JPH05303381A - Acoustic image localizing device - Google Patents

Abstract

PURPOSE:To distinctly localize an acoustic image with a simple constitution by reading out impulse response waveforms from independent acoustic image localizing positions (points) by right and left channels to set them to an output means. CONSTITUTION:A pan circuit 21 delays an inputted audio signal to process it into binaural signals, and a music signal is divided into binaural signals (signals of two channels L and R). Signals of respective channels are converted to analog signals by D/A converters 22L and 22R and are outputted. When a sound source is placed at a point P and both ears of a listener are placed in positions eL and eR, no-signal times tauL and tauR are generated after generation of an impulse in accordance with distances to ears. The pan circuit 21 realizes the no-signal times by a delay circuit and uses FIR filters to express the response waveforms in sections from arrival of direct sounds to attenuation of indirect sounds. Thus, the pan circuit 21 expresses times tauL and tauR by one system delay circuit and expresses these sections by right and left filters.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound image localization device capable of correctly localizing a sound image of a musical sound at an arbitrary position with a small number (about two) of speakers.

[0002]

2. Description of the Related Art When a human being hears an audio signal such as a musical sound, a person can perceptually recognize the position of the sound source. This is based on the difference in the characteristics of the signals reaching the left and right ears. That is, when the sound source is located at a position deviated from the front of the listener to the left and right, the time to reach the left and right ears is deviated. Further, there is a difference in acoustic characteristics (impulse response) between the case where the sound source is in front of the listener and the case where the sound source is behind it. This is due to the effects of the skull, ear lobes, and the like.

[0003]

However, the conventional electronic musical instruments and the like do not have a function for realizing the above-mentioned characteristics for localizing the sound image of the sound source, and simply control the left and right volume levels. Met. Therefore, sound image localization was unnatural.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a sound image localization device having a simple structure and capable of clearly localizing a sound image.

[0005]

SUMMARY OF THE INVENTION According to the present invention, a table in which impulse response waveforms from a plurality of points are separately stored for left and right ears and an input musical tone signal is realized as a predetermined impulse response waveform. Output means for outputting to the ears, and setting means for reading out impulse response waveforms from different points independently from the table and setting them in the output means.

[0006]

In the sound image localization apparatus of the present invention, the sound image is localized by reproducing the impulse response at the sound image localization position. Normally, the sound image is localized by reading the impulse response waveforms from the same localization position for both left and right channels (both ears), but in the present invention, the impulse response waveforms from different sound image localization positions (points) are read and output for the left and right channels. Set to means. This makes it possible to realize a unique sense of localization that has never existed before.

The present invention can also be configured as follows. That is, a plurality of delay circuits for inputting musical tone signals of the same sound source and delaying each for a predetermined time,
A sound image localization apparatus including a plurality of systems of FIR filters to which a plurality of signals extracted from respective delay circuits are input, and the delay circuit includes a blank time until a direct sound in the impulse response of the sound source arrives. The FIR filter has a characteristic to realize, and the FIR filter has a characteristic to realize a response characteristic until the reached direct sound is attenuated in the impulse response of the sound source. In this case, by realizing the blank time until the direct sound arrives with the delay circuit,
The number of stages of the FIR filter can be significantly reduced, and the configuration can be simplified.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is an overall block diagram of a sound image localization apparatus which is an embodiment of the present invention. It is a device that outputs an input audio signal as a binaural signal so as to be localized at a predetermined position. The operation of this device is controlled by the CPU 10. ROM 12, R to CPU 10 via bus 11
AM13, panel switch 14, indicator 15, MIDI
The interface 16 and the interface 17 are connected.

In addition to the control program, the ROM 12 stores the transfer function to be given to the FIR filter described later.
The panel switch 14 includes a switch group for inputting the localization position (left and right channels, distance, direction) of the sound image. A MIDI controller (electronic musical instrument) such as a keyboard is connected to the MIDI interface 16. When this MIDI controller is used as a man-machine interface, a control signal is input via this MIDI interface 16. A pan circuit 21 is connected to the interface 17. Bread circuit 21
Is a circuit that delays, reverberates, etc. the input audio signal and processes it into a binaural signal that is localized at a predetermined position. The A / D converter 20 is connected to the pan circuit 21. The A / D converter 20 is a circuit that converts a plurality of series of audio signals input from an external input terminal into digital signals. In the pan circuit 21, the tone signal is divided into binaural signals (2 channel signals of L and R).
The signals of each channel are respectively D / A converters 22L and 2L.
It is converted into an analog signal by 2R and input to the sound systems 23L and 23R. Sound system 23
The L and 23R amplify this signal and output it from a speaker (headphone).

Assuming that the sound source S is at the point P and both ears of the listener are at the positions eL and eR as shown in FIG.
The impulse response from the sound source S to both ears (eL, eR) is
For example, it becomes like FIG. 3 (A), (B). In this way, signalless times (time until direct sound reaches the ears) τL and τR occur depending on the distance from the impulse to the ear. In the pan circuit 21, this signalless time is realized by the delay circuit 50.

FIR filters 51 and 5 are used for the sections ab and cd from the arrival of the direct sound to the attenuation of the indirect sound.
2 is used to represent the response waveform. In the pan circuit 21, as shown in FIG. 4, the sections of .tau.L and .tau.R are represented by the delay circuit 50 of one system, and the sections of ab and cd are represented by the FIR filters 51 and 52 of the left and right systems. There is.

FIG. 4 is a diagram showing a basic configuration of the pan circuit 21. The delay circuit 50 can be configured by, for example, a plurality of stages of shift registers each including a read tap for each stage. The CPU reads the input signal from the taps corresponding to the delay times τl and τr of the left channel and the right channel, and inputs them to the FIR filters 51 and 52. The number of taps n is calculated by nL = (τL / c) sf nR = (τR / c) sf (c: sound velocity, sf: sampling frequency).

FIG. 5 is a diagram showing a filter coefficient table stored in the ROM 12. The filter coefficient is
F that represents the impulse response for localizing the sound image to a certain point
It is a transfer function of the IR filter. The table is provided for each of the left and right channels (HR, HL). The coefficient uses another coordinate system (x-yR, x-yL) whose origin is the ear for each of the left and right channels (ears). For both the left and right channels, 30 ° from the front for each short distance D1 and long distance D2 (P0 = P′0 = 0 °, P1 = P′1
= 30 °, P2 = P′2 = 60 °, ..., P11 = P′1
24 coefficients are stored for each of 12 angles (1 = 330 °). Here, both Pn and P'n are arguments representing an angle, but since the coordinate system (origin) is different between the left and right ears, one is attached with ('). Based on the operation of the panel switch 14 and the control signal input from the MIDI interface 16, one for each of the left and right channels is read out and set in the FIR filter. Unique effects can be obtained by reading the transfer function for localizing the sound image to another position for each of the left and right channels.

FIG. 6 is a diagram showing the configuration of the pan circuit 21. The pan circuit 21 is obtained by adding a volume control circuit and a reverb circuit to the basic configuration shown in FIG.
The pan circuit 21 has a plurality of similar circuits (modules) connected in parallel. One or a plurality of tone signals distributed by the selector 30 are input to each module. When a plurality of tone signals are input to one module, the plurality of signals are added and output in the selector 30.

Two FIR filters 32 and 33 and a reverb circuit 40 are externally attached to the module 31, and a delay circuit 34 and filters 35, 36 and 3 are provided.
7, multipliers M1 to M5, and adders A1 to A5 are incorporated. The delay circuit 34 is composed of a delay line having two or more variable taps. The signal input to the delay circuit 34 is delayed by a predetermined time and then output as OL and OR. The OL and OR are FIR filters 32 and 3
Input to 3. The delay time of the delay circuit 34 is determined by the CPU 10 based on the control data. FIR filter 3
Predetermined transfer functions are set at 2 and 33. The signals MIR and MI that have undergone the convolution operation by this transfer function
L is input to the filters 36 and 37. Filter 36,
37 is composed of a low-pass filter of the 1st to 2nd order,
It has the function of determining the frequency characteristics of the left and right direct sounds due to air or a shield. M2 and M3 are filters 36,
37 is a multiplier for controlling the gain of 37.
The multiplier is input from the CPU 10.

Here, a tone signal is directly input to the filter 35 from the selector 30. This filter 35 → multiplier M1 → adder A1 → reverb circuit 40 → multiplier M4
The system from M5 to the adders A2 and A3 is a system that adds an indirect sound to the tone signal. That is, it is a circuit that represents the sound that is diffusely reflected in space and remains as reverberation. For this reason, the filter 35 is composed of a 1st to 2nd order low-pass filter that filters a low-pitched sound portion that is less attenuated by irregular reflection. M1 is a multiplier for controlling the gain of the filter 35. The multiplier is input from the CPU 10. A1 is an adder that adds the F0 output (indirect sound output before reverb is added) other than this module (module A) to the indirect sound of the module A and outputs it to the reverb circuit 40.
That is, the indirect sound is a mixed sound due to a decrease in separability of a plurality of sound sources due to diffused reflection or the like, and is a circuit for expressing this. The reverb circuit 40 gives reverberation time due to diffuse reflection. M4 and M5 are multipliers for controlling the amount of indirect sound given to the left and right channels. Multiplier is C
It is given from PU10.

This indirect sound is added by the adders A2 and A3.
It is added to the musical sound of the direct sound system (system that has passed through the FIR filters 32 and 33). After this, adders A4 and A5
The musical tone signals (OUTL, OUT
R) is added, and D / is obtained as synthesized sounds OUTL and OUTR.
It is output to the A converters 22L and 22R. The OUTL and OUTR of the module B are the OUT of the module C.
Since L and OUTR are added and OUTL and OUTR of the lower modules are sequentially added in the following, OUTL and OUTR of module A are the combined outputs of all modules.

Selector signal, the number of L and R stages of delay,
The filter coefficient of each filter and the coefficient of each multiplier are
As described above, it is supplied from the CPU 10 via the interface. In the above circuit, all modules are FI.
Two R filters and a reverb circuit are provided.
The IR filter and the reverb circuit may be operated in a time-sharing manner so that a plurality of modules can be used in common. Furthermore, the outputs of the modules may not be sequentially added, but all outputs may be added at one time.

7 and 8 are flow charts showing the operation of the same sound image localization apparatus. FIG. 7 shows the main routine. When the power of the device is turned on, first, the initialization operation is executed (n1). The initialization operation includes resetting registers, reading preset data, and the like. Next, the panel processing (n2) and other processing (n3) are repeatedly executed.

FIG. 8 is a flow chart showing the panel processing operation. First, the mode is judged (n10). In the case of mode 1 (normal mode: sound image is localized at one point on a plane), the setting of the distance and direction of the position where the sound source (sound image) is desired to be localized is accepted (n11). The input distance / direction is up to the center of the face (origin O in FIG. 2). The distance / direction (rL, θ) from the sound source to the left and right ears (eL, eR in FIG. 2) based on the input distance and direction.
L, rR, θR) is calculated (n12). On the other hand, in the case of mode 2 (sound images are localized at different positions on both left and right channels), the sound source is localized at the distance and direction from the position where the sound source is localized at the left channel to the left ear and at the right channel. Input of the distance and direction from the position to the right ear is accepted (n13). After this, proceed to n14.

At n14, the number of OL and OR stages of the delay circuit 34 and the multipliers M2 and M3 are calculated from the set distance. The multipliers M2 and M3 are calculated based on the distance rL (rR). When rL ≦ D0 (D0 is the middle point of D1 and D2: see FIG. 2), HL1 (θL) × (D1 / rL) is calculated, and when rL> D0, HL1 (θL) × (D2 / r
L) ² is calculated. Further, the coefficients of the left and right filters are read out based on the distance / direction (n15). Various data obtained by the above operation are sent to the pan circuit 21 (n16). Various data are input to the designated module, the FIR filter of the module, and the reverb circuit. After that, other panel processing (n17) is executed and the process returns. Other panel processing includes setting operations such as module selection, direction / distance setting, data calculation, transmission, and selector control (which series is output to which module).

As described above, in this embodiment, since two types of transfer functions are provided according to the distance from the sound source to the ear,
A sense of distance in acoustic characteristics can also be realized. Further, the number of stages may be increased by further subdividing not only two types of short distance / long distance. Also, by allowing the sound image to be localized at different positions on the left and right channels, it is possible to produce unique effects not found in natural musical instruments.

In this embodiment, both the left and right signals are passed through the delay circuit 34 according to the distance to the sound source. However, the difference between the delay times of the left and right channels is calculated, and only the signal of the long delay channel is subjected to the difference time. You may make it delay.

Further, although the filters for controlling the characteristics of the direct sound in the pan circuit of FIG. 6 are separately provided for the left and right channels (36, 37), one filter is provided before the separation into the left and right channels (before the delay 34). The system may be provided.

A crosstalk cancel circuit may be provided between the pan circuit and the D / A converter. Furthermore, although two-dimensional localization has been described in this embodiment, it is also possible to apply it to three-dimensional localization.

[0026]

As described above, according to the present invention, since the impulse responses from the points (sound image localization positions) for the left and right channels are reproduced, it is possible to realize a unique sound image localization sensation that has never existed before.

[Brief description of drawings]

FIG. 1 is a block diagram of a sound image localization apparatus that is an embodiment of the present invention.

FIG. 2 is a diagram illustrating a localization function of the same sound image localization device.

FIG. 3 is a diagram illustrating a localization function of the same sound image localization device.

FIG. 4 is a diagram illustrating a localization function of the same sound image localization device.

FIG. 5 is a diagram showing an FIR filter coefficient table stored in the same sound image localization apparatus.

FIG. 6 is a configuration diagram of a pan circuit of the same sound image localization device.

FIG. 7 is a flowchart showing the operation of the same sound image localization apparatus.

FIG. 8 is a flowchart showing the operation of the same sound image localization apparatus.

[Explanation of symbols]

21-pan circuit, 34-delay circuit, 32, 33-F
IR filter.

Claims (1)

[Claims] 1. A table in which impulse response waveforms from a plurality of points are separately stored for the left and right ears, and an output means for outputting a signal of an input musical sound to the left and right ears by realizing a predetermined impulse response waveform. A sound image localization apparatus comprising: a setting unit configured to read out impulse response waveforms from different points independently from the table and set the output response unit to the output unit.