JP2827777B2 - Method for calculating intermediate transfer characteristics in sound image localization control and sound image localization control method and apparatus using the same - 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 reproducing signals processed by a plurality of signal conversion circuits having the same sound source and having predetermined transmission characteristics from a plurality of transducers arranged at a distance. The present invention relates to sound image localization control that makes a sound image appear to be localized at a desired arbitrary position different from a transducer (speaker) of the present invention. In particular, based on transfer characteristics actually measured (or calculated in advance), The present invention relates to a method for calculating an intermediate transfer characteristic (for example, at an intermediate sound image localization position) and a sound image localization control method and apparatus using the same.

[0002]

2. Description of the Related Art Conventionally, there is a sound image localization method in which a sound source can be felt at a specific position (specific direction) by a signal level difference and a phase difference (time difference) in both ears. This sound image localization method obtains a transfer characteristic for localizing a sound image at a specific position based on a measured HRTF (head-related transfer function), and realizes this transfer characteristic with a convolver (convolution operation circuit) or the like. It is.

In order to perform higher-quality sound image localization control, it is necessary to measure and collect transfer characteristics (data) at many sound image localization positions. For example, FIG.
As shown in FIG.
The use of the 24 sets of transfer characteristics for every fifth degree enables more realistic sound image localization control. Furthermore, when preparing a plurality of positional relationships between a pair of speakers and a listener used for sound image localization, and preparing a plurality of measurement data on which the transfer characteristics are calculated in consideration of individual differences (personal characteristics) of the listeners. In some cases, more transfer characteristics (data) are required.

However, it takes a great deal of time, labor, and cost to measure and collect highly accurate transfer characteristics. Further, it is necessary to store many transfer characteristics (data) measured and collected as data of the sound image localization control device.
The scale of the device increases. In other words, the capacity of the coefficient ROM for the digital filter (convolver) of the sound image localization control device increases dramatically. For this reason, conventionally, based on a transfer characteristic actually measured (hereinafter, referred to as a reference transfer characteristic), a transfer characteristic at an intermediate sound image localization position (hereinafter, referred to as an intermediate transfer characteristic) is actually measured. Attempts have been made to calculate without calculation.

Conventionally, such an intermediate transfer characteristic is defined as 2
It was calculated as the arithmetic average of two reference transfer characteristics. That is, in the calculation in the time axis region, the intermediate transfer characteristic at the intermediate position is approximated from the arithmetic average of the time response waveform of the reference transfer characteristic (the arithmetic average of the amplitude corresponding to time). In the calculation in the frequency domain, vector averaging is performed using two reference transfer characteristics as frequency responses. As a vector mean,
Specifically, the two reference transfer characteristics are expressed as a vector as a frequency response, and the intermediate transfer characteristics are obtained by arithmetically averaging the real part and the imaginary part. The details of this vector averaging are as follows. Here, the discrete time response of an arbitrary transfer characteristic is x (i), and the discrete frequency response of x (i) is X (i) (i = 0... N-1, X (i) is a complex vector).
And One of the two reference transfer characteristics is x (i), X
(I), the other is y (i), Y (i), and z (i), Z (i) are the intermediate transfer characteristics obtained by the arithmetic mean (vector average) of the prior art. And (Equation b).

[0006]

(Equation 1)

(Equation 2)

Therefore, the amplitude Q and phase q of the intermediate transfer characteristic are

[0008]

(Equation 3)

(Equation 4) Becomes

[0009]

However, according to the conventional calculation method for obtaining the intermediate transfer characteristic by arithmetic averaging (vector averaging) as described above, when the calculated intermediate transfer characteristic is used, deterioration of sound image localization and sound quality are caused. There is a problem that deterioration of the glass easily occurs. In order to solve this problem, the present applicant has searched for a cause of deterioration and found a new method of calculating an intermediate transfer characteristic. This will be described sequentially with reference to FIGS.

FIG. 13 is an example showing frequency-amplitude characteristics of two reference transfer characteristics having a relative opening angle of 15 degrees. FIG. 14B shows the frequency-amplitude characteristics of the intermediate transfer characteristics obtained from the reference transfer characteristics by the above-described conventional calculation method. On the other hand, FIG. 14A is a measured value of the intermediate transfer characteristic at an intermediate position of the reference transfer characteristic, and shows the frequency-amplitude characteristic. FIG. 15B shows the frequency-phase characteristics of the intermediate transfer characteristics obtained from the reference transfer characteristics by the above-described conventional calculation method. On the other hand, FIG.
(A) is a measured value of the intermediate transfer characteristic at an intermediate position of the reference transfer characteristic, and shows the frequency-phase characteristic thereof.

As apparent from FIGS. 15A and 15B, in the frequency-phase characteristics, there is no large difference between the value calculated by arithmetic averaging and the measured value. However, as is clear from FIGS. 14A and 14B, in the frequency-amplitude characteristics, there is a large difference between the value calculated by the arithmetic mean and the actually measured value.
There is a large difference in the high frequency range between kHz and 10 kHz, which seems to be the cause of deterioration of sound image localization and sound quality in practical use.

[0012]

The present applicant has further analyzed the frequency-amplitude characteristic of the intermediate transfer characteristic. FIG. 12A shows two reference transfer characteristics X (i) and Y
(I) Vector averaging when the difference between the phase angles is small. FIG. 12B shows that the difference between the phase angles of the two reference transfer characteristics X (i) and Y (i) is large and the phase difference is ± 180 degrees. It is a figure which shows the vector average of close case. As shown in the figure, especially when the phase angle is large, the amplitude of Z (i) obtained by arithmetic averaging (vector averaging) becomes much smaller than that of X (i) and Y (i). Obviously, the amplitude of the intermediate transfer characteristic is clearly ineligible.

That is, in a band where the phase changes drastically, or in a high frequency range where the phase difference is liable to be large in general, 2
There may be a case where the phase angle between the two reference transfer characteristics X (i) and Y (i) is large, and the amplitude of the arithmetic transfer averaged intermediate transfer characteristic is reduced in accuracy, causing deterioration in sound image localization and deterioration in sound quality. I think that the. This agrees with the result of comparison with the actually measured values described in detail with reference to FIGS. 14A and 14B. As shown in FIG. 15 described above, the two reference transfer characteristics (in FIG. 15, the intermediate transfer characteristic at the intermediate position in both cases) have a sharp change in phase in the high frequency range.
It can be said that the difference in phase angle between (i) and Y (i) is likely to be large, and as a result, the accuracy of the amplitude of the intermediate transfer characteristic is reduced.

Therefore, the present invention improves the method of calculating the amplitude of the intermediate transfer characteristics, particularly the amplitude, and correctly calculates the amplitude even when the phase angle between the two reference transfer characteristics is large, thereby deteriorating the sound image localization. It is possible to calculate an intermediate transfer characteristic in which deterioration of sound quality hardly occurs. That is, 2
The amplitudes of the two reference transfer characteristics were geometrically averaged to obtain the amplitude of the intermediate transfer characteristic.

According to the present invention, in order to solve the above problems, FIG.
As shown in FIG. 1 and the like, the same sound source (X) is supplied from a plurality of transducers (loudspeakers sp1 and sp2) disposed apart from each other, and a plurality of signal conversion circuits (convolvers; Coefficients are cfLx, cfR
a signal processed by a convolution operation processing circuit comprising a cancellation filter which is x, and a sound image that makes a listener (M) feel that the sound image is localized at an arbitrary position (x) different from the transducer. This is a method of calculating a transfer characteristic in localization control, wherein a frequency-amplitude characteristic (R) is obtained from a plurality of reference transfer characteristics X (i) and Y (i) measured or calculated in advance (for example, for different sound image localization positions). Is the geometric mean of the amplitude characteristics between the reference transfer characteristics, and the frequency-phase characteristic (r) is the phase component of the vector average between the reference transfer characteristics as the phase component between the reference transfer characteristics (for example, at an intermediate localization position). Intermediate transfer characteristic R
A method for calculating an intermediate transfer characteristic in sound image localization control characterized by calculating (i), and a sound image localization control method and apparatus using the same.

[0016]

According to the method for calculating the transfer characteristics in the sound image localization control as described above, even when the phase angle between the reference transfer characteristics X (i) and Y (i) is large, the intermediate transfer characteristics R (i) are obtained.
Is calculated as a value close to actual measurement as a geometric mean. Then, based on the calculated intermediate transfer characteristics,
The signal is processed by a signal conversion circuit, and a sound image is formed (for example,
The sound is localized at an intermediate position between the sound image localization positions based on the reference transfer characteristic.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for calculating an intermediate transfer characteristic in sound image localization control and a sound image localization control device using the same according to the present invention will be described below with reference to the drawings. At first,
The basic principle of the sound image localization control method will be described. This is a technique in which a sound image is localized at an arbitrary position in a space using a pair of transducers (hereinafter, a speaker will be described as an example) that are arranged apart from each other.

FIG. 5 is a principle diagram of sound image localization. sp1,
sp2 is a speaker arranged in front and left of a listener (which may be referred to as a listener in the embodiment).
The head transfer characteristic (impulse response) from 1 to the listener's left ear is h1L, the head transfer characteristic from the right ear to h1R is sp
The head transfer characteristics from 2 to the left and right ears are defined as h2L and h2R. Also, the head transfer characteristic to the listener's left and right ears when an actual speaker is placed at the target localization position x is pL.
x, pRx. Here, each transfer characteristic is obtained by appropriately measuring waveforms obtained by arranging a speaker in an acoustic space and microphones at both ear positions of a human head or a dummy head and performing actual measurement.

Next, signals obtained by passing a sound source X to be localized through signal converters cfLx and cfRx (transfer characteristics by a convolver or the like) are respectively converted into sp1 and sp
Consider playing on 2. At this time, assuming that the signals obtained from the left and right ears of the listener are eL and eR, eL = h1L · cfLx · X + h2L · cfRx · X (Equation 1) eR = h1R · cfLx · X + h2R · cfRx · X (〃) Let dL and dR be the signals obtained at the listener's left and right ears when the sound is reproduced from the target localization position, dL = pLx · X (Equation 2) dR = pRx · X (〃)

Here, if the signals obtained in the left and right ears of the listener by the reproduction of sp1 and sp2 coincide with the signals obtained when the source is reproduced from the target position, the listener can operate as if the speaker exists at the target position. It will recognize the sound image. From the conditions eL = dL, eR = dR and (Equation 1) and (Equation 2), X is eliminated and h1L · cfLx + h2L · cfRx = pLx (Equation 3) h1R · cfLx + h2R · cfRx = pRx (〃) (Equation 3) ), CfLx = (h2R · pLx−h2L · pRx) / H (Equation 4a) cfRx = (− h1R · pLx + h1L · pRx) / H (〃) where H = h1L · h2R−h2L · h1R (Equation 4b) Therefore, if a signal to be localized by a convolver (convolution operation processing circuit) or the like is processed using the transfer characteristics cfLx and cfRx calculated by (Equation 4a) and (Equation 4b), the target position x can be obtained. The sound image can be localized.

There are various possible ways of realizing the signal conversion device, but it is only necessary to use an asymmetric FIR digital filter (convolver). The final transfer characteristic when used in the FIR digital filter is a time response function. That is, as the transfer characteristics cfLx and cfRx at the required localization position x, those obtained by (Equation 4a) and (Equation 4b) are used as coefficients for realizing by one FIR filter process, and the coefficients cfLx and cfRx are used. It is created in advance and prepared as ROM data. R
If the coefficients of the required sound image localization position are transferred from the OM to the FIR digital filter, the signal from the sound source is convolution-processed, and reproduced from a pair of speakers, the sound image is localized at a desired arbitrary position.

The sound image localization control method based on the above principle and the method of calculating the intermediate transfer characteristic in the sound image localization control method will be described in detail with reference to FIGS. 3, 4, and 6 to 8. FIG. 3 shows steps of a method of sound image localization control, and FIG. 4 is a system configuration diagram of the sound image localization control device.

Head Related Transfer Function
Function; hereinafter referred to as HRTF) (Step 101) This will be described with reference to FIGS. FIG. 6 shows the HRTF
1 shows a measurement system. A pair of microphones ML and MR are installed at both ears of the dummy head (or human head) DM, receive a measurement sound from the speaker SP, and a recorder DAT.
Source sound (reference data) refL, refR
And the measured sound (measurement data) L and R are recorded in synchronization.
The position of the speaker SP is set at a plurality of angles θ (for example, as shown in FIG. 7, 24 points every 15 degrees) in a space where the front is determined to be 0 degrees (°), and each is set for a predetermined time. Record continuously.

HRTF impulse response (Impulse Re
calculation of sponse (hereinafter referred to as IR) (step 10)
2) In step 101, the source sounds (reference data) refL, refR and the measured sounds (measurement data) L, R, which are recorded synchronously, are transmitted to a workstation (not shown).
Process above. X is the frequency response of the source sound (reference data)
(S), the frequency response of the measured sound (measurement data) is represented by Y
(S) IR frequency response of HRTF at measurement position
Assuming (S), there is an input / output relationship shown in (Equation 5). Y (S) = IR (S) · X (S) (Equation 5) Therefore, the frequency response of the HRTF is expressed by IR (S) = Y (S) / X (S) (Equation 6) is there.

Therefore, the frequency response X of the reference
(S) As the frequency response Y (S) of the measurement data, the data obtained in the step 101 is cut out by a window synchronized with time, and each of the data is subjected to a finite Fourier series expansion by FFT transform to obtain a discrete frequency. Thus, the frequency response IR (S) of the HRTF is obtained by a well-known calculation method.
In this case, the accuracy of IR (S) is increased (improvement of SN ratio).
For several hundred windows that differ in time
It is good to calculate (S) and average them. And
The calculated HRTF frequency response IR (S) is subjected to inverse FFT transform, and the HRTF time axis response (impulse response) IR
(First IR).

IR (Impulse Response) Shaping Process (Step 103) Here, the IR obtained in Step 102 is shaped. First, for example, by FFT transform, the first
The IR band at discrete frequencies over the audio spectrum, and remove unnecessary bands (a large dip occurs in the high frequency band, which is an unnecessary one that does not significantly affect sound image localization) with a BPF (bandpass filter). I do. When the band is limited in this manner, unnecessary peaks and dips on the frequency axis are removed, and unnecessary coefficients are not generated in the cancel filter. Therefore, convergence is improved and the coefficients can be shortened.

Then, the band-limited IR (S) is converted to the inverse F
After performing FT conversion, an IR (impulse response) is multiplied by a cutout window (for example, a window of a cosine function) on the time axis, and a window process is performed (a second IR). By performing the window processing, the effective length of the IR is no longer long, the convergence of the cancel filter is improved, and the sound quality does not deteriorate.

Cancel filters cfLx, cfRx
(Step 104) The cancellation filters cfLx and cfRx, which are convolvers (convolution integrators), are, as shown in (Equation 4a) and (Equation 4b), cfLx = (h2R · pLx−h2L · pRx) / H (Formula 4a) cfRx = (− h1R · pLx + h1L · pRx) / H (() where H = h1L · h2R−h2L · h1R (Formula 4b).

Here, the speakers sp1 and sp
2 transfer characteristics h1L, h1R, h2L, h2R
The second IR (formed at steps 101 to 103, which has been shaped and processed for each angle θ, as the head-related transfer characteristics pLx and pRx when an actual speaker is placed at the target localization position x). Impulse response). The head transfer characteristics h1L and h1R correspond to the position of the L-channel speaker in FIG. 8, and if it is set at, for example, 30 degrees (θ = 330 degrees) to the left from the front, θ
= 330 degrees IR is used. Head transfer characteristics h2R, h2
L corresponds to the position of the R-channel speaker in the figure, and if it is installed at, for example, 30 degrees (θ = 30 degrees) from the front to the right, an IR of θ = 30 degrees is used (that is, the actual IR is used). (For example, a system close to the system at the time of reproducing the sound image (for example, shown in FIG. 4 described later) is selected).

The head-related transfer characteristics pLx and pRx are 90 degrees from the front which is the target sound source localization position.
The range of 180 degrees is of course (24 in FIG. 8).
The position of 0 degree is taken as an example), and the IR of every 15 degrees in a wide space (entire space) beyond that is substituted, so that the corresponding cfLx, cfR of the entire space is obtained.
x, that is, 24 sets of cancellation filters cfLx and cfRx are obtained for every 15 degrees actually measured. The group of cancellation filters cfLx and cfRx is finally obtained as an IR (impulse response) which is a response on the time axis. However, the coefficients (group) cfLx and cfRx of these cancellation filters are 2 for every 15 degrees actually measured.
There are only four sets.

Calculation of Intermediate Transfer Characteristics (Step 105) Then, based on the coefficients cfLx and cfRx of the cancel filter (convolver) calculated based on the actual measurement, the transfer characteristics at the intermediate sound image localization position ( Intermediate transfer characteristics) are obtained and 48 sets of transfer characteristics are set every 7.5 degrees. In the calculation of the intermediate transfer characteristic, in the present invention, the frequency-amplitude characteristic of the intermediate transfer characteristic is obtained as the geometric mean of the amplitude characteristics of the two reference transfer characteristics, and the frequency-phase characteristic is obtained by calculating the frequency complex vector of the two reference transfer characteristics. Obtained as the phase component of the vector average. This is the main part of the present application, and will be described below with reference to FIGS.

FIG. 1 is a conceptual diagram in which a method of calculating an intermediate transfer characteristic from two reference transfer characteristics is represented by a vector. One of the two reference transfer characteristics is x (i), X (i) and the other is y (i), Y (i), and the intermediate transfer characteristics determined by the above method are r (i), R (i). Then, the amplitude R and the phase r of the intermediate transfer characteristic are represented by the following (formula c) and (formula d).

[0033]

(Equation 5)

(Equation 6)

Therefore, for example, an intermediate transfer characteristic of 37.5 degrees, which is an intermediate position between 30 degrees and 45 degrees, is obtained by calculating the time axis response of the cancel filter at 30 degrees and 45 degrees obtained in the above step by two reference values. Transfer characteristics x (i), y
As (i), the time axis response r (i) of the intermediate transfer characteristic may be obtained. That is, the time axis response is x (i), y
(I) is subjected to FFT transformation (orthogonal transformation such as FFT transformation), and frequency responses X (i) and Y (i) are obtained from (Equation c) and (Equation d). i) is obtained, and this is subjected to inverse FFT transform to obtain a time axis response r (i) of the intermediate transfer characteristic.

According to such a method of calculating the intermediate transfer characteristics, as described above, the reference transfer characteristics X (i) and Y (i)
Is large, the amplitude R of the intermediate transfer characteristic R (i) with respect to the intermediate sound image localization position is calculated as a value close to the actual measurement as a geometric mean. Figure 2 shows this point.
This will be described schematically with reference to (A) and (B). FIGS. 2A and 2B show the intermediate transfer characteristics R (i) when the amplitude of X (i) is fixed at 1 and the amplitude of Y (i) is changed to 0.25, 0.5, 1, 2, and 4. (A) shows the geometrically averaged R (i) of the embodiment of the present invention, and FIG. (B) shows the arithmetically averaged Z (i) of the prior art. Show. The arithmetically averaged Z (i) is the reference transfer characteristic X (i) and Y
Significantly different from the amplitude of (i). On the other hand, the geometrically averaged R (i) has a substantially intermediate magnitude, and it can be said that the amplitude is substantially appropriate as the amplitude of the intermediate transfer characteristic.

As a result, as shown in FIGS. 14A and 14C, an intermediate transfer characteristic close to the actual measurement is obtained. FIG. 7A shows actual measured values of the frequency-amplitude characteristics of the intermediate transfer characteristics, and FIG. 7C shows the frequency-amplitude characteristics of the intermediate transfer characteristics calculated by geometric mean. Both have very similar characteristics. Therefore, it can be said that an intermediate transfer characteristic that closely approximates the actually measured value and hardly deteriorates the sound image localization feeling and the sound quality is obtained. In this way, 24 sets of intermediate transfer characteristics are calculated by calculation, and 48 sets of transfer characteristics (coefficients of cancel filters cfLx and cfRx) are added every 7.5 degrees, together with 24 sets of transfer characteristics actually measured every 15 degrees. Is obtained.

In general, 360 / M (M:
When the intermediate transfer characteristic is calculated N times (N: an integer of 1 or more) based on the reference transfer characteristic for each degree (an integer of 2 or more) degrees, 360
/ (M · 2 ^N ) degrees can control sound image localization. In the above embodiment, M = 24 and N = 1 are obtained every 7.5 degrees to perform sound image localization control. However, M = 16 and N = 2 are obtained every approximately 5.63 degrees to perform sound image localization control. May be. In addition, the resolution of human sound image recognition on the horizontal plane is
It is said to be about 3-4 degrees. Therefore, M = 24,
If the intermediate transfer characteristic is calculated twice with N = 2 and the sound image processing is performed about every 3.75 degrees, it is possible to localize the sound image at an arbitrary position almost continuously.

The scaling of the cancellation filter at each localization point x (step 106) The spectrum distribution of the sound source (source sound) actually subjected to sound image processing by the convolver (cancel filter) is statistically similar to pink noise. In some cases, the sound source is different from a single sound, so there is a danger that the convolution operation (integration) will overflow and cause distortion. is there.

Therefore, in order to prevent overflow,
All coefficients of the cancel filters cfLx and cfRx are scaled. Then, by using a window window (cosine window), both ends are set to 0 in accordance with the number of coefficients of an actual convolver (in this embodiment, a final cancellation filter that has been scaled is referred to as a convolver). , Window processing to shorten the effective length of the coefficient. A data group that is thus scaled and finally supplied as a coefficient to the convolver (in this example, 7.
48 sets of convolver coefficient groups) cfLx and cfRx capable of localizing the sound image every 5 degrees are obtained.

A signal from a sound source is convoluted and reproduced (step 107). For example, as a sound reproducing device for a game machine, as shown in FIG.
A pair of loudspeakers sp1 and sp2 are disposed at regular intervals so that sound signals processed by a pair of convolvers (convolution operation processing circuits) 1 and 2 are reproduced on the pair of loudspeakers sp1 and sp2. Constitute. Coefficient ROM3
Is 7.5 calculated in steps 101 to 106.
48 sets of convolver coefficients cfLx, cfR per degree
x is stored. A signal from the same sound source X (for example, a flight sound from a game synthesizer) is supplied to the pair of convolvers 1 and 2, and a desired coefficient c
fLx, cfRx (for example, when the sound image is to be localized at the position of 120 degrees left behind (θ = 240 degrees), θ =
240 °) is selected and set in the convolvers 1 and 2. For example, based on a sound image localization command from a main CPU (central processing unit) of a game machine or the like, the control sub CPU 4 transfers a coefficient at a desired localization position from the coefficient ROM 3 to the pair of convolvers 1 and 2.

Thus, the pair of convolvers 1 and 2
The signal from the sound source X is subjected to convolution calculation processing on the time axis, and a pair of speakers sp
1 and sp2. A pair of speakers sp1, s
The sound reproduced from p2 is canceled out of the crosstalk to both ears, the sound image is localized so that the sound source is located at a desired position, and the sound is heard by the game operator (listener) M, which is very realistic. Played as sound. As for the coefficients of the convolvers 1 and 2, for example, an optimal sound image position is sequentially selected and switched together with the transition of the movement of the airplane according to the operation of the operator M. When the sound is changed from the flight sound to, for example, a missile sound, the source sound from the sound source X is changed from the flight sound to the missile sound. In this way, the sound image can be freely localized at any position.

As a transducer for reproduction, a headphone can be used instead of the pair of speakers sp1 and sp2. In this case, since the measurement conditions of the HRTF are different, it is preferable to prepare a coefficient separately and switch according to the reproduction state. Further, the configuration described in the embodiment for reproducing a signal processed by a pair of convolvers to which the same sound source is supplied from a pair of transducers (speakers) disposed at a distance is used to obtain the effect of sound image localization. It shows a minimum configuration. Therefore, if necessary,
Of course, a pair, ie, two or more transducers and convolvers may be additionally configured. Further, when the convolver has a long coefficient, the coefficient may be divided into a plurality of convolvers. good.

In the above embodiment, the transmission characteristics of the cancel filters cfLx and cfRx calculated based on (Equation c) and (Equation d) at the stage of step 105 are compared with the intermediate transmission at the intermediate position. The characteristics were determined. However, an intermediate transfer characteristic at an intermediate position may be obtained with respect to the head transfer characteristic (HRTF) measured in steps 101 to 103. Even in this case, the calculation method according to the present invention has the effect. .

(Embodiment 2) Next, a description will be given of a sound image localization control device using the method for calculating the intermediate transfer characteristic described in the above embodiment. In the above embodiment, as shown in FIGS. 3 and 4, an intermediate transfer characteristic is calculated in advance and stored in the coefficient ROM of the sound image localization control device. The sound image localization control device described here is configured so that the intermediate transfer characteristic is calculated on the device side as necessary.

FIG. 9 shows a sound image localization control device. This device has the same purpose as that of the sound image localization control device shown in FIG. 4, and particularly shows its structural features. The coefficient ROM 5 is storage means for storing basic transfer characteristics (coefficients of 24 sets of convolvers for every 15 degrees of the sound image localization position based on actual measurement, for example, calculated in advance in steps 101 to 106). The FFT circuit 6, the intermediate transfer characteristic calculating circuit 7, and the inverse FFT circuit 8 are means for calculating an intermediate transfer characteristic in the frequency domain. The intermediate transfer characteristic calculating circuit 9 calculates the intermediate transfer characteristic in the same manner as in step 106 described above. It is to be calculated.

The coefficient supply means (sub CPU) 9 supplies coefficients of the sound image position to the convolvers 1 and 2 in response to a sound image localization command from the main CPU. When supplying coefficients,
If the coefficient is stored in the coefficient ROM 5, the coefficient supply means 9 supplies it to the convolvers 1 and 2 from the coefficient ROM 5. On the other hand, coefficients not stored in the coefficient ROM 5;
That is, if the coefficient is at the intermediate position, two reference coefficients for calculating the intermediate position (as described above, the reference transfer characteristics on both sides of the intermediate position) are supplied from the coefficient ROM 5 to the FFT circuit 6, and the FFT circuit 6 Convert to response. And
Based on the two converted reference transfer characteristics (frequency response), the intermediate transfer characteristic calculation circuit 7 calculates the intermediate transfer characteristics of both. The calculated intermediate transfer characteristic is converted to a time axis response by the inverse FFT circuit 8 and supplied to the convolvers 1 and 2.

As described above, if the sound image localization control device is configured to calculate the intermediate transfer characteristic as required, the sound image localization position is continuously changed while the capacity of the coefficient ROM is kept small. Thus, smooth movement of the sound image can be easily realized.

The coefficient supply means (sub CPU) 9
In response to a sound image localization command from the main CPU, a subsequent sound image localization position is predicted, an intermediate transfer characteristic is calculated and stored in advance, and a coefficient corresponding to the next sound image localization command can be immediately supplied. May be.

(Third Embodiment) In the second embodiment, the intermediate transfer characteristics are calculated to correspond to a plurality of sound image localization positions. However, the convolvers 1 and 2 use the transfer characteristics cfLx and cfRx at the localization positions as the transfer characteristics cfLx and cfRx. , (Equation 4a), (Equation 4b)
Was obtained by one FIR filter process.

On the other hand, in this embodiment, as shown in FIG. 10, the transfer characteristics cfLx, cfR
As x, the value obtained by (Equation 4a) and (Equation 4b) is 3
It is configured by a convolver that divides one side into three parts so as to be realized by the FIR filtering process. This corresponds to speaker sets sp1 and sp2 used for sound image localization.
This is an example suitable for a case where a plurality of positional relations between the user and the listener M are prepared.

From the above (Equation 4a) and (Equation 4b), the transfer characteristics cfLx and cfRx of the convolver are cfLx = (h2R · pLx−h2L · pRx) / H (formula 4a) cfRx = (− h1R · pLx + h1L · pRx) / H (〃) where H = h1L · h2R−h2L · h1R (Equation 4b).

Therefore, focusing on the common coefficients pLx and pRx, as shown in FIG.
A pair of convolvers 10 to 15 and a pair of adders 16 and 17 can be used. With this configuration,
These are coefficients of the convolvers 10 and 13. The sound image localization position can be changed only by changing only pLx and pRx. Therefore, even when the sound image is localized at an intermediate position not prepared in the coefficient ROM 3, the coefficients (transfer characteristics) pLx, pR
Only for x, the intermediate transfer characteristic may be calculated and supplied to the convolver by the method described above. Therefore, FF
The burden on the means for calculating the intermediate transfer characteristic, which is composed of the T circuit 6, the intermediate transfer characteristic calculation circuit 7, and the inverse FFT circuit 8, is reduced.

When preparing a plurality of positional relationships between a speaker set used for sound image localization and a listener, a pair of speakers sp1 and sp2 are separated from each other by, for example, 30 degrees left and right around a game operator (listener) M. When corresponding to a case where the speaker is arranged (shown in FIG. 4) and a case where a pair of speakers sp1 and sp2 are arranged narrower and spaced apart by 15 degrees left and right (shown in parentheses in FIG. 4). Then, depending on the actual system installation state, the convolvers 11, 12, 14,
H2R / H, -h2L / H, h1L /
H, -h2L / H may be changed. Also at this time, the intermediate transfer characteristic may be calculated and supplied to the convolver by the method described above for a state (angular position) not stored and stored in the coefficient ROM 5.

The transfer characteristics also differ depending on the individual differences (personal characteristics, for example, head size, ear shape, etc.) of the listener in the measurement data on which the calculation is based. In response to this, for a plurality of states corresponding to individual differences, transfer characteristics of only representative positions of the states (for example, 12 sets every 30 degrees) are prepared, and then, as necessary. By repeating the calculation of the intermediate transfer characteristic, the positional relationship can be finely controlled. On the other hand, on the contrary, the transfer characteristics of only the representative state corresponding to individual differences (for example, only two states of an adult with a large head and a small infant) are prepared, and the necessary Accordingly, it is also possible to calculate an intermediate transfer characteristic between the two, for example, in the case of an elementary school student between an adult and an infant.

Further, utilizing the calculation of the intermediate transfer characteristic, that is, from the transfer characteristic using the representative head transfer characteristic,
It is also possible to calculate a transfer characteristic based on a virtual intermediate head transfer characteristic that cannot be originally obtained by measurement or the like, and to absorb a greater number of individual differences. For example, in the measurement of the head related transfer function (HRTF) in step 101, the dummy head to be measured is merely an abstracted model, and does not necessarily require a generally accepted transfer characteristic. Therefore, a head related transfer function (HRTF) is measured for a plurality of different heads, and measurement data reflecting conditions such as different ears, faces, and heads is obtained for each person, and this is used as a reference transfer characteristic. Calculate the intermediate transfer characteristics. Then, since the calculated intermediate transfer characteristic is a transfer characteristic that is generally accepted in which a large number of individual differences are absorbed, even if the sound image is localized using a game machine or the like, the effect is given to most operators. Thus, an extremely practical sound image localization device can be provided.

As described above, when a plurality of coefficient groups are switched in order to cope with the positional relationship between a speaker set used for sound image localization and a listener and individual differences among the listeners, information on these system configurations is used. (System information)
It is preferable to provide a means for detecting and inputting the information, and to input such system information to the control means 4 so that the coefficient group can be switched.

(Embodiment 3) By the way, in the above embodiment,
In this example, the impulse response (time axis response) data is stored and prepared in the coefficient ROMs 3 and 5 of the apparatus. In this embodiment, as the frequency response, the data is the coefficient RO of the device.
This is an example in which storage is prepared in M23. In the device shown in FIG. 11, the convolution operation circuit includes the windowing means 18,
FFT transformation means 19, multiplication means 20, inverse FFT means 2
1, an adding means 22, which is configured to perform a convolution operation on a signal from a sound source in a frequency domain. To calculate the transfer characteristic at the intermediate position with reference to the basic transfer characteristic stored in the coefficient ROM 23, the intermediate transfer characteristic calculating circuit 7 directly calculates the intermediate transfer characteristic by the same method as in step 106 described above. Calculate the transfer characteristics.
In this device, F for calculating the intermediate transfer characteristic is used.
No FT circuit or inverse FFT circuit is required.

[0056]

As described above in detail, the method for calculating the intermediate transfer characteristic in the sound image localization control according to the present invention uses the reference transfer characteristic based on a plurality of reference transfer characteristics measured or calculated in advance. Since the geometrical average of the amplitude characteristics between the reference transfer characteristics is calculated, and the frequency-phase characteristic is calculated as the phase component of the vector average between the reference transfer characteristics, the intermediate transfer characteristic between the reference transfer characteristics is calculated. Is large, the amplitude of the intermediate transfer characteristic is calculated as a geometric mean as a value close to the actual measurement. Therefore, when the calculated intermediate transfer characteristics are used, deterioration of sound image localization and deterioration of sound quality do not occur. Therefore, the sound image can be localized at an arbitrary position almost continuously over a wide range of 360 degrees using the intermediate transfer characteristic.

Further, since the intermediate transfer characteristic close to the actually measured value is calculated from the reference transfer characteristic, even if the calculation of the intermediate transfer characteristic by calculation is repeated several times, a transfer characteristic that can be used practically is obtained, and actual measurement is required. Transfer characteristics are reduced. Therefore,
The process of collecting and collecting data (transfer characteristics) is reduced. Further, since the capacity of data (transfer characteristics) stored in the coefficient ROM or the like of the sound image localization control device is reduced, the size of the device can be reduced.

Further, by calculating a virtual intermediate transfer characteristic that cannot be obtained by actual measurement from the measured representative transfer characteristics, the transfer characteristics absorbing the differences in the measurement conditions, the individual differences of the listeners, and the like can be obtained. can get. Therefore, by utilizing the calculated intermediate transfer characteristics, the effect of sound image localization does not significantly differ from person to person, and an extremely practical sound image localization control device can be provided.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a method of calculating an intermediate transfer characteristic in sound image localization control according to the present invention, and is a diagram particularly illustrating a principle of calculating an amplitude.

FIGS. 2A and 2B are diagrams for explaining an amplitude of an intermediate transfer characteristic, wherein FIG. 2A is an embodiment and FIG.

FIG. 3 is a chart showing steps of a sound image localization control method using a calculation method of an intermediate transfer characteristic according to the present invention.

FIG. 4 is a basic configuration diagram of a sound image localization apparatus using a method of calculating an intermediate transfer characteristic according to the present invention.

FIG. 5 is a configuration diagram showing a basic principle of sound image localization control.

FIG. 6 is a configuration diagram showing an HRTF (head-related transfer function) measurement system.

FIG. 7 is a diagram illustrating points of HRTF measurement.

FIG. 8 is a diagram illustrating calculation of a cancel filter.

FIG. 9 is a configuration diagram showing a second embodiment of the sound image localization control device using the calculation method of the intermediate transfer characteristic according to the present invention.

FIG. 10 is a configuration diagram showing a third embodiment of a sound image localization control device using a method for calculating an intermediate transfer characteristic according to the present invention.

FIG. 11 is a configuration diagram showing a fourth embodiment of the sound image localization control device using the calculation method of the intermediate transfer characteristic according to the present invention.

12A is a diagram illustrating an arithmetic average when the difference between the phase angles of the reference transfer characteristics is small, and FIG. 12B is a diagram illustrating the arithmetic average when the difference between the phase angles of the reference transfer characteristics is large.

FIG. 13 is an example of frequency-amplitude characteristics of two reference transfer characteristics.

14A is a measured value of a frequency-amplitude characteristic of an intermediate transfer characteristic, FIG. 14B is a frequency-amplitude characteristic of an intermediate transfer characteristic obtained from a reference transfer characteristic by a conventional calculation method, and FIG. It is a frequency-amplitude characteristic of the intermediate transfer characteristic obtained from the characteristic by the calculation method of the present invention.

FIG. 15A is a measured value of the frequency-phase characteristic of the intermediate transfer characteristic, and FIG. 15B is a frequency-phase characteristic of the intermediate transfer characteristic obtained by arithmetic averaging from the reference transfer characteristic.

[Explanation of symbols]

1, 2, 10 to 15 Convolver 3, 5, 23 Coefficient ROM 4 Coefficient supply means 7 Intermediate transfer characteristic calculation means 101 Step of measuring head related transfer function (HRTF) 102 Step of calculating impulse response of HRTF 103 IR (Impulse) Response shaping step 104 Computing the cancel filter (reference transfer characteristic) 105 Computing the intermediate transfer characteristic 106 Scaling the cancel filter 107 Convolving and reproducing the signal from the sound source sp1, sp2 Speakers h1L, h1R Head transfer characteristics from speaker sp1 to listener's left and right ears h2L, h2R Head transfer characteristics from speaker sp2 to listener's left and right ears pLx, pRx An actual speaker is arranged at a desired localization position x. Sometimes Head transfer characteristics to the listener's right and left ears cfLx, cfRx Cancel filter (coefficient of) DM Dummy head (or head) M Listener (game operator, listener) X Sound source x Target sound image localization position X (i) , Y (i) reference transfer characteristic R (i) intermediate transfer characteristic R amplitude r phase (phase angle)

Claims (6)

(57) [Claims] A signal reproduced from a plurality of transducers arranged at a distance and supplied by a plurality of signal conversion circuits having the same sound source and having a predetermined transmission characteristic is reproduced by a plurality of transducers. A method of calculating a transfer characteristic in sound image localization control for causing a sound image to be localized at a different arbitrary position, wherein a frequency-amplitude characteristic is obtained from a plurality of reference transfer characteristics measured or calculated in advance. Sound image localization control, wherein an intermediate transfer characteristic between the reference transfer characteristics is calculated as a phase component of a vector average between the reference transfer characteristics, and a frequency-phase characteristic is a geometric mean of amplitude characteristics between the reference transfer characteristics. Calculation method of intermediate transfer characteristics in. 2. A plurality of transducers arranged at a distance reproduce signals processed by a plurality of signal conversion circuits having the same sound source and having predetermined transmission characteristics, and provide the listener with a signal from the plurality of transducers. In a sound image localization control device that makes it feel as if a sound image is localized at a different arbitrary position, a frequency-amplitude characteristic is calculated by calculating a geometric mean of amplitude characteristics between the reference transfer characteristics from a plurality of reference transfer characteristics measured or calculated in advance. The frequency-phase characteristic has means for calculating an intermediate transfer characteristic between the reference transfer characteristics as a phase component of a vector average between the reference transfer characteristics, and a sound image based on the calculated intermediate transfer characteristic is provided. A sound image localization control device characterized by performing localization control. 3. A signal processed by a plurality of signal conversion circuits provided with the same sound source and having a predetermined transfer characteristic is reproduced from a plurality of transducers arranged at a distance, and the listener is provided with the transducer. What is claimed is: 1. A sound image localization control system configured to make a sound image seem to be localized at a different arbitrary position, wherein a plurality of reference transfer characteristics measured or calculated in advance and the frequency-amplitude characteristics are the reference transfer characteristics. Transfer characteristic storing means for storing a geometric mean of amplitude characteristics between the frequency characteristics and a frequency-phase characteristic as a phase component of a vector average between the reference transfer characteristics and an intermediate transfer characteristic calculated from the plurality of reference transfer characteristics. Means for appropriately selecting a transfer characteristic of the transfer characteristic storage means in accordance with the configured system information and supplying the selected transfer characteristic to the signal conversion circuit. Sound image localization control apparatus being characterized in that so as to sound image localization control according to Temu. 4. A virtual intermediate transfer characteristic is calculated from representative measured transfer characteristics in accordance with the method for calculating an intermediate transfer characteristic in the sound image localization control according to claim 1. A sound image localization control method, wherein sound image localization control is performed based on the sound image localization control. 5. A plurality of transducers arranged at a distance reproduces signals processed by a plurality of signal conversion circuits having the same sound source supplied thereto and having predetermined transfer characteristics. A method of calculating a transfer characteristic in sound image localization control that makes a sound image appear to be localized at a different arbitrary position, wherein a frequency-amplitude characteristic is obtained from a plurality of reference transfer characteristics measured or calculated in advance for different sound image localization positions. Is the geometric mean of the amplitude characteristics between the reference transfer characteristics, and the frequency-phase characteristic is the phase component of the vector average between the reference transfer characteristics, and calculates an intermediate transfer characteristic for an intermediate sound image localization position of the reference transfer characteristics. A method for calculating an intermediate transfer characteristic in sound image localization control, characterized in that: 6. A method for calculating an intermediate transfer characteristic in the sound image localization control according to claim 5, wherein the intermediate transfer characteristic is calculated on the basis of the measured reference transfer characteristic for each 360 / M (M: an integer of 2 or more) degrees. A sound image localization control method, wherein a transfer characteristic is calculated N times (N: an integer equal to or greater than 1) and sound image localization control is performed every 360 / (M · ^2N ) degrees.