ₁ (ω) to q ₃ (ω). 0 except for ₃₁ . Further, the second to fourth rows of the transformation matrix corresponding to the original acoustic signals s ₂ (ω) to s ₄ (ω) are the same as the first row of the transformation matrix. Then, for elements other than 0, the transformation matrix setting means 50 obtains a value that minimizes the square error, and sets the transformation matrix.

  As described above, the optimization problem is divided into regions and handled locally, so that the direction in which the sound of a channel in the original sound field can be heard and the direction in which the sound of that channel can be heard in the playback sound field are completely different. Can be prevented.

  Here, the divided area R surrounded by the three reproduction speakers ζ (1) to ζ (3) has been described. However, the transformation matrix setting unit 50 performs the above operation on the divided areas surrounded by other reproduction speakers. The divided regions may be performed. Further, the transformation matrix setting means 50 may perform the above-described area division on a divided area surrounded by two reproduction speakers or a divided area surrounded by four or more reproduction speakers. Note that the divided area surrounded by the two reproduction speakers is a linear area.

[Operation of acoustic signal converter]
Hereinafter, the operation of the acoustic signal converter will be described (see FIG. 1 as appropriate). FIG. 3 is a flowchart showing the operation of the acoustic signal converter of FIG.

  The acoustic signal converter 1 performs Fourier transform on the input original sound signal by the Fourier transform means 10 into an original sound physical quantity signal indicating an acoustic physical quantity in the original sound field for reproducing the original sound signal (step S1). In the acoustic signal conversion device 1, the original sound field vector generation means 20 generates an original sound field vector whose component is an acoustic physical quantity in the original sound field, based on the original sound physical quantity signal Fourier-transformed by the Fourier transform means 10. Step S2).

  Following step S2, the acoustic signal conversion device 1 reproduces the reproduced acoustic signal from the original acoustic physical quantity signal Fourier-transformed by the Fourier transform unit 10 based on a predetermined transformation matrix by the acoustic signal conversion unit 30. Is converted into a reproduced acoustic physical quantity signal indicating the acoustic physical quantity at step S3. In step S <b> 3, the acoustic signal conversion device 1 converts the original acoustic physical quantity signal into a reproduced acoustic physical quantity signal by using the acoustic signal conversion unit 30 using the conversion matrix as an initial value.

  Subsequent to step S3, the acoustic signal conversion device 1 uses the reproduction sound field vector generation unit 40 to reproduce the reproduction sound having the component of the acoustic physical quantity in the reproduction sound field based on the reproduction acoustic physical quantity signal converted by the acoustic signal conversion unit 30. A field vector is generated (step S4). Further, the acoustic signal conversion device 1 sets the conversion matrix that minimizes the square error between the original sound field vector and the reproduced sound field vector in the predetermined sound receiving region by the conversion matrix setting means 50 (step S5). In step S <b> 5, the acoustic signal conversion device 1 may perform the above-described region division by the conversion matrix setting unit 50.

  Then, based on the conversion matrix set in step S5, the acoustic signal conversion device 1 reconverts the original acoustic physical quantity signal into a reproduced acoustic physical quantity signal by the acoustic signal conversion means 30 (step S6). Further, the acoustic signal conversion device 1 performs Fourier inverse transform on the reproduced acoustic physical quantity signal retransformed by the acoustic signal transforming unit 30 into the reproduced acoustic signal by the Fourier inverse transform unit 60 (step S7).

[Conversion result of acoustic signal]
Hereinafter, the conversion result of the acoustic signal by the acoustic signal conversion device will be described in comparison with the prior art. FIG. 4 shows 10 channels (distance between the speaker and the sound receiving area center point or the sound receiving point is 3 m) of the original sound signal in the intermediate layer of 22.2 channels, and 5 channels of the reproduced sound signal in 5.1 channel surround ( It is a graph which shows the upper limit of the square error when the distance between the speaker and the sound receiving area center point or the sound receiving point is converted to 2 m), and (a) is the least square error at the sound receiving point in the prior art. (B) is a graph in the case of optimization with the square error minimized in the sound receiving region in the present invention. In FIG. 4, the vertical axis represents the normalized square error, and the horizontal axis represents the position in the xy plane.

  Here, FIG. 4 normalizes the square error (see Equation (12)) obtained by experiment and shows it for each sound receiving point. Further, in FIG. 4A, the origin of the xy plane is set as the sound receiving point, and in FIG. 4B, the vertical 0.4 m × horizontal 0.4 m centered on the origin of the xy plane. The sound receiving area was set.

  In FIG. 4A, the square error at the origin (sound receiving point) on the plane is small, but the square error around the sound receiving point is large. That is, when the square error is minimized at the sound receiving point indicating one point, it is considered that the listening environment around the sound receiving point is not a good listening environment. On the other hand, in FIG. 4B, the square error is small even at and around the origin of the plane. That is, when the square error is minimized in the sound receiving area as in the present invention, it is considered that the listening environment is good including the origin of the plane and its surroundings.

  In each embodiment, the acoustic signal conversion device according to the present invention has been described as an independent device. However, in the present invention, a general computer can be operated by a program that functions as each of the above-described units. This program may be distributed via a communication line, or may be distributed by writing in a recording medium such as a CD-ROM or a flash memory.

  In each embodiment, an example of controlling sound pressure and particle velocity as an acoustic physical quantity has been described. However, the acoustic signal conversion device according to the present invention uses either sound pressure or particle velocity as an acoustic physical quantity. You may control, and you may control acoustic physical quantities, such as the instantaneous sound intensity which is a product of a sound pressure and a particle velocity, and the sound intensity which is a time average of an instantaneous sound intensity.

It is a block diagram of an acoustic signal converter concerning an embodiment of the present invention. It is explanatory drawing which looked at the plane where the speaker was arrange | positioned in order to demonstrate the area | region division in this invention. It is a flowchart which shows operation | movement of the acoustic signal converter of FIG. It is a graph which shows the upper limit of the square error at the time of converting 10 channels of the original sound signal in the intermediate layer of 22.2 channels into 5 channels of the reproduction | regeneration sound signal in 5.1 channel surround, (a) is conventional. It is a graph when the square error is minimized at the sound reception point in the technology, and (b) is a graph when optimization is performed with the square error is minimized in the sound reception region in the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Acoustic signal converter 10 Fourier transform means 20 Original sound field vector generation means 30 Acoustic signal conversion means 40 Reproduction sound field vector generation means 50 Transformation matrix setting means 60 Fourier inverse transform means 70 Storage means

Claims (5)

An acoustic signal conversion device that converts an original sound signal corresponding to one or more channels into a reproduced sound signal corresponding to a number of channels different from the number of channels of the original sound signal,
Fourier transform means for Fourier transforming the input original sound signal into an original sound physical quantity signal indicating an acoustic physical quantity in an original sound field for reproducing the original sound signal;
Based on the original acoustic physical quantity signal Fourier-transformed by the Fourier transform means, an original sound field vector generating means for generating an original sound field vector whose component is an acoustic physical quantity in the original sound field;
Acoustic signal conversion means for converting the original acoustic physical quantity signal Fourier-transformed by the Fourier transform means into a reproduced acoustic physical quantity signal indicating an acoustic physical quantity in a reproduced sound field for reproducing the reproduced acoustic signal based on a predetermined transformation matrix; ,
Based on the reproduced acoustic physical quantity signal converted by the acoustic signal converting means, a reproduced sound field vector generating means for generating a reproduced sound field vector having an acoustic physical quantity in the reproduced sound field as a component;
Transformation matrix setting means for setting the transformation matrix that minimizes a square error between the original sound field vector and the reproduced sound field vector in a predetermined sound receiving area;
A Fourier inverse transforming means for inversely transforming the reproduced acoustic physical quantity signal converted by the acoustic signal converting means into the reproduced acoustic signal;
The acoustic signal converter is characterized in that the acoustic signal converter reconverts the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal based on the transformation matrix set by the transformation matrix setting means. Storage means for storing the position of the original speaker corresponding to the channel of the original acoustic signal and the position of the reproduction speaker corresponding to the channel of the reproduced acoustic signal;
The transformation matrix setting means includes
Based on the position of the original speaker and the position of the reproduction speaker stored by the storage means, a divided area surrounded by the two or more reproduction speakers adjacent in the reproduction sound field is obtained, and a virtual line to the sound reception area is obtained. For the original speaker that passes through the divided region, the conversion matrix for assigning the channel of the original acoustic signal corresponding to the original speaker to the channel of the reproduced acoustic signal corresponding to the two or more reproduced speakers is set. The acoustic signal converter according to claim 1. The Fourier transform means includes, as the acoustic physical quantity in the original sound field, the sound pressure in the original sound field that is a scalar quantity and the particle velocity in the original sound field that is a three-dimensional vector quantity. Fourier transform to the original acoustic physical quantity signal,
The original sound field vector generating means generates the original sound field vector which is a four-dimensional vector having sound pressure in the original sound field and particle velocity in the original sound field as components,
The acoustic signal conversion means uses the original acoustic physical quantity signal as the acoustic physical quantity in the reproduced sound field, and the sound pressure in the reproduced sound field as a scalar quantity and the particle in the reproduced sound field as a three-dimensional vector quantity. Converting to the reproduced acoustic physical quantity signal including velocity,
The reproduction sound field vector generation means generates the reproduction sound field vector which is a four-dimensional vector having sound pressure in the reproduction sound field and particle velocity in the reproduction sound field as components. Item 3. The acoustic signal converter according to Item 1 or 2. An acoustic signal conversion method for converting an original sound signal corresponding to one or more channels into a reproduced sound signal corresponding to a number of channels different from the number of channels of the original sound signal,
A Fourier transform step of Fourier transforming the input original sound signal into an original sound physical quantity signal indicating an acoustic physical quantity in an original sound field for reproducing the original sound signal;
An original sound field vector generation step for generating an original sound field vector having an acoustic physical quantity in the original sound field as a component based on the original sound physical quantity signal Fourier-transformed in the Fourier transform step;
An acoustic signal conversion step of converting the original acoustic physical quantity signal Fourier-transformed in the Fourier transformation step into a reproduced acoustic physical quantity signal indicating an acoustic physical quantity in a reproduced sound field for reproducing the reproduced acoustic signal based on a predetermined transformation matrix; ,
Based on the reproduced acoustic physical quantity signal converted in the acoustic signal converting step, a reproduced sound field vector generating step for generating a reproduced sound field vector having an acoustic physical quantity in the reproduced sound field as a component;
A transformation matrix setting step for setting the transformation matrix that minimizes a square error between the original sound field vector and the reproduced sound field vector in a predetermined sound receiving region;
Based on the transformation matrix set in the transformation matrix setting step, an acoustic signal reconversion step for reconverting the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal;
A Fourier inverse transform step for inversely transforming the reproduced acoustic physical quantity signal retransformed in the acoustic signal transforming step into the reproduced acoustic signal;
An acoustic signal conversion method comprising: In order to convert the original sound signal corresponding to the number of channels of 1 or more into a reproduced sound signal corresponding to the number of channels different from the number of channels of the original sound signal,
Fourier transform means for Fourier transforming the input original sound signal into an original sound physical quantity signal indicating an acoustic physical quantity in an original sound field for reproducing the original sound signal;
An original sound field vector generating means for generating an original sound field vector whose component is an acoustic physical quantity in the original sound field, based on the original sound physical quantity signal Fourier-transformed by the Fourier transform means;
An acoustic signal converting means for converting the original acoustic physical quantity signal Fourier-transformed by the Fourier transform means into a reproduced acoustic physical quantity signal indicating an acoustic physical quantity in a reproduction sound field for reproducing the reproduced acoustic signal based on a predetermined transformation matrix;
Based on the reproduced acoustic physical quantity signal converted by the acoustic signal converting means, a reproduced sound field vector generating means for generating a reproduced sound field vector having an acoustic physical quantity in the reproduced sound field as a component,
Conversion matrix setting means for setting the conversion matrix that minimizes a square error between the original sound field vector and the reproduced sound field vector in a predetermined sound receiving region;
The reproduced acoustic physical quantity signal converted by the acoustic signal converting means functions as inverse Fourier transform means for inversely transforming the reproduced acoustic signal into the reproduced acoustic signal,
The acoustic signal conversion means reconverts the original acoustic physical quantity signal into the reproduced acoustic physical quantity signal based on the transformation matrix set by the transformation matrix setting means.