JP2000324600A - Sound image localization device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize an excellent sound feeling by means of small arithmetic quantity and memory quantity and to adjust the strength of a sound image positioning effect outside the head by providing first filter parts for generating two L and R direct signals and the two second filter parts for generating L and R reflecting signals. SOLUTION: The first filter parts 303b to 303e of a filter part 350 execute two kinds of arithmetic processings for L and R (L<R) concerning inputted respective digital signals Ql(n), Qr(n)... and generate two direct signals for L and R. The second filter part 305a for L executes the arithmetic processing to the L direct signal to generate the L reflecting signal during outputs from the first filter parts 303b to 303e and generates the R reflecting signal with respect to the R direct signal among the outputs from the first filter pats. Multiplying equipment 302a to 302p multiplies the outputs from the prescribed filter parts by a prescribed multiplier and an adder 303a adds the outputs from the prescribed multiplying equipment.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound image localization apparatus for realizing a virtual speaker outside a head using headphones in, for example, an AV (audio / visual) device.

[0002]

2. Description of the Related Art In recent years, media and devices for handling multi-channel audio signals, such as DVDs, have become widespread. FIG. 6 is a diagram showing a speaker arrangement when a general multi-channel audio signal such as a DVD (Digital Versatile Disk) is reproduced by a speaker.
Playback speakers 1a, 1b, and 1c for outputting L, C, and R audio signals are arranged in three directions, left (Left), front (Center), and right (Right). Then, in two directions, left (Left) and right (Right), behind the listener 10, reproduction speakers 1d and 1e for outputting audio signals for surround SL and SR are provided. Are located. Also, HLl, HLr, HCl, HCr, HRl, H in FIG.
Rr, HSLl, HSLr, HSRr, and HSRl are speakers 1a,
The impulse responses from 1b, 1c, 1d and 1e to the left and right ears of the listener 10 are shown.

Conventionally, when such a multi-channel audio signal is reproduced by headphones, the audio signals L (n) and L (n) output from the reproduction speakers 1a to 1e are used.
Down-mixing of the five channels C (n), R (n), SL (n), and SR (n) into two-channel signals Lh (n) and Rh (n) was performed and obtained. A method of reproducing two-channel signals Lh (n) and Rh (n) with headphones is adopted.

Here, the downmix coefficients are represented by aF, aC,
aS, and if the multiplication is ×, the signals Lh (n) and Rh (n) of the two channels are

[0005]

(Equation 1) as well as

[0006]

(Equation 2) Is obtained by the following equation. However, in this method, the channel signal Lh (n) is a downmix coefficient corresponding to L (n) from the left front, C (n) from the front, and SL (n) from the left rear. aF, aC, and aS multiplied by each other, and the channel signal Rh (n) is R
(n), C (n) from the front, and SL (n) from the right rear are multiplied by the corresponding downmix coefficients aF, aC, and aS, respectively. Is generated on the line connecting the left and right ears in the head of the listener, thereby causing the listener 1
In the case of 0, the user feels that all sounds are sounding in his / her head (the sound image is localized in the head), and it is difficult to distinguish the position of the sound image by the sound signal, and a sense of spaciousness cannot be obtained. there were.

[0007] In order to solve such a problem, when the listener 10 listens to the headphones, the sound image is localized outside the head, that is, the listener 10 hears a three-dimensional sound as if he or she is listening to live performance. Attempts have been made to apply the processing to get the feeling. For example, there is a method in which an input signal is processed by a filter that approximates a spatial transmission method (impulse response) from a desired sound image position to a listener's both ears to reproduce the audio signal from headphones. .

FIG. 7 is a block diagram showing a configuration of a conventional sound image localization apparatus 200. The sound image localization device 200 is a one-channel analog signal S (n) generated from the signal source 201.
A / D converter 202 that converts the digital signal into a digital signal, and filter units 203a and 203b that adjust the amplitude and phase of the input digital signal for each frequency and generate a control signal for localizing a sound image at a virtual speaker position. And a D / A converter 204a, which converts control signals (digital signals) from the filter units 203a, 203b into analog signals.
204b, amplifiers 205a and 205b for amplifying and outputting the input analog signal, and audio signals yL (n) and yR.
(n).

The sound image localization apparatus 200 includes a filter coefficient recording unit 212 for recording the filter coefficients of the filter units 203a and 203b, a filter coefficient reading unit 213 for reading out the filter coefficients from the filter coefficient recording unit 212, Coefficient reading means 21
And control mode instructing means 214 for instructing the control mode for sound image localization to the filter coefficient reading means 2.
13 is configured to read out a filter coefficient from the filter coefficient recording means 212 based on an instruction from the control mode instruction means 214. Reference numeral 201 denotes a source that generates an analog signal, 207 denotes a virtual speaker, and 10 denotes a listener.

Next, the operation will be described. First, one-channel analog signal S (n) generated from signal source 201
Is input to the sound image localization apparatus 200. Sound image localization device 2
00, when the analog signal S (n) is input, the analog signal S (n) is converted into a digital signal Q by the A / D converter 202.
The converted digital signal is output to the filter units 203a and 203b.

When digital signals Q (n) are input to filter sections 203a and 203b, virtual speakers 207
The input digital signal Q (n) is adjusted in amplitude and phase for each frequency so that the sound image is localized at the position, and the control signals QL (n) and QR (n ) Is generated respectively.

At this time, the control signal QL (n) generated by the filter unit 203a is supplied to the D / A converter 204a, and the control signal QR (n) generated by the filter unit 203b is supplied to the D / A converter 204a.
/ A converter 204b.

In the D / A converter 204a, the filter unit 20
The control signal QL (n) input from 3a is an analog signal yL
(n), and in the D / A converter 204b, the filter unit 2
03b is the analog signal yR
(n), and the analog signal yL (n) is output to the amplifier 205a, and the analog signal yR (n) is output to the amplifier 205b.

In each of the amplifiers 205a and 205b, the analog signal yL (n) or yR (n) input from each of the D / A converters 204a and 204b is amplified and output to the headphone 206. signal
yL (n) and yR (n) are reproduced.

The filter units 203a and 203b
With respect to the filter coefficient, the control mode instructing means 214 issues an instruction of the sound image localization control mode to the filter coefficient reading means 213, and the filter coefficient reading means 213
The filter coefficients are read out from the filter coefficient recording means 212 based on the instruction from the control mode instruction means 214, so that the filter units 203a,
The filter coefficient of 3b is set.

The filter units 203a and 203b
Is performed so that when the listener 10 listens to the sound reproduced from the headphones 206, the listener 10 can hear the sound from the position of the virtual speaker 207.

More specifically, the impulse responses hL (n) and hR (n) from the headphone 206 to the binaural eardrum position of the listener 10 (or the entrance of the ear canal), and the binaural eardrum of the listener 10 from the virtual speaker 207 The impulse response dL (n) and dR (n) up to the position (or the entrance of the ear canal) are obtained in advance,
next

[0018]

(Equation 3) as well as

[0019]

(Equation 4) The filter coefficients xL and xR in the filter units 203a and 203b are determined so as to satisfy the following equation. Here, * indicates a convolution operation. Generally, these filter units 203
a and 203b are F calculated using the least squares method or the like.
This is realized by an IR (Finite Impulse Response) filter or the like. Using the filter coefficients xL and xR obtained in this way, the following

[0020]

(Equation 5) as well as

[0021]

(Equation 6) , YL (n) and yR (n) are calculated from the calculated audio signal yL.
By outputting (n) and yR (n) from the headphones, a sound image of the signal S (n) can be realized at the virtual speaker position.

[0022]

As described above, in order to realize the out-of-head sound image localization, it is necessary to show how the sound signal is transmitted in space from the position where the sound source is placed to each of the left and right ears. A transfer function (impulse response) is obtained in advance for each of the left and right directions, and is realized by convolving the obtained impulse response with an audio signal. To do this, the filter
You need one. Therefore, in order to realize a sound image at each virtual speaker position with respect to a multi-channel signal, a filter unit having twice the number of input signal channels is required. However, there is a problem that a large number of

In order to realize a sound image with high accuracy,
Since the number of filter coefficients (the number of taps) in the filter unit is required to be very large, if a large number of taps is used, a huge amount of calculation and a large amount of memory are required, and a sound image cannot be realized in real time. There is a point.

Further, when the effect of the above-mentioned out-of-head sound image localization is increased or decreased, it is realized by appropriately changing the filter coefficient. For example, a plurality of impulse responses from the virtual speaker to the listener's both ears at a plurality of locations having different reverberations are measured, and filter coefficients are obtained by the above (Equation 3) and (Equation 4). Is stored, and a predetermined filter coefficient is appropriately read from the stored plurality of filter coefficients, thereby changing the filter coefficient. However, this configuration has a problem in that a plurality of filter coefficients must be stored, and a large-capacity memory is required.

The present invention has been made to solve such a problem, and realizes an excellent sound spreading feeling with a small amount of calculation and a small amount of memory, and can adjust the strength of the effect of localization of an out-of-head sound image. It is an object to provide a sound image localization device.

[0026]

According to a first aspect of the present invention, there is provided a sound image localization apparatus, comprising: a filter coefficient recording unit for recording a filter coefficient; Filter coefficient reading means for reading the digital audio signal, and a control signal for localizing the sound image at the virtual speaker position based on the filter coefficient read from the filter coefficient reading means. A sound image localization apparatus comprising: a D / A converter for converting a control signal from the filter unit into an analog signal; an amplifier for amplifying an analog signal of the D / A converter; And headphones that radiate the amplified analog signal. A first filter unit that performs two types of arithmetic processing for L (Left) and R (Right) on the filter to generate two direct signals for L and R, and an output signal of the first filter unit , And two second filter units for generating reflected signals for L and R.

A sound image localization apparatus according to a second aspect of the present invention is the sound image signal localization apparatus according to the first aspect, further comprising control mode instructing means for instructing a control mode of the sound image localization, wherein the filter coefficient readout is performed. The means reads out a filter coefficient from the filter coefficient recording means based on an instruction from the control mode instruction means.

According to a third aspect of the present invention, in the sound image localization apparatus according to the first or second aspect, the second filter unit may include an IIR (Infinite Impulse).
e Response) filter.

A sound image localization apparatus according to a fourth aspect of the present invention is the sound image localization apparatus according to any one of the first to third aspects, wherein the second filter section includes an output signal of the first filter section. Among them, the second L filter unit for performing an arithmetic process on the direct signal for L to generate a reflected signal for L, and the direct signal for R among the output signals of the first filter unit And a second R filter for generating a reflected signal for R by performing an arithmetic operation on the output. The output of the second filter for L is multiplied by a multiplier having a predetermined multiplier, and the L Input to the second filter unit for R and to the second filter unit for R.
The output of the second R filter unit is multiplied by a multiplier having a predetermined multiplier, and is again input to the second R filter unit and input to the second L filter unit. There is a feature.

A sound image localization apparatus according to a fifth aspect of the present invention is the sound image localization apparatus according to the fourth aspect, wherein the multiplier of the multiplier is variable.

[0031]

Embodiments of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram for explaining a sound image localization apparatus 100 according to a first embodiment. The sound image localization apparatus 100 of the present invention is the same as the conventional sound image localization apparatus 200.
, Filter processing is performed on the 5-channel audio input signal S (n) input from the A / D converter 202 in place of the filter units 203a and 203b, and a sound image is formed at the positions of the virtual speakers 207a to 207e. A filter section 350 for generating two-channel control signals QL (n) and QR (n) to be localized is provided, and C (Center), L (Left), and R (Righ
t), two-channel control signals QL (n), Q generated by filtering from five-channel audio input signals S (n) for SL (Surround Left) and SR (Surround Right)
R (n) is converted into analog signals yL (n) and yR (n), amplified, and reproduced from the headphones 206, whereby the sound emitted from the headphones 206 is transmitted to the listener 10 This makes the user feel as if they were heard from the positions of the virtual speakers 207a to 207d.

The sound image localization apparatus 100 converts the 5-channel analog signals S (n) generated by the signal source 201a into digital signals (Qc (n), Ql (n), Qr (n), Qsl (n), Qsr (n)), an A / D converter 202 that converts the input digital signal into a filter, and performs a filtering process on each of the virtual speakers 207a to 207a.
Control signals QL (n) and QR (n) for localizing the sound image at the position 207d
And the control signals QL (n) and QR (n) from the filter unit 350 are converted into analog signals yL (n) and yR (n).
D / A converters 204a and 204b that convert the analog signals yL (n) and yR (n) into amplifiers 205a and 205b that amplify and output the analog signals yL (n) and yR (n).
and headphones 206 for reproducing analog signals yL (n) and yR (n) from b as audio signals.

The sound image localization apparatus 100 includes a filter coefficient recording unit 212 for recording a filter coefficient of the filter unit 350, and a filter coefficient reading unit 21 for reading out a filter coefficient from the filter coefficient recording unit 212.
And a control mode instructing means 214 for instructing the filter coefficient reading means 213 of a control mode of sound image localization. The filter coefficient reading means 213 The filter coefficient is read from the coefficient recording unit 212.

The signal source 201a is an AC-
In some cases, an audio digital signal such as a decoded output of 3 or DTS is used as a signal source. In such a case, the digital signal can be directly input to the filter unit 350 and used without using the A / D converter 202. Other configurations of the sound image localization device 100 include:
The configuration is the same as that of the conventional sound image localization apparatus 200.

FIG. 2 is a block diagram showing the configuration of the filter section 350. The filter unit 350 includes C, L,
Each audio digital signal Qc for R, SL, SR
(n), Ql (n), Qr (n), Qsl (n), and input terminals 300a to 300e for inputting Qsr (n), and the input digital signal for C Qc (n) And a C filter 301 for generating a center signal, and two L and R filters for each of the input digital signals Ql (n), Qr (n), Qsl (n), and Qsr (n). First filter sections 303b to 303e that perform two types of arithmetic processing to generate two direct signals for L and R, and operate on the direct signal for L among the outputs from the first filter sections 303b to 303e A second L filter unit 305a that performs processing to generate a reflected signal for L, and performs an arithmetic process on the direct signal for R among the outputs from the first filter units 303b to 303e to generate a reflected signal for R R that produces
Second filter unit 305b and multipliers 302a to 302p for multiplying an output from a predetermined filter unit by a predetermined multiplier.
And an adder 303a for adding an output from a predetermined multiplier
To 303f.

Here, each of the first filter sections 303b
In 303e to 303e, the arithmetic processing for the direct sound directly reaching the left and right ears of the listener from the respective signal sources of L, R, SL, and SR is performed for each signal for the left (L) and right (R). The second filter units 305a and 305b use the output from the first filter unit to reflect the sound image for an out-of-head feeling, a reverberation feeling, or a sound spreading feeling. Arithmetic processing is performed on the sound. In the C filter section 301, the C input signal is output to the output terminals 304e and 30e.
4f, the input signals for the other L, R, SL, and SR are output to the output terminal 3 after the first and second arithmetic processing.
An arithmetic process is performed to match the timing of output to 04e and 304f.

In the filter section 350, the multiplier 3
02a to 302p, the multiplier 302a is connected to the output of the C filter unit 301, and the multiplier 302b
To 302i are respectively connected to the respective outputs for L and R from the first filter units 303b to 303e. The multipliers 302p, 302l, and 302o are
The multipliers 302q, 302m, and 302n are connected to the output of the second filter unit 305a.
5b is connected to the output.

In the filter section 350, of the adders 303a to 303f, the adder 304a adds the outputs of the multipliers 302b to 302e.
The adder 304b adds the outputs of the multipliers 302f to 302i. Further, the multipliers 302j, 30
2k multiplies the outputs of the adders 304a and 304b by a predetermined multiplier.

The adders 304e and 304f are connected to the multipliers 302a, 302j and 302p or the multiplier 3
The output of the adder 304e is output as a control signal for L, and the output of the adder 304f is output as a control signal for R.

The adders 304c and 304d are connected to the multipliers 302l and 302m or the multipliers 302n and 302n.
02o, the output of the adder 304c is output to the second filter unit 305a as a feedback signal fb1, and the output of the adder 304d is output as the feedback signal fb2 to the second filter unit 305b. What is output.

That is, in the filter unit 350, the output of the second filter unit 305a is multiplied by a predetermined multiplier, and then input to the second filter unit 305a again.
The input to the second filter unit 305b, and similarly, the output of the second filter unit 305b is also multiplied by a predetermined multiplier, and then input to the second filter unit 305b again and input to the second filter unit 305a. It is configured to be.

The multipliers of the multipliers 302a to 302q are constants while the multipliers of the multipliers 302b to 302i are constants, but the multipliers of the multipliers 302a and 302l to 302q are variable and can be adjusted. it can.

FIG. 3 is an example of a block diagram showing the configuration of the first filter section 303.

The first filter section 303-1 has an input terminal 400 for inputting a digital signal and a filter coefficient bL _0.
FIR (Finite Impulse Response) for L having ~ bL _N-1
Filter and R FI having filter coefficients bR _{0 to} bR _N-1
An R filter and the FIR filter for L or the R
The output from the FIR filter for L to the signal SL for L or R
Output terminals 405a and 405b for outputting as the use signal SR
And the FIR filter for L and the FIR for R
The filters are arranged and arranged in parallel.

The L and R FIR filters delay the input signal by a predetermined time and output it.
403f, 403h to 403p, multipliers 402a to 402l for multiplying the outputs of the unit delay units 403b to 403f, 403k to 403p by predetermined coefficients, and outputs of the multipliers 402a to 402f, 402g to 402l. And adders 404a and 404b for adding.

The unit delay unit 40 of the FIR filter for L
3a to 403g are connected in series from the input terminal side 400, and the bDL-th unit delay unit 403b to the unit delay unit 40
The output of each unit delay unit up to 2f is output to the next connected unit delay unit, and the multipliers 402a to 402f
It is configured to be output to

The unit delay units 403h to 403p of the R FIR filter are connected in series from the input terminal side 400, and the outputs from the bDR-th unit delay unit 403k to the unit delay unit 403p are It is configured to output to the connected unit delay device and output to the multipliers 402g to 402l.

Each of the multipliers 402a to 402f,
The outputs from 402g to 402l are output to adders 404a or 404b, and the outputs from adders 404a are
As a direct signal for L, the output from adder 404b is
It is configured to be output as a direct signal for R.

Note that bDL and bDR are delay amounts of a portion obtained by subtracting a common delay existing in each path from an initial delay existing in an impulse response from a plurality of virtual sound sources to be realized to both ears. That is, in the first filter unit 303-1,
It is configured such that a direct signal is generated by removing the common delay existing in each path from the initial delay existing in the impulse response from a plurality of virtual sound sources to be realized to both ears.

The multipliers 402a to 402f, 4
02g to 402l are predetermined multipliers bL _{0 to} bL _N−1 , respectively.
bR _{0 to} bR _N−1 are provided as filter coefficients. The above N
Is the number of filter coefficients (the number of taps). In the first embodiment, the number of taps N of the L FIR filter and the number of taps N of the R filter are described as being equal, but different values may be set. it can.

FIG. 4 shows a first filter section 303 in FIG. 3 in which the unit delay units of the L and R FIR filters are shared by the L and R filters. 9 is a second example of the filter unit 303. In the first filter unit 303-2 in FIG. 4, the unit delay units 502a to 502i
Are common for L and R. First filter unit 3
In 03-2, the multipliers 503a to 503i for L include unit delay units 502b to 50g from the bDL-th unit delay unit 502b to the unit delay unit 502g from the input terminal 501 side.
2g output is obtained and multiplied by a predetermined multiplier, respectively.
The adder 504a is configured to combine one direct signal for L. Further, multipliers 503g to 503 for R
l is the bDR-th unit delay unit 50 from the input terminal 501 side
Each unit delay unit 502 from 2d to unit delay unit 502i
The outputs of d to 502i are obtained, multiplied by predetermined multipliers, and added to one direct signal for R by an adder 504b.

Each of the bDL and bDR is a delay amount obtained by subtracting a common delay existing in each path from an initial delay existing in an impulse response from a plurality of virtual sound sources to be realized to both ears. That is, the first filter unit 303-2 is configured to generate a direct signal by removing the common delay existing in each path from the initial delay existing in the impulse response from a plurality of virtual sound sources to be realized to both ears. Have been.

Each of the L and R multipliers 503a
５０503f, 503g〜503l are predetermined multipliers bR ₀ bb
R _N−1 is provided as each filter coefficient. N is the tap length (number) of each FIR filter coefficient.
It is.

The first filter units 303-1 and 303-2 shown in FIGS.
It is possible to use a filter that realizes the first half of the filter coefficient obtained by (Equation 4), and it can be realized by a very low-order FIR filter of the order of about 30 or an IIR filter.

FIG. 5 is a block diagram showing the configuration of the second filter unit 305. In the second filter unit 305, the output HL from the first filter units 302b to 302e is output.
l, HRl, HSLl, HSRl and the feedback signal fb1, or the output HL from the first filter units 302b to 302e.
r, HRr, HSLr, HSRr and the feedback signal fb2 are input to control the processing for generating the second half of the impulse response of the filter to be realized, that is, to control the feeling outside the head, the reverberation, and the feeling of the spread of the sound image. Processing is performed.

In the second filter unit 305, the first filter for L
Input terminal for inputting output HLl or HLr of the filter unit 303b
To the output HRl of the first filter section 303c for R
Or an input terminal c for inputting HRr, and a first filter unit for SL.
Input terminal 60 for inputting output HSLl or HSLr of 303d
0d and the output HSRl of the SR first filter unit 303e or
Is the input terminal 600d for inputting HSRr and the feedback
Input terminal 600f for inputting signal fb1 or fb2
And the output HLl or HLr of the first filter unit 303b for L
For a given multiplier rl₀~ Rl₆601a to multiply by
601g, the output HRl of the first R filter unit 303c,
Is a given multiplier rr for HRr₀~ Rr₆Multiplier 6 to multiply by
01h to 601n and the SL first filter unit 303d.
A predetermined multiplier rsl for the output HSLl or HSLr₀~ Rsl₆Squared
Multipliers 601o to 601u to be calculated and a first fill for SR
A predetermined multiplier for the output HSRl or HSRr of the
rsr₀~ Rsr ₆Multipliers 601v to 601b 'for multiplying by
A predetermined number of unit delay units are connected, and each of the multipliers 601a is connected.
601b 'introduces the output from the delay line 602;
Output terminal 6 for outputting the generated reflection signal DHL or DHR
03, and the first filter units 303b to 303e
These outputs are multiplied by a constant, and this constant multiplication is delayed by a predetermined time.
Output so that the reflected sound is approximated.
Has been established.

Next, the operation will be described. First, a 5-channel analog signal S (n) generated from the signal source 201a
Is input to the sound image localization apparatus 100. Sound image localization device 1
At 00, when the analog signal S (n) is input, the A / D converter 202 converts the 5-channel analog signal S (n) to 5
Channel digital signal Qc (n), Ql (n), Qr (n), Qsl
(n), Qsr (n), and the converted digital signal is
Output to the filter unit.

In the filter section 350, the digital signal Qc
When (n), Ql (n), Qr (n), Qsl (n), and Qsr (n) are input, the input digital signal Q is input so that the sound image is localized at the positions of the virtual speakers 207a to 207e. For (n), the amplitude and phase are adjusted for each frequency, and control signals QL (n) and QR (n) for L and R are generated, respectively. At this time, the control signal QL (n) generated by the filter unit 203a is
The control signal QR (n) generated by the A converter 204a and generated by the filter unit 203b is output to the D / A converter 204b.

In the D / A converter 204a, the filter unit 20
The control signal QL (n) input from 3a is an analog signal yL
(n), and in the D / A converter 204b, the filter unit 2
03b is the analog signal yR
(n), and the analog signal yL (n) is output to the amplifier 205a, and the analog signal yR (n) is output to the amplifier 205b.

In each of the amplifiers 205a and 205b, the analog signal yL (n) or yR (n) input from each of the D / A converters 204a and 204b is amplified and output to the headphone 209. yL
(n) and yR (n) are reproduced.

The filter units 203a and 203b
With respect to the filter coefficient, the control mode instructing means 214 issues an instruction of the sound image localization control mode to the filter coefficient reading means 213, and the filter coefficient reading means 213
The filter coefficients are read out from the filter coefficient recording means 212 based on the instruction from the control mode instruction means 214, so that the filter units 203a,
The filter coefficient of 3b is set.

Next, the operation of the filter section 350 will be described. First, the digital signals Qc (n), Ql (n), and Qr are input from the A / D converter 202 to the input terminals 300a to 300e.
(n), Qsl (n), and Qsr (n) are input.

In the C filter 301, predetermined arithmetic processing is performed on the input C digital input signal Qc (n), and the generated center signal Hc is output to the multiplier 302a. Specifically, the input signal for C is output terminal 3
04e and 304f are output to other L,
The input signals for R, SL, and SR are adjusted to match the timing at which a direct signal and a reflected signal are generated by arithmetic processing and output to the output terminals 304e and 304f.
In the multiplier 302a, the output Hc from the C filter unit 301 is further multiplied by a predetermined multiplier and adjusted. The output aHc of the multiplier 302a is added to adders 304e and 3
04b.

In the first filter section 303b, the L and R digital input signals Ql (n) are
Two types of arithmetic processing are performed, direct signals for L and R
HLl and HLr are output. At this time, of the output of the first filter unit 303b, the direct signal HL1 for L is output to the second filter unit 305a for L and the multiplier 3
02b. Further, the first filter unit 30
3b, the direct signal HLr for R is output to the second filter unit 305b for R and the multiplier 302
output to f.

Similarly, the first filter units 303c to 303c
At 303e, each input digital input signal Qr
(n), Qsl (n) and Qsr (n) are subjected to two types of arithmetic processing for L and R, and the direct signals HRl, HSLl and HSRl for L are
Each is output to the second L filter unit 305a, and is output to each of the multipliers 302c, 302d, and 302e. The direct signals HRr, HSLr, and HSRr for R are each output to the second R filter unit 305b. Is output to each of the multipliers 302g, 302h, and 302i.

The multipliers 302b to 302e multiply the respective direct L signals HLl, HRl, HSLl, HSRl inputted from the first filter sections 303b to 303e by a predetermined multiplier. Adder 304a
Is output to the adder 304a.
The outputs bHLl, cHRl, dHSLl, and eHSRl from to 302e are combined into one L signal QL and output to the multiplier 302j. In the multiplier 302j, the L signal QL from the adder 304a is multiplied by a predetermined multiplier and output to the adder 304e.

Similarly, in the multipliers 302f to 302i,
The R direct signals HLr, HRr, HSLr, and HSRr input from the first filter units 303b to 303e are respectively multiplied by a predetermined multiplier, and are added to the adder 304.
b, and output to the multiplier 302b at the adder 304b.
Outputs bHLr, cHRr, dHSLr, eHSRr from f to 302i are 1
The signals are combined into one R signal QR and output to the multiplier 302k. In multiplier 302k, R from adder 304b
Signal QR is multiplied by a predetermined multiplier to adder 304f
Output to

On the other hand, in the second L filter section 305a,
L input from the first filter units 303b to 303e
The direct signals HLl, HRl, HSLl, and HSRl are subjected to arithmetic processing for multiplying them by a constant.
The reflection signal DHL for L is generated using b1. The reflected signal DHL for L is output to multipliers 302p, 302l, 30
Output to 2o. Upon receiving the reflection signal DHL for L, the multiplier 302p multiplies the DHL by a predetermined multiplier and outputs the result to the adder 304e. The adders 304e output the outputs aQcc, jQ from the multipliers 302a, 302j, 302p.
L and pDHL are combined into one and output from the output terminal 306a as a control signal QL (n) for L.

Similarly, the second R filter unit 305b performs an arithmetic process for multiplying the R direct signals HLr, HRr, HSLr, and HSRr input from the first filter units 303b to 303e by a constant. The reflection signal DHR for R is generated using the constant multiple value and the feedback signal fb2.
The reflected signal DHR for R is output to multipliers 302q, 302m,
302n. When the reflection signal DHR for R is input to the multiplier 302q, the DHR is multiplied by a predetermined multiplier and output to the adder 304f. In the adder 304f,
The outputs aQcc from the multipliers 302a, 302k, 302q,
kQR and qDHR are combined into one and output from the output terminal 306b as an R control signal QR (n).

The output DHL from the second filter unit 305a is input to the multipliers 302l and 302o, and the output DHR from the second filter unit 305b is input to the multipliers 302m and 302n. In each of the multipliers, the output from each of the input filter units is multiplied by a predetermined multiplier, and an adder 304
c or 304d.

In the adder 304c, the outputs of the multipliers 302l and 302m are combined into one feedback signal fb1, and the feedback signal fb1 is output to the second L filter section 305a. In the adder 304d, the multiplier 302n and the The output of the multiplier 302o is combined into one feedback signal fb2, and the feedback signal fb2
2 is output to the second R filter unit 305b.

Next, referring to FIG. 3, the first filter unit 3
The operation of 03-1 will be described. Here, a case where L digital signal Ql (n) is input to first filter section 303-1 will be described as an example. First, the digital signal Ql
(n) from the input terminal 400, the FIR filter for L and the R
Input to the FIR filter.

In the L FIR filter, the digital signal
When Ql (n) is input, bDL unit delay units 403a to 403a-4
According to 03b, the input digital signal Ql (n) is output with a predetermined time delay for each unit delay unit.
The amount of delay by the bDL unit delay units 403a to 403b is obtained by subtracting the common delay existing in each path from the initial delay existing in the impulse response from a plurality of virtual sound sources to be realized to both ears. Minutes delay.

The output of the unit delay unit 403b is output to the unit delay unit 403c to be connected next and to the multiplier 402a. Similarly, the unit delay unit 4
03c to the output of the unit delay unit 403g are output to the next connected unit delay unit, and the multipliers 402b to
Output to 402f.

In each of the multipliers 402b to 402f, the input signal is multiplied by a predetermined multiplier.
4a. In the adder 404a, the multiplier 4
Outputs from 02b to 402f are combined into one L signal and output from an output terminal 405a. In addition, F for R
In the IR filter, the same processing as that of the L FIR filter is performed on the input digital signal Ql (n).

Next, with reference to FIG.
The operation of 3-2 will be described. Here, a case where L digital signal Ql (n) is input to first filter section 303-2 will be described as an example. In the first filter unit 303-2, when the digital signal Ql (n) is input from the input terminal 501, the unit delay units 502a to 502i delay the input signal by a predetermined time, and Output to the container. At this time, in each of the unit delay units 502c to 502d, the input signal is output to the next connected unit delay unit with a predetermined time delay, and the L multiplier 50
3a to 503b. Also, 502d
In each of the unit delay units て to 502 g, the input signal is output to the next connected unit delay unit with a delay of a predetermined time, and to the L multipliers 503 c to 503 f and the R multipliers 503 g to 503 j. Each is output. Also, 50
In the unit delay units 2h to 502i, the input signal is output to the next connected unit delay unit with a delay of a predetermined time and output to the R multipliers 503k to 503l.

In each of the L multipliers 503a to 503f, the output of each of the unit delay units 502b to 502g is multiplied by a predetermined multiplier and output to the adder 504a. In the R multipliers 503g to 503l, the outputs of the unit delay units 502d to 502i are respectively multiplied by predetermined multipliers and output to the adder 504b.

In the adders 504a and 504b, each of the multipliers 503a to 503f or 503g
To 503l are combined into one L or R direct signal HLl, HLr, respectively, and output terminals 505a, 505
b.

Next, referring to FIG. 5, the second filter unit 30
Operation 5 will be described. Here, the operation of the second filter unit 305a will be described as an example. First filter section 303
Outputs HLl, HRl, HSLl, and HSRl from b to 303e are applied to input terminals 600b to 600e and to a feedback signal f, respectively.
b1 is input to the input terminal 600f. The feedback signal fb1 is input from the input terminal 600f to the delay line 602.

The first filter units 303b to 303e
The output HLl, HRl, HSLl, HSRl of each input terminal 600b ~
600e, each of the 1 multipliers 601a-6
01g, r multipliers 601h-601n, sl multipliers 601o-601u, sr multipliers 601v-601
At b ′, predetermined multipliers rl _{0 to} rl ₆ , rr _{0 to} rr ₆ , rs
is multiplied by _{l 0 ~rsl 6, rsr 0 ~rsr} 6, is input to the delay line 602.

In the delay line 602, each unit delay unit on the delay line 602 outputs the feedback signal fb1 input from the input terminal 600f to the next unit delay unit with a delay of a predetermined time, Vessels 601a-60
1g, 601h to 601n, 601o to 601u, 60
The outputs from 1v to 601b are introduced, a reflection signal DHL is generated, and output from the output terminal 603. That is, in the second filter unit 305a, the first filter units 303b to 30b
The direct signals HLl, HRl, HSLl, HSRl output from 3e are
Each of the multipliers 601a to 601g, 601h to 601n, 6
01o to 601u and 601v to 601b are multiplied by a constant,
By introducing this into the delay line 602, the reflected signal DH
L is generated.

As described above, the sound image localization apparatus 100 according to the first embodiment has the five-channel audio input signal S
(n) is provided with a filter unit 350 that performs filter processing for localizing a sound image at the positions of the virtual speakers 207a to 207e and generates two-channel control signals QL (n) and QR (n). One filter section 350 can localize each sound image at a plurality of virtual speaker positions with respect to audio input signals S (n) of a plurality of channels.

The filter section 350 is a first filter section 3 for performing an arithmetic operation on the input digital signal.
03b to 303e, and second filter units 305a and 305a that perform arithmetic processing on the output signal of the first filter unit.
5b, and the first filter units 303b to 303e.
In, the direct sound is generated except for the common delay present in each path among the initial delays present in the impulse responses from the multiple virtual sound sources to be realized to both ears, so that the amount of memory required for the calculation is reduced. Also, the second filter units 305a and 305b can
Since the reflected sound is generated by delaying the direct sound from 〜303e by a constant, the number of filter coefficients (the number of taps) can be reduced.

Also, the reflected sound of the sound emitted from the actual speaker reaches both the left and right ears, and
The two filter units 305a and 305b are configured to cause the outputs DHL and DHR of the second filter units 305a and 305b to be input again to the second filter units 305a and 305b and to be input to the second filter units 305b and 305a, respectively. Therefore, an apparently long impulse response can be realized, and more natural and spacious control can be performed with a small amount of calculation.

In the filter section 350, each multiplier can change the multiplier set in each multiplier based on an instruction from the control mode instructing means 214 for instructing a control mode of sound image localization. Therefore, it is possible to add the strength of the reverberation by changing the coefficient without changing the filter configuration, and to adjust the feeling outside the head and the feeling of the spread. For example, in a mode in which reverberation is strong (hereinafter, strong mode), the coefficients (feedback coefficients) of the multipliers 302l to 302o and the multiplier 32 of the first filter unit output are used.
0j, 302k, and multiplier 3 of the second filter unit
The values of the coefficients 02p and 302q can be realized by changing the settings so that the filter coefficients of the respective paths approximate long impulse responses. Also, the second
By setting the feedback coefficients of the filter units 305a and 305b to 0, a mode with less reverberation than in the strong mode can be realized. Also, the second filter unit 3
Multiplier 302 for outputs DHL and DHR of 05a and 305b
By setting the multiplier of p and 302q to 0, the operations of the second filter units 305a and 305b are not performed, so that the processing for the reflected sound is not performed, and only the minimum necessary processing is performed.

In the filter unit 350, the first filter units 303b to 303e and the second filter units 305a and 305a
Multipliers 302b-302 for each output
Since i, 302p, and 302q are provided to enable individual level adjustments, the coefficient can be changed so that the sound volume does not change audibly in each mode having different effects. For example, in the weak mode, only the first filter portions 303b to 303e are processed, so that the energy of the processed signal is smaller than in the strong mode, and a volume difference occurs when the mode is switched. Therefore, for example, the second filter units 305a and 305b
Are not performed, the multipliers 302b to 302i
Is adjusted, the second filter unit 305
It is possible to prevent a volume difference from occurring when the processing of a and 305b is performed. Similarly, the 3 mentioned above
By adjusting the value of the multiplier for each of the examples of the two modes, it is possible to eliminate the volume difference between the modes.

In the filter section 350, a multiplier is
Since the first filter units 302b to 302e and the output units of the second filter units 305a and 305b are installed, the quantization error of the output signal can be reduced.
For example, when the filter coefficient of the second filter unit has a small value as compared with the filter coefficient of the first filter unit, the filter coefficient is normalized and set so as not to overflow during calculation. However, by performing multiplication by a constant at the time of output and adding, it is possible to reduce quantization errors that occur when performing filter processing.

Note that the C filter section 301 is
Although it can be realized by an R filter, an IIR filter, or the like, a configuration in which the arithmetic processing is performed by the first filter unit and the second filter unit as in the case of other input signals such as L, R, SL, and SR can also be employed. .

[0089]

According to the sound image localization apparatus according to the first aspect of the present invention, the filter coefficient recording means for recording the filter coefficient, the filter coefficient reading means for reading out the filter coefficient from the filter coefficient recording means, and the filter A sound image localization apparatus comprising: a filter unit that performs a filtering process on an input digital acoustic signal based on a filter coefficient read from a coefficient reading unit, and generates a control signal for localizing a sound image at a virtual speaker position. , A D / A converter for converting a control signal from the filter unit into an analog signal, an amplifier for amplifying the analog signal of the D / A converter, and headphones for radiating the analog signal amplified by the amplifier , And the filter unit applies L (Lef) to the input digital audio signal.
t) and R (Right) are processed, and L and R
A first filter unit that generates two direct signals for L and R, and two second filter units that perform an arithmetic process on the output of the first filter unit and generate reflected signals for L and R. With this configuration, a single filter unit can localize each sound image at a plurality of virtual speaker positions with respect to digital input signals S (n) of a plurality of channels, with a small amount of computation and a small amount of memory. The effect is that the amount of sound can be more widespread than before.

A sound image localization apparatus according to a second aspect of the present invention is the sound image signal localization apparatus according to the first aspect, further comprising control mode instructing means for instructing a control mode of sound image localization, wherein the filter coefficient reading is performed. The means reads out the filter coefficient from the filter coefficient recording means based on the instruction from the control mode instruction means. Therefore, it is possible to adjust the strength of the effect only by changing the coefficient without changing the filter configuration. This makes it possible to adjust the feeling outside the head and the feeling of spreading.

According to a third aspect of the present invention, there is provided the sound image localization apparatus according to the first or second aspect, wherein the second filter unit is provided with an IIR (Infinite Impulse).
e Response) filter, so that an excellent sound spreading feeling can be realized.

According to a fourth aspect of the present invention, in the sound image localization apparatus according to any one of the first to third aspects, the second filter section includes an output signal of the first filter section. Among them, the second L filter unit for performing an arithmetic process on the direct signal for L to generate a reflected signal for L, and the direct signal for R among the output signals of the first filter unit And a second R filter for generating a reflection signal for R. The output of the second filter for L is multiplied by a multiplier having a predetermined multiplier, and The signal is input to the second filter unit for L and is also input to the second filter unit for R.
The output of the second R filter unit is multiplied by a multiplier having a predetermined multiplier, and is again input to the second R filter unit and input to the second L filter unit. Therefore, there is an effect that the number of filter coefficients (the number of taps) can be reduced while realizing a feeling of sound spreading more than before.

Further, in the sound image localization apparatus according to the fifth aspect of the present invention, since the multiplier of the multiplier is variable in the sound image localization apparatus according to the fourth aspect, only the coefficient is changed. This has the effect that the strength of the effect can be adjusted without changing the filter configuration.

[Brief description of the drawings]

FIG. 1 is a sound image localization apparatus 1 according to a first embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of a 00.

FIG. 2 shows a filter section 350 according to the first embodiment of the present invention.
FIG. 2 is a block diagram for explaining the configuration of FIG.

FIG. 3 shows a first filter unit 3 according to the first embodiment of the present invention.
FIG. 3 is a block diagram for explaining a configuration of No. 03-1.

FIG. 4 shows a first filter unit 3 according to the first embodiment of the present invention.
It is a block diagram for demonstrating the structure of 03-2.

FIG. 5 shows a second filter unit 3 according to the first embodiment of the present invention.
FIG. 5 is a block diagram for explaining a configuration of the power supply unit 05.

FIG. 6 is a diagram for explaining an installation position of a conventional multi-channel speaker.

FIG. 7 is a block diagram showing a configuration of a conventional sound image localization apparatus 200.

[Explanation of symbols]

1a-1e Speaker 10 Listener 201, 201a Signal source 200 Sound image localization device 202 A / D converter 203a, 203b Filter unit 204a, 204b D / A converter 205a, 205b Amplifier 206 Headphone 207, 207a-207e Virtual speaker 212 Control Mode instructing means 213 Filter coefficient reading means 214 Filter recording means 350 Filter section 300a to 300e, 400, 501, 600b to 60
0f input terminal 301C filter section 302a-302q, 402a-402l, 503a-
503l, 601a to 601b 'Multipliers 303b to 303e First filter units 304a to 304f, 404a to 404b, 504a to
504b Adders 305a, 305b Second filter units 306a-306b, 405a-405b, 505a-
505b, 603 Output terminals 403a-403p, 502a-502i Delay unit 602 Delay line

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Shoji Matsumoto 1006 Kazuma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Inventor Masahiro Sueyoshi 1006 Kadoma Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co. 72) Inventor Akihisa Kawamura 1006 Kadoma, Kazuma, Osaka Pref. Matsushita Electric Industrial Co., Ltd. (72) Inventor Yoshinori Kumamoto 1006 Odaka, Kadoma, Kadoma, Osaka Pref. 1006 Kadoma Kadoma, Matsushita Electric Industrial Co., Ltd. (72) Inventor Kosuke Nishio 1006 Kadoma, Kadoma, Osaka Pref. Matsushita Electric Industrial Co., Ltd. F term (reference) 5D062 AA26 AA71 AA74

Claims (5)

[Claims] A filter coefficient recording unit that records a filter coefficient; a filter coefficient reading unit that reads a filter coefficient from the filter coefficient recording unit; and a filter coefficient input based on the filter coefficient read from the filter coefficient reading unit. A filter unit for performing a filtering process on the digital acoustic signal and generating a control signal for localizing the sound image at the position of the virtual speaker. A filter for converting the control signal from the filter unit into an analog signal. A / A converter, an amplifier for amplifying an analog signal of the D / A converter, and headphones for radiating the analog signal amplified by the amplifier, wherein the filter unit converts the input digital audio signal into For L (Left) and R (r
ight) for two types of arithmetic processing to generate two direct signals for L and R, and an arithmetic process for the output signal of the first filter unit to generate two direct signals. A sound image localization apparatus, comprising: two second filter units that generate a reflection signal for R. 2. The sound image signal localization apparatus according to claim 1, further comprising control mode instruction means for instructing a control mode of sound image localization, wherein said filter coefficient reading means includes: A sound image localization device, wherein a filter coefficient is read from the filter coefficient recording means. 3. The sound image localization device according to claim 1, wherein the second filter unit is configured by an IIR (Infinite Impulse Response) filter. 4. The sound image localization device according to claim 1, wherein the second filter unit performs an arithmetic process on a direct signal for L in an output of the first filter unit, The second for L that generates the reflected signal for L
A filter unit, an R second second signal for performing an arithmetic process on the direct signal for R among the outputs of the first filter unit and generating a reflected signal for R;
And an output of the second L filter unit is multiplied by a multiplier having a predetermined multiplier, and is again input to the second L filter unit.
The output of the second filter unit for R is multiplied by a multiplier having a predetermined multiplier, and is again input to the second filter unit for R. Second
A sound image localization device, which is input to a filter unit. 5. The sound image localization device according to claim 4, wherein a multiplier of the multiplier is variable.