JP2870422B2 - Filter setting method for sound image localization device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for setting a filter of a sound image localization apparatus capable of localizing an imaginary sound image at an arbitrary position in an AV (audio or visual) device.

[0002]

2. Description of the Related Art In recent years, in the field of multimedia centering on video and audio, devices capable of reproducing three-dimensional images have been developed and put into practical use, and accordingly, development of three-dimensional sound reproducing devices has been desired. .

FIG. 4 is a block diagram showing a conventional sound image localization apparatus. In FIG. 4, 1 is an analog input signal source, and 2 is an A that converts the input signal 1 into a digital signal.
The / D converters, 3L and 3R are filters for realizing sound image localization, and the output signal of the A / D converter 2 is input. 4
L and 4R are D / A converters for the output signals of the filters 3L and 3R, and are amplifiers for amplifying the signals with the same gain.
L and 5R are acoustic transducers for converting the signal amplified by the amplifier into sound waves, and a speaker is used in this conventional example. 6
Is a listener who listens on the center line of the speakers 5L and 5R.
Reference numeral 7 denotes a position of a target virtual sound image.

A description will be given of a case where the virtual sound image 7 is realized by the sound image localization apparatus configured as described above.

When a transfer function from the speaker 5L to the right ear of the listener 6 is expressed as a discrete digital signal, HL
R (n), the transfer function from the left ear to the left ear is HLL (n), the transfer function from the speaker 5R to the right ear of the listener 6 is HRR (n), and the transfer function from the left ear to the left ear is HRL (n). I do.

The transfer function from the virtual sound image 7 to the right ear of the listener 6 is KL (n), and the transfer function from the right ear to the right ear is KR.
(N).

When the input signal 1 is given as S (n) and output from the virtual sound image 7, the sound reaching the ears of the listener 6 is represented as follows.

The sound pressure L (n) reaching the left ear is: L (n) = S (n) * KL (n) (1) The sound pressure R (t) reaching the right ear is R (n) = S (n) * KR (n) (2) where *: convolution operation Similarly, sound in which signal S (n) reaches both ears of listener 6 from speakers 5L and 5R. Is shown as follows.

The sound pressure L '(n) reaching the left ear is L' (n) = S (n) * (HLL (n) + HRL (n)) * FL (n) (3) The sound pressure R '(t) reaching the right ear is as follows: R' (n) = S (n) * (HLR (n) + HRR (n)) * FR (n) (4)

At this time, if the head related transfer functions are equal to each other, it is generally assumed that sounds can be heard from the same direction (this assumption is generally correct).

Then, in order for the listener 6 to radiate sound waves from the speakers 5L and 5R and perceive the sound image at the position of the virtual sound image 7,
The following equation should be satisfied.

L (n) = L '(n) (5) R (n) = R' (n) (6) From the above equations (1) to (6), FL (n) = KL (n) / (HLL (n) + HRL (n)) ・ ・ ・ (7) FR (n) = KR (n) / (HLR (n) + HRR (n)) ・ ・ ・ (8) Is obtained.

That is, if FL (n) given by equation (7) is set for the filter 3L shown in FIG. 4 and FR (n) given by equation (8) is set for the filter 3R, a desired position is obtained. It is possible to localize the virtual sound image 7.

For example, FIG. 5 is a diagram showing set values of filters used in a sound image localization apparatus and set by a conventional filter setting method. SFL is the data set in the filter 3L shown in FIG. 4, SFR is the data set in the filter 3R shown in FIG. 4, the horizontal axis represents the time axis, and the vertical axis represents the amplitude value. .

As shown in FIG. 5, when the tap length can be made sufficiently long (128 taps in this example), it is possible to realize a clear sound image localization of a virtual sound image.

[0016]

However, in the sound image localization apparatus shown in FIG. 4, when the tap length of the filter cannot be made large as shown in FIG. 6 (32 taps in the example of FIG. 6) due to the limitation of the hardware scale. Is the S shown in FIG.
Since it is not possible to provide a filter such as FL with a data sequence required for realizing sound image localization, sound image localization has been difficult.

Further, as shown in FIG. 7, when the difference between the integral values of SFL and SFR is large, there is also a problem that the sound image localization becomes difficult. In FIG. 7, the horizontal axis represents the time axis, the vertical axis represents the amplitude value, and the black circles represent filter setting data.

An object of the present invention is to improve the above-mentioned conventional problems and to provide a filter setting method for realizing a sufficient sound image localization even when the hardware scale of the sound image localization device cannot be large. It is assumed that.

[0019]

According to a first aspect of the present invention, there is provided a filter setting method comprising: an A / D converter for converting an input signal into a digital signal; and an output of the A / D converter. First and second filters to which signals are input, first and second amplifiers for amplifying the signals by D / A converting the output signals of the first and second filters, respectively, and the first and second amplifiers In a sound image localization device including first and second acoustic transducers each converting an output signal of an amplifier into a sound wave, the acoustic transducer can realize the input signal at an arbitrary position as an imaginary sound image. A first, sufficiently long tap length prepared for the first and second filters
And when the second data string is set in the first and second filters, between a first predetermined value and a second predetermined value,
The data up to the point where the axis crosses the zero axis are the first and second data.
The first and second filters are configured by using a data string of (1) and thereafter (0).

The filter setting method according to the second invention is characterized in that:
An A / D converter for converting an input signal into a digital signal;
First and second receiving the output signal of the A / D converter
A filter, first and second amplifiers for D / A converting the output signals of the first and second filters, respectively, and amplifying the signals, and converting output signals of the first and second amplifiers to sound waves, respectively. In the sound image localization device including the first and second sound transducers, the sound transducer is provided for the first and second filters, which can realize the input signal as an imaginary sound image at an arbitrary position. When the first and second data strings having sufficiently long tap lengths are set in the first and second filters, the first and second data strings are used within a predetermined range. , Otherwise 0
Is used to constitute the first and second filters.

Further, a filter setting method according to a third aspect of the present invention provides an A / D converter for converting an input signal into a digital signal, and first and second filters to which an output signal of the A / D converter is input. A first and second amplifier for amplifying the signals by D / A converting the output signals of the first and second filters, respectively, and a first and a second amplifier for converting the output signals of the first and second amplifiers into sound waves, respectively. And a sound image localization device configured with a second acoustic converter, wherein the acoustic converter can realize the input signal at an arbitrary position as an imaginary sound image, and is provided for the first and second filters. When first and second data strings having sufficiently long tap lengths are set in the first and second filters, the initial delay is set to the first delay.
When the second data string is larger than the data string of the above, when the absolute value of the first predetermined value of the first data string is X, the data immediately before the X and the absolute value of the first predetermined value are the predetermined value. Is replaced with 0, the tap is set as an initial tap, and a point of the second data string, which crosses the 0 axis between the first predetermined value and the second predetermined value. The data is a final tap, and first and second data strings in the range from the initial tap to the final tap are set in the first and second filters, respectively.

[0022]

According to the structure of the first aspect of the invention, the number of taps important for sound image localization is ensured, and taps of low importance are replaced with zeros. By cutting the portion, the filter length can be shortened, and furthermore, the error of the integrated value between LR is reduced, so that sufficient sound image localization can be realized.

Further, according to the configuration of the second aspect, the first aspect is provided.
The sound image localization can be realized while further shortening the filter length as compared with the configuration of the invention of the third aspect.

Further, according to the configuration of the third aspect of the present invention, a good sound image localization can be realized while shortening the filter length by keeping the binaural time difference reaching the both ears in a pseudo manner.

[0025]

An embodiment of the present invention will be described below with reference to the drawings.

The hardware configuration of the sound image localization apparatus for realizing the filter setting method of the present invention is the same as that shown in FIG. 4 described in the conventional example. That is, the filter setting method of the present invention has a feature in the setting method of the filters 3L and 3R in the configuration shown in FIG.

FIG. 1 is a diagram showing a filter setting method according to the first embodiment of the present invention. In (a) and (b), the black circles (●) localize the sound image to the virtual sound image 7 which is discretized by the sampling frequency f (the localization position is shown in FIG. 4).
Is naturally not limited to the position of the filter 3R and 3L).
(N) and the data string of FL (n) are shown. Also,
(C) SFR and SFL shown in (d) are filter characteristics to be set for the filters 3R and 3L shown in FIG. 4, and data (tap) in a range surrounded by a rectangle is set. That is, it is composed of taps cut out from the data sequence of FR (n) and FL (n).

Next, a method of setting a filter will be described. FL (n) and FR with sufficiently long tap length
The data string of (n) is, for example, 2 for each of SFL and SFR.
When setting 2-tap data, first
From (n), as shown in FIG.
Data of 1 to 14 taps, which is a portion corresponding to the wavelength, is cut out from the data string. And as shown in the figure, 15tap to 22t
Substitute 0 for ap.

Next, from FL (n), as shown in FIG. 1D, 1 tap including a portion corresponding to one wavelength of the sin curve is obtained.
Cut out ~ 20 taps of data from the data string. And 21, 22
Substitute 0 for tap.

According to the above setting method, the filters 3L, 3L
The energy difference (integral difference) between SFL and SFR set for each of R is reduced. In the example of FIG.
It is also possible to reduce 2 taps.

According to the above setting method, when the filter coefficients are created, even if the hardware scale of the sound image localization apparatus cannot be made large, the filter length can be shortened by cutting the zero portion, and the LR interval can be further reduced. Since the error of the integral value is reduced, sufficient sound image localization can be realized.

Hereinafter, a second embodiment of the present invention will be described with reference to the drawings. FIG. 2 is a diagram illustrating a filter setting method according to a second embodiment of the present invention.
In (a) and (b), black circles (•) indicate the sound image localization on the virtual sound image 7 that is discretized by the sampling frequency f (the localization position is not limited to the position in FIG. 4). Calculated to set to 3L,
7 shows a data string of FR (n) and FL (n) described in the conventional example. Also, SFR and SFL shown in (c) and (d)
Is a filter characteristic to be set in the filters 3R and 3L shown in FIG. 4, and data (tap) in a range surrounded by a rectangle is set. That is, it is composed of taps cut out from the data sequence of FR (n) and FL (n).

Next, a method of setting a filter will be described. FL (n) and FR with sufficiently long tap length
The data string of (n) is, for example, 2 for each of SFL and SFR.
When setting 2-tap data, first
From (n), as shown in FIG.
Data of 1 to 10 taps, which is a portion corresponding to a half wavelength, is cut out from the column. And, as shown in the figure, 11 tap to 22 tap
Is substituted with 0.

Next, from FL (n), as shown in FIG. 2D, a portion corresponding to a half wavelength of the sine curve is included.
Data of 1 to 15 taps is cut out from the data string. And one
0 is substituted for 6 tap to 22 tap.

According to the above setting method, the filters 3L, 3L
The energy difference (integral difference) between SFL and SFR set for each of R is reduced. In the example shown in FIG.
7 taps can also be reduced.

With the above setting method, when the filter coefficients are created, even if the hardware scale of the sound image localization apparatus cannot be made large, the filter length can be shortened by cutting the zero portion, and the LR interval can be further reduced. Since the error of the integral value is reduced, sufficient sound image localization can be realized.

Hereinafter, a third embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a diagram showing a filter setting method according to the third embodiment of the present invention.
In (a) and (b), black circles (•) indicate the sound image localization on the virtual sound image 7 that is discretized by the sampling frequency f (the localization position is not limited to the position in FIG. 4). Calculated to set to 3L,
7 shows a data string of FR (n) and FL (n) described in the conventional example. Also, SFR and SFL shown in (c) and (d)
Is a filter characteristic to be set in the filters 3R and 3L shown in FIG. 4, and data (tap) in a range surrounded by a rectangle is set. That is, it is composed of taps cut out from the data sequence of FR (n) and FL (n).

Next, a filter setting method will be described. In FIG. 3, when the number of taps required to localize the ischemic image is 7 taps, the SFL having a large initial delay is used.
The last tap of the portion corresponding to one wavelength of the sin curve is the 18th FL (n), and this is the last tap.

Next, an initial tap is obtained.
Is the sixth peak. The absolute value of this amplitude value is X
Then, the third having the absolute value of X / 2 or less and having the earliest value in time is FR (n), which is set as the initial tap.

Further, the tap length required for the initial delay is 7 taps.
At this time, the ninth data of FL (n) is not 0. Here, if the value at this time is equal to or less than １ ／ of the absolute value X of the next peak value, it can be approximated to 0, but the ninth data applies to this.

As described above, first, from FR (n), SFR
Are set to data of 4 to 18 taps and a value obtained by substituting 0 for the third tap. Also, 10 tap to 18 tap data is cut out from the data string from FL (n). Then, 0 was substituted into 6 taps to 9 taps and set to SFL.

According to the above setting method, even when the initial delay difference between the SFL and the SFR cannot be sufficiently obtained, the filter coefficient can be created by substituting 0 under a predetermined condition, and sufficient sound image localization can be realized.

[0043]

As is apparent from the above description, the filter setting method of the present invention can realize a sufficient sound image localization even when the hardware scale of the sound image localization device cannot be made large.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining a filter setting method according to a first embodiment of the present invention;

FIG. 2 is a diagram for explaining a filter setting method according to a second embodiment of the present invention;

FIG. 3 is a diagram for explaining a filter creation method according to a third embodiment of the present invention;

FIG. 4 is a block diagram of a conventional sound image localization apparatus.

FIG. 5 is a diagram showing a conventional filter setting value;

FIG. 6 is a diagram showing filter setting values according to a conventional filter setting method.

FIG. 7 is a diagram showing filter setting values according to a conventional filter setting method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Analog input signal source 2 A / D converter 3L, 3R filter 4L, 4R amplifier 5L, 5R Speaker 6 Listener 7 Virtual image

──────────────────────────────────────────────────の Continuing on the front page (72) Inventor Riko Ogawa 1006 Kazuma Kadoma, Osaka Prefecture Inside Matsushita Electric Industrial Co., Ltd. (56) References JP-A-4-169974 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) H04S 1/00

Claims (11)

(57) [Claims] An A / A converter for converting an input signal into a digital signal.
A D converter, first and second filters to which the output signal of the A / D converter is input, and a D / A converter for D / A converting the output signals of the first and second filters, respectively, to amplify the signal. In a sound image localization device including first and second amplifiers and first and second acoustic transducers for converting output signals of the first and second amplifiers into sound waves, an arbitrary position is determined by the acoustic transducer. The first and second filters, provided for the first and second filters, having sufficiently long tap lengths, which can realize the input signal as an imaginary sound image.
Is set in the first and second filters, between the first predetermined value and the second predetermined value, the data up to the point that crosses the zero axis is the first and second filters. 2. A method for setting a filter for a sound image localization apparatus, wherein the first and second filters are configured by using two data strings and thereafter using 0. 2. The sound image localization according to claim 1, wherein the first predetermined value is a first peak value of the data sequence, and the second predetermined value is a second dip value of the data sequence. Device filter setting method. 3. The sound image localization according to claim 1, wherein the first predetermined value is a first dip value of the data sequence, and the second predetermined value is a second peak value of the data sequence. Device filter setting method. 4. An A / D converter for converting an input signal into a digital signal.
A D converter, first and second filters to which the output signal of the A / D converter is input, and a D / A converter for D / A converting the output signals of the first and second filters, respectively, to amplify the signal. In a sound image localization device including first and second amplifiers and first and second acoustic transducers for converting output signals of the first and second amplifiers into sound waves, an arbitrary position is determined by the acoustic transducer. The first and second filters, provided for the first and second filters, having sufficiently long tap lengths, which can realize the input signal as an imaginary sound image.
Is set in the first and second filters, the first and second data strings are used in a predetermined range, and the first and second filters are set to 0 in other ranges. And a second filter. A filter setting method for a sound image localization apparatus, comprising: 5. The filter setting method for a sound image localization apparatus according to claim 4, wherein the predetermined range is a range that first crosses the 0 axis before and after the first peak value of the data string. 6. The method according to claim 4, wherein the predetermined range is a range that first crosses the 0 axis before and after the first dip value of the data string. 7. An A / D converter for converting an input signal into a digital signal.
A D converter, first and second filters to which the output signal of the A / D converter is input, and a D / A converter for D / A converting the output signals of the first and second filters, respectively, to amplify the signal. In a sound image localization device including first and second amplifiers and first and second acoustic transducers for converting output signals of the first and second amplifiers into sound waves, an arbitrary position is determined by the acoustic transducer. The first and second filters, provided for the first and second filters, having sufficiently long tap lengths, which can realize the input signal as an imaginary sound image.
Is set in the first and second filters, when the initial delay is larger in the second data string than in the first data string, the absolute value of the first predetermined value in the first data string is When the value is X, the tap immediately preceding the X and whose absolute value is equal to or smaller than a predetermined value is replaced with 0, and the tap is set as an initial tap. The data at the point crossing the 0 axis between the first predetermined value and the second predetermined value is defined as the final tap, and the first and second filters are used to determine the data in the range from the initial tap to the final tap. A filter setting method for a sound image localization apparatus, wherein a first data string and a second data string are respectively set. 8. The value of the predetermined magnitude is equal to 1 / X of the X.
8. The method for setting a filter for a sound image localization apparatus according to claim 7, wherein the value is 2. 9. The sound image localization according to claim 7, wherein the first predetermined value is a first peak value of the data string, and the second predetermined value is a second dip value of the data string. Device filter setting method. 10. The sound image localization according to claim 7, wherein the first predetermined value is a first dip value of the data sequence, and the second predetermined value is a second peak value of the data sequence. Device filter setting method. 11. The number of the peak value of the data sequence is assigned to the maximum value of the maximum value in the range where the data line crosses the 0 axis for the first time before and after the peak value, and the number of the dip value of the data sequence is 11. The method according to claim 2, wherein the minimum value is added to a minimum value in a range where the value first crosses the zero axis before and after the value.
The method for setting a filter for a sound image localization apparatus according to any one of the above.