JP5010148B2 - 3D panning device - Google Patents

Description

  The present invention relates to a three-dimensional panning apparatus, and more particularly to a three-dimensional panning apparatus that can three-dimensionally pan an acoustic signal regardless of the arrangement of speakers.

  In an audio playback device such as a 2-channel audio system or 5.1 surround system, the volume of the sound source output from each speaker is changed by a panpot of a mixing console, and the sound image of the sound source is localized or moved. Was adjusting.

  However, in the above audio playback device, all speakers need to be arranged on a sound receiving point, that is, on a circle with a certain radius centered on the listener, so that the playback system is a rectangular playback field such as a theater. It was impossible to apply.

  Therefore, it is possible to move the position of the sound image by arranging a plurality of speakers along the periphery of a rectangular playback field such as a theater and delaying the output of one speaker relative to the output of another speaker. An audio processing apparatus has been proposed (see, for example, Patent Document 1).

  However, in the above sound processing apparatus, since it is necessary to arrange the speakers in the same horizontal plane, it is possible to pan the sound image two-dimensionally in the speaker arrangement plane, but the sound image is three-dimensionally arranged. It was impossible to pan.

  Therefore, in order to enable three-dimensional panning, it is possible to three-dimensionally pan a sound image by arranging speakers three-dimensionally and applying a vector reference amplitude panning method. A three-dimensional panning apparatus has already been proposed (see Non-Patent Document 1, for example).

Further, there has been proposed a three-dimensional panning device that can pan a sound image not only in the horizontal direction but also in the vertical direction by applying two FIR filters to two speakers arranged on the same plane (for example, Patent Documents). 2).
US Patent Publication No. 20020048380 ([0026], FIG. 3) Japanese Patent No. 3177714 ([0005], FIG. 1) “Virtual Sound Source Positioning Using Vector Base Amplitude Panning” VILLE PULKKI, J. Audio Eng. Soc. Vol. 45, No, 6, 1997 June

  However, in the conventional three-dimensional panning apparatus described above, the speaker needs to be installed on a spherical surface having a constant radius centered on the sound receiving point, that is, the position of the listener's ear, but in a reproduction field close to a rectangular parallelepiped, When a speaker is arranged on a spherical surface having a certain radius centered on the sound receiving point, there is a problem that the audible range is narrowed.

  The present invention has been made to solve the conventional problems, and an object of the present invention is to provide a three-dimensional panning apparatus capable of three-dimensionally panning an acoustic signal regardless of the arrangement of speakers. .

The three-dimensional panning apparatus of the present invention includes a pre-panning acoustic signal reading unit that reads a pre-panning acoustic signal that is a time function, a post-panning acoustic signal output unit that outputs a post-panning acoustic signal that is a time function, and the pre-panning acoustic signal. Arrangement for storing arrangement information of conversion means for converting a signal into an acoustic signal before frequency domain panning, localization position information generation means for generating localization position information to be added to the acoustic signal before panning, and acoustic signal output means after panning Information storage means, frequency domain panned acoustic signal generation means for generating a frequency domain panned acoustic signal obtained by panning the frequency domain pre-panned acoustic signal based on the localization position information and the arrangement information, and the frequency domain Post-panning acoustic signal as a function of time Seen including inverse transform means for inverse transformation to the frequency domain panning after acoustic signal generation means, acoustic physical quantity after panning at the sound receiving point when the emitted acoustic signal after the panning from the audio signal output means after the panning The frequency domain post-panning acoustic signal is generated such that the vector is equal to the pre-panning acoustic physical quantity vector at the sound receiving point when the pre-panning acoustic signal is emitted from a sound source .

With this configuration, the acoustic signal can be panned three-dimensionally.
Also, with this configuration, the acoustic signal can be panned three-dimensionally regardless of the arrangement of the speakers.

The three-dimensional panning apparatus of the present invention includes a pre-panning acoustic signal reading unit that reads a pre-panning acoustic signal that is a time function, a post-panning acoustic signal output unit that outputs a post-panning acoustic signal that is a time function, and the pre-panning acoustic signal. Arrangement for storing arrangement information of conversion means for converting a signal into an acoustic signal before frequency domain panning, localization position information generation means for generating localization position information to be added to the acoustic signal before panning, and acoustic signal output means after panning Information storage means, frequency domain panned acoustic signal generation means for generating a frequency domain panned acoustic signal obtained by panning the frequency domain pre-panned acoustic signal based on the localization position information and the arrangement information, and the frequency domain Post-panning acoustic signal as a function of time Inverse transform and inverse transform means for, wherein the said frequency domain panning after acoustic signal generation means, acoustic physical quantity vector after panning in the sound receiving area when radiated acoustic signal after the panning from the audio signal output means after the panning The frequency domain panned sound signal is generated so as to be equal to the pre-panning acoustic physical quantity vector in the sound receiving area when the sound signal is emitted from the sound source.

With this configuration, the acoustic signal can be panned three-dimensionally.
Also, with this configuration, a sound image can be localized for a large number of listeners.

  The present invention provides an arrangement for speaker arrangement by providing a frequency domain panned acoustic signal generating means for generating a frequency domain panned acoustic signal obtained by panning the frequency domain pre-panned acoustic signal based on the localization position information and the arrangement information. Regardless of this, it is possible to provide a three-dimensional panning apparatus having an effect that the acoustic signal can be panned three-dimensionally.

  Hereinafter, embodiments of a three-dimensional panning apparatus according to the present invention will be described with reference to the drawings.

  In this specification, an acoustic physical quantity vector is a vector indicating an acoustic physical quantity at a sound receiving point that receives an acoustic signal radiated from a point sound source, that is, a vector having at least one of sound pressure or particle velocity as a component. Alternatively, it means a sound intensity vector obtained by multiplying the particle velocity vector by a sound pressure that is a scalar quantity and integrating it over a certain time interval.

  Further, in the following description, boldface letters indicate vectors or matrices, and thin letters indicate scalar quantities.

(First embodiment)
As shown in the block diagram of FIG. 1, the embodiment of the three-dimensional panning apparatus 1 according to the present invention includes a pre-panning acoustic signal reading unit 11 that reads a pre-panning acoustic signal s (t) that is a time function, and a time function. A post-panning acoustic signal output means 12 for outputting a certain post-panning acoustic signal q (t), a conversion means 13 for converting the pre-panning acoustic signal s (t) into a frequency domain pre-panning acoustic signal S (ω), and before panning Localization position information generation means 14 for generating localization position information to be added to the acoustic signal s (t), arrangement information storage means 15 for storing arrangement information of the post-panning acoustic signal output means 12, localization position information, and arrangement information Frequency domain that generates a frequency domain post-panning acoustic signal Q (ω) obtained by panning the frequency domain pre-panning acoustic signal S (ω) based on Includes an acoustic signal generator 16 after N'ningu, and inverse transformation means 17 for inverse conversion after panning audio signal q (t) is an acoustic signal Q (omega) the time function after the frequency domain panning.

Then, the three-dimensional panning apparatus 1 according to the present invention has a post-panning acoustic physical quantity vector _a P at the sound receiving point when the post-panning acoustic signal q _m is radiated from the post-panning acoustic signal output means 12 of the number of channels M. panning front acoustic signal s _n from each of the sound source number n to generate a panning after acoustic signal q _m to equal the panning front acoustic physical quantity vector _{b P} at the sound receiving point k when the radiation.

  First, the sound pressure at the sound receiving point when there is one point sound source and one sound receiving point will be described.

The sound pressure p (t, r) on the spherical surface having a radius r from the origin of the acoustic signal radiated from the point sound source arranged at the origin of the three-dimensional space is determined by the wave equation of [Equation 1].

［数２］の周波数領域表現は、［数３］となる。

Therefore, if the acoustic signal radiated from the point sound source is s (t), the sound pressure p (t, r) on the spherical surface having the radius r from the origin is expressed by [Equation 2].

The frequency domain representation of [Equation 2] is [Equation 3].

In the above equation, the sound pressure at one sound receiving point of the pre-frequency domain panning sound signal S (ω) radiated from one point sound source is inversely proportional to the distance from the point sound source to the sound receiving point, and is received from the point sound source. The propagation time of the acoustic signal to the sound point is delayed, and the product of the sound pressure transfer function G _p (ω, r), which is a function of the distance from the point sound source to the sound receiving point, and the sound signal S (ω) before frequency domain panning. It is shown that.

  Since the distance from the point sound source to the sound receiving point and the propagation time of the acoustic signal are uniquely determined if the point sound source coordinates and the sound receiving point coordinates are determined in an arbitrary coordinate system, the sound pressure transfer function is also determined in an arbitrary coordinate system. If the point sound source coordinates and the sound receiving point coordinates are determined, they are uniquely determined.

［数４］を解いて、粒子速度ｖ（ｒ）は［数５］により表すことができる。

［数５］の周波数領域表現は［数６］となる。

If the particle velocity on the spherical surface with the radius r from the origin of the acoustic signal radiated from the point sound source arranged at the origin of the three-dimensional space is v (r, t), the equation of motion on the spherical surface with the radius r is 4].

Solving [Equation 4], the particle velocity v (r) can be expressed by [Equation 5].

The frequency domain representation of [Equation 5] is [Equation 6].

Therefore, the acoustic physical quantity vector P _k having both the sound pressure p (t, r) and the particle velocity v (t, r) at the sound receiving point k as components can be defined by [Equation 7].

The acoustic physical quantity vector P _k may include only one of the sound pressure p (t, r) or the particle velocity v (t, r) as a component.

Further, the acoustic physical quantity vector P _k is an instantaneous acoustic intensity II (t, r) that is a product of the sound pressure p (t, r) and the particle velocity v (t, r), or a time interval in which the instantaneous acoustic intensity is present. The sound intensity I (t, r), which is the integral value of

Here, the instantaneous sound intensity II (t, r) is defined by [Equation 8], and the sound intensity I (t, r) is defined by [Equation 9].

In the following embodiments, a case where the acoustic physical quantity vector P _k includes only the sound pressure p (t, r) as a component will be described.

  In the above description, the reproduced sound field is represented by polar coordinates, but in the following, the reproduced sound field is represented by orthogonal coordinates having an X axis, a Y axis, and a Z axis that are orthogonal to each other.

The virtual sound source position, which is the localization position of the pre-panning acoustic signal s _n (t) in the three-dimensional reproduction sound field, is represented by ξ ⁽ⁿ⁾ = [ξ ⁽ⁿ⁾ _x , ξ ⁽ⁿ⁾ _y , ξ ⁽ⁿ⁾ _z ] ^T (Assuming that n is an integer from 1 to N), assuming that the position of the sound receiving point is r = [r _x , r _y , r _z ] ^T , radiation from N sound source positions ξ ⁽ⁿ⁾ The pre-panning sound pressure vector _b P = [ _b P _x , _b P _y , _b P _z ] ^{T at} the sound receiving point position r of the frequency domain pre-panning acoustic signal S _n (ω) is expressed by [Equation 10]. Is done.

Here, if a fixed value sound source position xi] ⁽ⁿ⁾ pre-panning acoustic signal s _{n (t)} can be localized at the sound source position xi] ^(n), by the sound source position xi] a ⁽ⁿ⁾ and the time function value For example, the pre-panning acoustic signal s _n (t) can be moved at a predetermined direction and speed.

In the following description, for simplicity, it is assumed that the number of pre-panning acoustic signals s _n (t) is 1, that is, N = 1.

Further, M-number of positions zeta ⁽¹⁾ a predetermined, ζ ⁽²⁾ ··· ζ after panning the arranged speakers ^(M) audio signals _{q 1 (t), q 2} (t) ·· The post-panning sound pressure vector _a P = [ _a P _x , _a P _y , _a P _z ] ^{T at} the sound receiving point position r when q _M (t) is output is expressed by [Equation 11].

Here, frequency-domain panned acoustic signals Q ₁ (ω), Q ₂ (ω)... Q _M (ω) are determined so as to satisfy the following [Equation 12], and inversely transformed into the time domain. Output from the speaker, the post-panning acoustic signals q ₁ (t), q ₂ (t)... Q _M (t) obtained by panning the pre-panning acoustic signal s (t) are set to a predetermined arrangement position ζ ^{( 1)} , ζ ⁽²⁾ ... Ζ ^(M) can be reproduced by a speaker. However, in [Equation 10], N = 1.

Then, in order to determine the post-panning acoustic signals q ₁ (t), q ₂ (t)... Q _M (t) so that [Equation 12] holds, panning represented by [Equation 13] It is only necessary to obtain q ₁ (t), q ₂ (t)... Q _M (t) that minimizes the square error E between the previous sound pressure vector _b P and the post-panning sound pressure vector _a P.

ここで、［数１５］を使用して［数１４］を書き換えると、［数１６］となる。

When [Equation 10] and [Equation 11] are substituted into [Equation 13] and expanded, [Equation 14] is obtained.

Here, when [Equation 14] is rewritten using [Equation 15], [Equation 16] is obtained.

よって、二乗誤差Ｅを最小とするＱ（ω）は［数１８］となる。

Q (ω) that minimizes the square error E can be determined by [Equation 17] in which E is partially differentiated with respect to Q (ω) and set to zero.

Therefore, Q (ω) that minimizes the square error E is [Equation 18].

  FIG. 2 is a block diagram showing a hardware configuration of the three-dimensional panning apparatus 1 according to the present invention, and outputs an analog / digital (A / D) converter 21 that reads a pre-panning sound signal and a post-panning sound signal. A digital / analog (D / A) converter 22 that performs, a CPU 23 that executes a three-dimensional panning process, a memory 24 that stores a three-dimensional panning program, and peripheral devices for operating the three-dimensional panning apparatus are connected. An interface (I / F) 25 is connected to the bus 20.

  A display panel 26, a keyboard 27, and a mouse 28 are connected to the I / F 25. An operation panel can be applied instead of the display panel 26, the keyboard 27, and the mouse 28.

  That is, the three-dimensional panning apparatus of the present invention is configured by installing a three-dimensional panning program in a computer.

  Next, the operation of the three-dimensional panning apparatus of the present invention will be described.

  FIG. 3 is a flowchart of the three-dimensional panning program installed in the memory 24. First, the CPU 23 reads the pre-panning acoustic signal s (t) via the A / D converter 21 (step S31).

  The CPU 23 performs Fourier transform on the pre-panning acoustic signal s (t) to calculate the frequency-domain pre-panning acoustic signal S (ω) (step S32).

Next, the CPU 23 executes a panning routine (step S33) to calculate the frequency domain panned acoustic signals Q ₁ (ω), Q ₂ (ω)... Q _M (ω). Explained.

Further, the CPU 23 performs inverse Fourier transform on the frequency-domain panned acoustic signals Q ₁ (ω), Q ₂ (ω)... Q _M (ω) to perform the panned acoustic signals q ₁ (t), q ₂ (t )... Q _M (t) is calculated (step S34), and the panned acoustic signals q ₁ (t), q ₂ (t)... Q _M (t) ( The vector display q (t)) is output (step S35), and this routine is terminated.

FIG. 4 is a flowchart of the panning routine executed in step S33 of the three-dimensional panning program in FIG. Here, the sound source position that is the localization position of the pre-panning acoustic signal s _n (t) in the three-dimensional reproduction sound field is ξ = [ξ _x , ξ _y , ξ _z ] ^T , and the sound receiving point position is r = [r _x , R _y , r _z ] ^T. It is assumed that the number of sound sources is N.

First, the CPU 23 reads the sound receiving point position r = [r _x , r _y , r _z ] ^T (step S331).

Next, the CPU 23 generates information of the sound source position ξ = [ξ _x , ξ _y , ξ _z ] ^T (step S332). For example, a trajectory such that the sound source moves between a predetermined start point and end point may be calculated. It is also possible to input sound source position information from the outside.

  Next, the CPU 23 calculates the coefficient shown in [Equation 15] (step S333), further calculates the vector h and the matrix H shown in [Equation 16] (step S334), and the frequency domain from [Equation 18]. The post-panning acoustic signal Q (ω) is calculated (step S335), and this routine is finished.

  Here, as shown in FIG. 5, a case is assumed in which a sound image is localized by acoustic signals output from two speakers.

  The sound receiving point J is the origin of the XY coordinates, the left speaker SL is disposed at a position d away from the sound receiving point J at an angle of 30 degrees to the left of the Y axis, and the right speaker SR is from the sound receiving point J. It is arranged at an angle of 30 degrees to the right of the Y axis at a position d apart. The sound source SS is assumed to exist at a position away from the sound receiving point J by an angle θ with respect to the Y axis. Note that θ is 0 degree on the Y axis, and takes a positive value on the right side of the Y axis and a negative value on the left side of the Y axis.

Then, the coefficient of [Equation 15] becomes [Equation 19].

[Expression 16] becomes [Expression 20].

Thus, is calculated by the output to _{Q 1} left speaker SL by using the number 18] (omega) and the right speaker SR output _{Q 2} (omega) is [number 21].

  Here, if d = D, it corresponds to the panning by the well-known “tangent law”, and therefore, it also matches the panning by the vector base amplitude panning method disclosed in Non-Patent Document 1.

  As described above, the three-dimensional panning apparatus according to the present invention can handle the case of d ≠ D, and can provide a panning method that is an extension of the conventional “tangent law” and the vector-based amplitude panning method. It is.

(Second Embodiment)
The first embodiment targets one sound receiving point, but the second embodiment targets one reproduction region V.

That is, in the second embodiment, using the number 22 as the square error E of panning Maeoto圧vector _{b P} and panning Kooto圧vector _{a P.}

  In the above description, it is assumed that two speakers are arranged in the same plane, but it is obvious that the three-dimensional panning apparatus according to the present invention can be applied even when the speakers are arranged in a three-dimensional space.

  That is, according to the second embodiment, a sound image can be localized for a large number of listeners.

  In the first embodiment and the second embodiment described above, the pre-panning acoustic signal is Fourier-transformed into the frequency domain, and then the panning process is performed in the frequency domain, and the calculated post-frequency-domain panning sound The sound signal is calculated by performing inverse Fourier transform on the signal after panning, but the Fourier transform processing, downmix processing, and inverse Fourier transform processing are composed of delay elements and filters, and panning in the time domain is also possible. It will be apparent to those skilled in the art.

  As described above, the three-dimensional panning apparatus according to the present invention has the effect that the pre-panning acoustic signal can be three-dimensionally panned and reproduced from a speaker arranged at a predetermined position. It is effective as a signal processing device.

The block diagram which shows the structure of 1st Embodiment of the three-dimensional panning apparatus which concerns on this invention. The block diagram which shows the hardware constitutions of the three-dimensional panning apparatus which concerns on this invention Flowchart of a three-dimensional panning program executed by the panning apparatus according to the present invention Flowchart of panning routine executed by panning apparatus according to the present invention Layout of two speakers

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 3D panning apparatus 11 Pre-panning acoustic signal reading means 12 Post-panning acoustic signal output means 13 Conversion means 14 Localization position information generation means 15 Arrangement information storage means 16 Frequency domain post-panning acoustic signal generation means 17 Inverse conversion means

Claims (2)

A pre-panning acoustic signal reading means for reading a pre-panning acoustic signal which is a time function;
A post-panning acoustic signal output means for outputting a post-panning acoustic signal as a time function;
Conversion means for converting the pre-panning acoustic signal into a frequency domain pre-panning acoustic signal;
Localization position information generating means for generating localization position information to be added to the pre-panning acoustic signal;
Arrangement information storage means for storing arrangement information of the post-panning acoustic signal output means;
A frequency domain post-panning acoustic signal generating means for generating a frequency domain post-panning acoustic signal obtained by panning the frequency domain pre-panning acoustic signal based on the localization position information and the arrangement information;
Look including a reverse conversion means for inversely converting an acoustic signal after the frequency domain panning panning after the acoustic signal is a function of time,
The frequency domain post-panning acoustic signal generation means,
The post-panning acoustic physical quantity vector at the sound receiving point when the post-panning acoustic signal is emitted from the post-panning acoustic signal output means, and the pre-panning acoustic physical quantity vector at the sound receiving point when the pre-panning acoustic signal is emitted from the sound source A three-dimensional panning apparatus that generates an acoustic signal after frequency domain panning that is equal to a vector . A pre-panning acoustic signal reading means for reading a pre-panning acoustic signal which is a time function;
A post-panning acoustic signal output means for outputting a post-panning acoustic signal that is a time function;
Conversion means for converting the pre-panning acoustic signal into a frequency domain pre-panning acoustic signal;
Localization position information generating means for generating localization position information to be added to the pre-panning acoustic signal;
Arrangement information storage means for storing arrangement information of the post-panning acoustic signal output means;
A frequency domain post-panning acoustic signal generating means for generating a frequency domain post-panning acoustic signal obtained by panning the frequency domain pre-panning acoustic signal based on the localization position information and the arrangement information;
An inverse transform means for inversely transforming the frequency domain panned acoustic signal into a panned acoustic signal that is a time function;
The frequency domain post-panning acoustic signal generation means,
The post-panning acoustic physical quantity vector in the sound receiving area when the post-panning acoustic signal is emitted from the post-panning acoustic signal output means, and the pre-panning acoustic physical quantity vector in the sound receiving area when the pre-panning acoustic signal is emitted from the sound source A three-dimensional panning apparatus that generates an acoustic signal after frequency domain panning that is equal to a vector.

Images