JPH09187100A - Sound image controller - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide the sound image controller in which a processing amount of arithmetic processing for convolution arithmetic processing to control a sound image formed virtually to a listener is reduced so as to realize a high speed processing. SOLUTION: The sound image controller 1 is provided with a delay means 13 delaying a sound source signal, a delay amount setting means 14 setting the delay by the delay means 13 to be value corresponding to a presence direction, and a convolution arithmetic processing means 11 having a time response characteristic of a filter coefficient operator the presence direction, a time response characteristic of the filter coefficient corresponding to the presence direction, applying convolution processing between a period when the filter coefficient is at a prescribed level or over at all times and a delayed sound signal from the delay means 13 to provide an output of left/right channel signals. The delay amount setting means 14 sets the delay amount to be a value corresponding to the first half period in the time response characteristic where the filter coefficient is smaller than a prescribed level.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound image control device for allowing a listener to perceive a virtual sound image of a sound source (source position created in the head).

[0002]

2. Description of the Related Art Conventionally, as a sound image control device for causing a listener to perceive a sound image that is virtually formed, a sound emitted from a plurality of speakers is controlled so that a sound source seems to be present at that position. Various types of sound image forming devices have been proposed. For example, as shown in FIG. 3, a sound source O is arranged diagonally rearward to the left of the listener 50, and sounds output from the left and right speakers SL and SR arranged in front of the listener 50 are heard in both ears of the listener 50. When the input sound is a signal equivalent to the sound input from the sound source O to both ears of the listener 50 when actually input, the listener 50 seems to have the sound source at the O position. Will be heard.

That is, in the figure, O: sound source KL, KR: transfer function from the sound source to each of the left and right ears H: transfer function from the left and right speakers to each of the right and left ears EL, ER: input to both ears of the listener 50 TL, TR: left and right convolution processing units (FIR)
Filter characteristics SL, SR: Left and right speakers.

The transfer function H from the left and right speakers to the left and right ears includes a transfer function HLL from the left speaker SL to the left ear, a transfer function HRR from the right speaker SL to the right ear, and a left speaker SL. Transfer function HLR from the speaker to the right ear, transfer function HR from the right speaker SR to the left ear
There is L. Note that the transfer function HLL is a time response of a sound signal input to the left ear of the listener 50 when a unit impulse is applied to the left speaker SL at time t = 0. The same applies to the transfer functions HRR, HLR, and HRL.

When the sound signal S is output from the position of the sound source O, the left and right ear signals are as shown in the following equation (1).

[0006]

[Equation 1]

[0007] Further, the sound signal S is converted into filter characteristics TL, TR
When reproduced from the speakers SL and SR through the FIR, signals at the left and right ears are as shown in the following Expression 2.

[0008]

[Equation 2]

From the above equations (1) and (2), filter characteristics TL,
When TR is obtained, the following equation 3 is obtained.

[0010]

(Equation 3)

The sound image virtually formed for the listener is controlled by convolution processing of the sound source signal by the FIR having the filter characteristics TL and TR satisfying the relational expression (3), and its control direction. A sound image is formed as if there is a sound source.

Then, as shown in FIG. 4, sound sources are provided around the listener 50 in eight directions (0 °, 45 °, 90 °, 135).
°, 180 °, 225 °, 270 °, 315 °), and when the sound image is arranged in any one of eight directions, the sound image is arranged in each direction. Having a filter characteristic determined so that
For example, when a sound image is formed in the 0 ° direction, FIR0 for L and FIR0 for R are used, and when a sound image is formed in the 45 ° direction, FIR1 for L and FIR1 for R are used. Also, 90 °, 135 °, 180 °, 225
Similarly, when forming a sound image in the directions of °, 270 °, and 315 °, the FIRs 3 to 8 for L and the FIRs 2 to 7 for R are used, respectively.

[0013]

In this conventional sound image control device, FIR has a filter characteristic as shown in FIG. 5, and is represented by a relative amplitude indicating a filter coefficient for each time. Here, the filter characteristic of the solid line shows the FIR0 for L corresponding to the sound image direction 0 °, and the filter characteristic of the dotted line shows the FIR1 for L corresponding to the sound image direction 45 °.

As can be seen from this figure, the FIR filter characteristics differ depending on the direction of the virtual sound image, and the sections in which the filter coefficient in the first half and the latter half of the time have a small ratio to the peak value also differ depending on the direction of the virtual sound image. ing.

For this reason, conventionally, a section where the filter coefficient is small in the first half of time (for example, Δt1 and Δt in FIG. 5).
2) In FIR having the longest filter characteristic,
The convolution calculation processing is performed in each FIR over the entire section (for example, Ta in FIG. 5) until the filter coefficient becomes smaller than a predetermined level. In FIG. 5, each FI is
In R, the first half sections in which the filter coefficient is smaller than 0.05 with respect to the relative amplitude of the peak value are Δt1 and Δt2, respectively, and the section with 0.05 or more relative to the relative amplitude of the peak value is T. It is represented by.

Therefore, since the filter coefficient is as small as a predetermined level or less and the calculation processing is performed even in the section where the convolution calculation processing is not required, the processing amount becomes enormous.
It was an obstacle to speeding up the processing.

As a solution to this problem, a method is conceivable in which the convolution operation processing is not performed only in the section where the filter coefficient is small in the first half of the time (for example, Δt1 in FIG. 5) in all FIRs, but as shown in FIG. As described above, since the filter characteristics of each FIR are different, in this case, the processing amount can be slightly shortened. For this reason,
The convolution operation processing still has to be performed for the sections where the filter coefficient is small in the first half and the second half of the time, and the speedup of the processing cannot be realized.

The present invention has been made in view of the above point, and reduces the processing amount of the convolution calculation processing for controlling the sound image virtually formed for the listener, and It is an object of the present invention to provide a sound image control device that realizes high speed.

[0019]

According to the present invention, a sound source signal is corrected based on direction information on a circumference in a horizontal plane centered on the listener, which represents a virtual positional relationship between the listener and the sound source. In the sound image control device that outputs the left and right sound signals and generates a sound image corresponding to the localization direction based on the direction information, a delay unit that delays the sound source signal, and a delay amount by the delay unit that corresponds to the localization direction. And a delay amount setting means to be set to, and a time response characteristic of a filter coefficient corresponding to the localization direction, a section in which the filter coefficient is always above a predetermined level, and a convolution operation processing of the delayed sound source signal from the delay means. And a convolution calculation processing means for outputting left and right sound signals, wherein the delay amount setting means compares the delay amount with the first half section of the time response characteristic in which the filter coefficient is smaller than a predetermined level. It is to set the value.

By using this configuration, the sound source signal is delayed by a time corresponding to an interval in which the filter coefficient in the first half portion of the time response characteristic of each convolution operation processing means with respect to the control sound image direction is smaller than a predetermined level, and Since the delayed sound source signal is subjected to the convolution operation processing in the section of the time response characteristic in which the filter coefficient is equal to or higher than the predetermined level, the processing amount of the operation processing in the convolution operation processing means can be shortened and the processing speed can be increased. .

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings showing an embodiment thereof. FIG. 1 is a functional block diagram showing an embodiment of a sound image control device of the present invention.

As shown in the figure, the sound image control apparatus 1 of the present invention
Is a convolution operation processing unit (FIR) 11, coefficient setting means 12, delay means 13, delay amount setting means 14, multiplier 15
And adders 16 and 17.

The FIR 11 is a sound source according to filter characteristics which are determined separately for right and left so that a direction in a horizontal plane centering on a listener is virtually divided into 8 parts to form a sound image in each direction as in the case of FIG. 4 described above. It is provided with an FIR FIR 11a for L and an FIR 11b for R that correct the signal and output the left and right sound signals. The filter characteristic of the FIR 11 is determined by the time response characteristic of the multiplication coefficient (filter coefficient). And
The multiplication coefficient data is set based on the input data from the coefficient setting means 12 described later.

Here, the FIR0 for L and the FIR0 for R are
It has a filter characteristic for forming a sound image in the 0 ° direction. That is, the sound source signals are FIR0 for L and FIR0 for R.
When they are output as left and right sound signals respectively, the sound image is formed in the 0 ° direction. That is, the listener sounds as if there is a speaker in the 0 ° direction.

The FIR1 for L and the FIR1 for R have filter characteristics for forming a sound image in the direction of 45 °. That is, when the sound source signal is output as the left and right sound signals through the FIR 1 for L and the FIR 1 for R, a sound image is formed in the direction of 45 °. That is, the listener sounds as if there is a speaker in the 45 ° direction.

FIR for L2 to FIR7 for L and FIR for R
Similarly, the filter characteristics are determined so that sound images are formed in the corresponding directions for the FIRs 2 to R for 7 as well. still,
Since the method of determining the filter characteristic of the FIR 11 is the same as the conventional method described above, the details are omitted.

The coefficient setting means 12 determines the filter characteristics of each of the FIR 11a and FIR 11b based on the input direction information, and the FIR 11 corresponding to the input direction.
a, the multiplication coefficient of FIR 11b is read out, and its FIR1
1a and FIR 11b, respectively. Here, a section in which the multiplication coefficients of the FIR 11a and the FIR 11b are equal to or higher than a predetermined level (in the present embodiment, set when the filter coefficient is 0.05 or more with respect to the relative amplitude of the peak value in each FIR) (for example, the figure The multiplication coefficient for each time in T) in 5 is output. This is FI
Since the convolution operation processing is not performed in the first half and the second half of the time when the multiplication coefficient of R11 is equal to or lower than the predetermined level, the multiplication coefficients of only the sections where the multiplication coefficient of R11 is equal to or higher than the predetermined level are output, and FIR11a and FIR1 are output based on the values.
The multiplication coefficient is set in 1b.

Specifically, FIR1 for each sound image direction
Each of the FIRs 11a and 11b corresponding to the input direction has a multiplication coefficient memory unit in which the respective multiplication coefficient data of 1a and 11b is stored, and which is based on the input direction information.
Read the multiplication coefficient data of the
11a and 11b, respectively. Then, in the FIRs 11a and 11b corresponding to the input direction, h (1) to h (m) are set as the multiplication coefficient data set based on the input data from the coefficient setting means 12, and the sound source from the delay means 13 described later. The convolution operation is performed with the signal data. Here, m is the number of data of the multiplication coefficient in FIR11, and h (1) is the first multiplication coefficient value with respect to the time axis,
h (m) represents a multiplication coefficient value after a lapse of a predetermined time.

The delay unit 13 has its delay amount set by the delay amount setting unit 14, delays the sound source signal for a predetermined time, and inputs it to the FIR 11. Specifically, the delay unit 13 has a memory unit 21 for storing and holding the input sound source signal for a predetermined time as shown in FIG. 2, and the past n data of the input sound source signal corresponds to the address A ( n) is stored and held, and address A (n) is input every time new data is input.
Is updated and the previous data is changed to A (n-1).
Then, according to the input of the control signal from the delay amount setting means 14 which will be described later, the head address A (nm−m−
(d + 1) is designated, and data from the designated address to the m-th address A (n-d) is output at any time. Here, the number of input data is n, the number of data of the multiplication coefficient in the FIR 11 is m, and the delay amount is d (d ≦ n−m).

However, the amount of input data stored and held in the memory section 21 is the number of data obtained by adding at least the number of data for the maximum delay time to the number of data of the multiplication coefficient which is subjected to the convolution operation processing by the FIR 11. .

The delay amount setting means 14 sets the delay amount of the delay means 13 according to the direction information, and each multiplication coefficient of the FIR 11a and FIR 11b in the control sound image direction has a predetermined level (in this embodiment, in each FIR). A section (for example, Δt1 and Δt in FIG. 5) where the multiplication coefficient is smaller than 0.05 when the relative amplitude of the peak value is set.
Source signal delayed by the time corresponding to 2) is FIR
11a and FIR 11b are input.

Therefore, in the FIR 11, the convolution operation processing shown by the following equation is performed, and the result y (n) is multiplied by the multiplier 1
5 is being sent.

[0033]

(Equation 4)

The multiplier 15 receives the output of the FIR 11 and
It is multiplied by a predetermined coefficient based on the direction information. That is, each FI
The sound signal that has passed through R11 is weighted. For example, if the direction information indicates that the direction of the sound image in the zone 0 is close to the 45 ° direction, the ratio of the left and right sound signals of the L FIR1 and the R FIR1 is more than the left and right sound signals of the L FIR0 and the R FIR0. The multiplication coefficient is set so that
On the contrary, if the direction information indicates that the sound image of zone 0 is close to the 0 ° direction, L0 FIR0 and R1 FI
The ratio of the left and right sound signals by R0 is calculated by FIR1 for L and F for R.
The multiplication coefficient is set so as to be higher than the left and right sound signals due to IR1. Therefore, an inverse interlocking type multiplier can be used as the multiplier 15.

The adder 16 receives the F from the multiplier 15 for L.
The IR outputs are added and output as a left sound signal. The adder 17 adds the FIR outputs from the multiplier 15 for R and outputs the result as a right sound signal.

With the above configuration, the sound source signal is delayed by the time corresponding to the section in which the multiplication coefficient in the first half of the time of each FIR 11 with respect to the control sound image direction is smaller than the predetermined level, and the delayed sound source signal is multiplied by the multiplication coefficient. Since the convolution operation processing is performed by each FIR 11 having a filter characteristic of a predetermined level or higher, the processing amount of the operation processing in the FIR 11 can be shortened and the processing speed can be increased.

The description of the above embodiments is for the purpose of explaining the present invention, and should not be construed as limiting the invention described in the claims or reducing the scope thereof.
In addition, the configuration of each part of the present invention is not limited to the above-described embodiment, and it goes without saying that various modifications can be made within the technical scope described in the claims.

[0038]

As described above, according to the present invention, the processing amount of the convolution calculation processing for controlling the sound image virtually formed for the listener in the localization direction is shortened and the processing speed is increased. It can be realized.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing an embodiment of a sound image control device of the present invention.

FIG. 2 is a memory section 21 of a delay unit 13 in the apparatus of FIG.
FIG. 3 is an explanatory diagram for explaining the method.

FIG. 3 is a diagram showing a state of transfer characteristics of a speaker and a sound image to a listener.

FIG. 4 is a diagram showing eight directions for a listener.

FIG. 5 is a diagram showing a filter characteristic for each sound image direction in a conventional convolution operation unit (FIR).

[Explanation of symbols]

 1 Sound Image Control Device 11 Convolution Operation Processing Unit (FIR) 12 Coefficient Setting Means 13 Delay Means 14 Delay Amount Setting Means 15 Multipliers 16 and 17 Adders

Claims (1)

[Claims] 1. A sound source signal is corrected based on direction information on a circumference in a horizontal plane centering on the listener, which represents a virtual positional relationship between the listener and the sound source, and outputs left and right sound signals, In a sound image control device for generating a sound image corresponding to a localization direction based on the direction information, delay means for delaying the sound source signal, and delay amount setting means for setting a delay amount by the delay means to a value corresponding to the localization direction. , Having a time response characteristic of a filter coefficient corresponding to the localization direction, and a convolution operation processing of a section in which the filter coefficient is always above a predetermined level and a delayed sound source signal from the delay means, and outputs left and right sound signals. Convolution operation processing means, wherein the delay amount setting means sets the delay amount to a value corresponding to the first half section of the time response characteristic in which the filter coefficient is smaller than a predetermined level. Characteristic sound image control device.