JP3197077B2 - Localized sound image generator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a localized sound image generating apparatus for generating a sound image localized at a predetermined position in space.

[0002]

2. Description of the Related Art In a conventional sound image localization apparatus for localizing a sound image at a predetermined position, a technique for localizing a sound image using a head-related transfer function representing a sound field is known. This technique will be described below. First, how sound generated from a sound source at a predetermined position in space is transmitted to a listener, that is,
Measure the HRTF. As this measurement, a so-called impulse response method in which an impulse generated from a sound source at a predetermined position is captured by small microphones installed in the right and left ears of a listener is preferable.

[0003] An FIR filter is prepared by using each amplitude value of the measured impulse response waveform as a coefficient, and a pair of left and right is used as a set to produce a convolution device for performing a convolution operation of a sound image localization. A convolution operation is performed on the obtained waveform data. Next, by hearing the convoluted waveform data independently with both ears using headphones or earphones, the listener can obtain a sense that the sound image of the sound being heard is localized at a predetermined position. Can be.

[0004]

In the above-described conventional sound image localization apparatus, the directional device for performing the convolution operation is composed of an FIR filter. However, this FIR filter requires enormous processing (several hundred times). ~ Thousands of multiplications)
(E.g., one sampling clock of 20 μs or less) must be an expensive circuit, which has contributed to an increase in manufacturing cost.

Further, in the above-described sound image localization apparatus, in order to obtain a sufficiently clear sense of localization, it is necessary to provide the directing devices densely corresponding to positions that sufficiently subdivide a space that can be a sound source position. For this reason, the circuit scale tends to be huge. The present invention has been made in view of such a background, and provides a localized sound image generating apparatus which has an inexpensive and small-scale circuit configuration and can realize a high-quality localization feeling and a feeling of movement of a sound image. With the goal.

[0006]

According to a first aspect of the present invention, there is provided a localized sound image generating apparatus, wherein a storage means for storing a plurality of waveform data corresponding to a plurality of predetermined sound image positions in a space, and a position in the space. Position specifying means for reading waveform data from the storage means; and waveform data generating means for selectively generating waveform data corresponding to the position specified by the position specifying means . It is characterized by having.

According to a second aspect of the present invention, there is provided a localized sound image generating apparatus, comprising: storage means for storing a plurality of waveform data corresponding to a plurality of predetermined discrete sound image positions in a space; Means for reading waveform data from the storage means, and comprising weighting and adding means for weighting and adding to the plurality of waveform data read according to the position specified by the position specifying means . And

[0008] A localized sound image generating apparatus according to claim 3.
Is a localization sound image generating apparatus according to claim 1 or 2.
The storage means is vacant for each of the plurality of sound image positions.
Stored waveform data corresponding to a plurality of predetermined positions between
It is characterized by: The localization sound image generating apparatus according to claim 4 , wherein the storage means stores a plurality of waveform data corresponding to a plurality of positions in the vertical direction of the space; a position specifying means for specifying a position in the space; a from the storage unit and reads the waveform data, the waveform data generating means for generating a selectively waveform data in accordance with the component of the vertical angle direction of the specified position by the position specifying means, specified by the position specifying means depending on the horizontal angle direction component of position, characterized by comprising a conversion means for providing a delay time difference in the right channel component and the left-channel component and the amplitude difference of the waveform data generated by said waveform data generating means It is set to.

[0009]

SUMMARY OF] According to the configuration of claim 1, Te waveform data generating means smell, corresponding to a predetermined plurality of sound image position in space
A plurality of waveform data stored in the Te SL憶means, position
Waveform data corresponding to the position of the specified me by the location designation means space is output is read out to select specific. Further, according to the configuration of the second aspect, the weighting and adding means records a plurality of predetermined discrete sound image positions in the space.
From multiple waveform data stored in the憶unit, position the finger
Multiple of wave in accordance with the position of the designated space I'm in constant means
The shape data is read out and weighted and added, thereby obtaining waveform data corresponding to positions between discrete sound image positions.

[0010] According to a third aspect of the present invention , a contractor is provided.
The configuration according to claim 1 or 2, wherein a plurality of sound image positions
Waveform data corresponding to a plurality of predetermined positions in the space.
Data is stored in the storage means. Further, according to the configuration of claim 4, a plurality of waveform data stored in the memorize means in response to the vertical angle direction of the space, the position finger
Waveform data corresponding to a component of the vertical angle direction of the designated position by the constant means, selectively in the waveform data generating means
It is read out to. Next, in the conversion means, a delay time difference and an amplitude difference are given to the right channel component and the left channel component of the selected waveform data according to the horizontal angle component at the designated position.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A localized sound image generating apparatus according to an embodiment of the present invention stores waveform data of a plurality of sampled waveforms corresponding to a vertical angle .theta. Since the sense of localization in the horizontal angle ψ direction can be sufficiently expressed by the amplitude difference and the delay time difference between the R / L channel components of the waveform data, the sample waveform corresponds to only the vertical angle θ. First, a process of collecting waveform data of a sample waveform will be described with reference to FIG.

In FIG. 5, a horizontal angle ψ from the origin O,
The sound source P specified by the vertical angle θ and the distance D is assumed to exist on a spherical surface S having a radius r (r = D) centered on the origin O. DH is a dummy head having a shape imitating a human head placed at the center thereof at the origin O, and a microphone corresponding to the right and left ears for sampling a sound generated from the sound source P. MR and ML are attached.

The points A to G are points on the line of intersection of the spherical surface S with a plane specified by the origin O, the front direction and the vertical direction of the dummy head DH, and are arranged at equal intervals.
The vertical angle θ with respect to the point A is 0 °, the vertical angle θ with respect to the point B is 30 °, and so on.
Is 180 °. Reference numeral 1 denotes a waveform memory for storing waveform data of sounds collected by the microphones MR and ML.

In such a configuration, when a sound is generated by playing a predetermined musical instrument, for example, a trumpet, from a sound source P arranged on one of the points A to G, the generated sound is transmitted through the air, Microphones MR and M of dummy head DH
L and converted to electrical signals. When this electric signal is subjected to A / D conversion by an A / D converter (not shown), the electric signal is converted into digital data and stored in the waveform memory 1. That is, the entire waveform is sampled, and the longer the sampling time is, the better (of course, the length is less than the limit of the memory capacity). Here, the same sound is generated for each of the points A to G, and sampling is performed for about three seconds each. As described above, the process of collecting the waveform data of the sample waveform ends.

Next, an embodiment of the present invention based on the waveform memory 1 created through the above process will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a localized sound image generating apparatus according to an embodiment of the present invention. In FIG. 1, reference numeral 1 'denotes a waveform memory storing the same waveform data as the waveform memory 1. The waveform data of the timbre corresponding to the timbre data ST supplied from the keyboard or the like is output. Here, the output waveform data is seven series of waveform data corresponding to points A to G.

Reference numeral 2 denotes a localization control device for setting a position at which the sound image is to be localized (hereinafter, referred to as a target sound image position). As shown in FIG. 2, the localization control device 2 comprises an instruction unit 2a and a control unit 2b. I have. Further, the pointing unit 2a rotates the vertical dial 2a to specify a vertical angle.
₁ and, by rotating, the horizontal angle dial 2a ₂ that specifies the horizontal angle, by sliding the knob, and a distance slider 2a ₃ to specify the distance. Then, the vertical angle θ, the horizontal angle す る, and the distance D specifying the target sound image position are supplied to the control unit 2b.

[0017] Here, in the present embodiment, since in correspondence with the upper hemisphere points A-G, a vertical angle dial 2a ₁ is rotatably set in the 0 ° to 180 °. In addition,
When sound is generated and sampled at a position below the dummy head DH, that is, when the whole sphere is used, the rotation is freely set in the range of 0 ° to 360 °, that is, any number of revolutions. Further, the horizontal angle dial 2a ₂ is as free rotation at 0 ° to 360 °.

Reference numeral 3 denotes seven series of waveform data (serial data) read out from the waveform memory 1 'as a series of data (serial data) by time-division processing. Demultiplexers 4a to 4g are multipliers for multiplying the waveform data output from the demultiplexer 3 by the corresponding multiplication coefficients G _{A to} G _G , and the multiplication coefficients G _{A to} G
_G is a control unit 2b of the localization control device 2 as shown in FIG.
Is controlled according to the vertical angle θ supplied from the controller.

Next, 5 is an adder for adding the seven series of waveform data output from the multipliers 4a to 4g and outputting one waveform data. As shown in FIG.
A multiplier that multiplies the waveform data output from the adder 5 by multiplication coefficients a and b corresponding to the distance D supplied from the control unit 2b. The multiplier 7 converts the waveform data output from the multiplier 6a into R
/ L channel components, each of which is a delay unit for delaying. Next, 8a and 8b are multipliers for multiplying the waveform data of the R channel component and the L channel component output from the delay unit 7 by multiplication coefficients c and d corresponding to the horizontal angle ψ supplied from the control unit 2b. is there.

A reverberation generator 9 separates the waveform data output from the multiplier 6b into R / L channel components, and generates reverberation data corresponding to each component.
The adder 10b that adds the R-channel waveform data output from a and the R-channel reverberation data output from the reverberation generator 9 includes the L-channel waveform data output from the multiplier 8b and the reverberation. The adder adds the L-channel reverberation data output from the generation unit 9.

Next, reference numeral 11 denotes a crosstalk canceller serving as an inverse characteristic circuit of crosstalk generated in accordance with the positional relationship between the speaker and the listener in the reproduction space, and an adder 10a
And the waveform data of the R / L channel components output from 10b are subjected to processing for canceling crosstalk in advance. Reference numeral 12 denotes a D / A converter for converting waveform data (digital data) output from the crosstalk canceller 11 into an audio signal (analog signal). Reference numeral 13 denotes a D / A converter 1.
2 is an amplifier that amplifies the sound signal converted in 2 and supplies it to left and right speakers (not shown).

In such a configuration, the sound of the timbre at the target sound image position specified by the vertical angle θ ₁ , the horizontal angle ψ ₁ , and the distance D ₁ can be localized by the operation described below. For example, when the timbre data ST is given to the waveform memory 1 'by an operation such as pressing a key (not shown),
Waveform data of seven series of timbres corresponding to the timbre data ST is output as time-division serial data. Thereafter, in the demultiplexer 3, the time-division serial data output from the waveform memory 1 ′ is converted into 7-series parallel data, and each waveform data is converted into a multiplier 4 a corresponding to the corresponding points A to G. ~ 4g.

The vertical angle θ ₁ and the horizontal angle ψ ₁ of the target sound image position are determined by rotating the vertical angle dial 2a ₁ and the horizontal angle dial 2a ₂ of the indicator 2a. The distance D ₁ is determined by sliding the distance slider 2a _3. The vertical angle θ ₁ , the horizontal angle ψ _1, and the distance D ₁ are transmitted to the control unit 2 b, and the control unit 2 b
b, the multipliers 4a to 4g have multiplication coefficients G _{A to} G _G corresponding to the vertical angle θ ₁ , and the multipliers 6a and 6b have
Distance D multiplication factor a and b corresponding to _1, the multiplier 8a and 8b, the multiplication coefficients c and d corresponding to the horizontal angle [psi ₁
Are supplied respectively.

Here, the relationship between the multiplication coefficients G _{A to GG} and the vertical angle θ ₁ is as shown in FIG. In FIG. 3, for example, when the vertical angle θ _{1 is} 30 °, in the multiplication coefficients G _{A to} G _G , only GB is 1 and the other multiplication coefficients are 0. Next, the vertical angle operation of the dial 2a ₁ of the indicator 2a, if the vertical angle theta ₁ is changed to greater than 30 °, the multiplication factor G _B becomes small, the multiplication coefficient G _C becomes larger. At this time, the other multiplication coefficients remain at 0. Further, the sum of the multiplication coefficient G _B and G _C are adapted to be 1, the vertical angle θ
_{When 1} = 45 °, G _B = G _C = 0.5, and the waveform data at the position where no waveform data is collected can be synthesized by interpolation of the collected waveform data.

Next, the multiplier 4a-4g, the waveform data multiplied with the multiplication coefficient G _A ~G _G are summed in adder 5 is supplied to the multiplier 6a and 6b. The multipliers 6a and 6b, as shown in FIG. 4, the multiplication coefficient distance D ₁ is small as becomes large and,
Conversely, a large multiplication coefficient b is supplied from the localization control device 2. Therefore, in the multiplier 6a, the waveform data multiplied by the multiplication coefficient a is input to the delay unit 7,
Channel and L channel components,
And 8b.

[0026] multiplier 8a and 8b, when the horizontal angle [psi ₁ is near 90 °, the multiplication factor of the multiplier 8a as a major c,
On the other hand, a multiplication coefficient d that becomes smaller is supplied from the localization controller 2. Therefore, multipliers 8a and 8b
Are input to the adders 10a and 10b. Further, in the multiplier 6b, the waveform data multiplied by the multiplication coefficient b is input to the reverberation generation unit 9, separated into R channel and L channel components, and processed into reverberation data. next,
The reverberation data of the R channel and L channel components is added to multipliers 8a and 8b in adders 10a and 10b.
Are added to the waveform data supplied from the crosstalk canceller 11 and supplied to the crosstalk canceller 11.

Next, in the crosstalk canceller 11, the crosstalk of the audio signals of the R / L channel components output from the adders 10a and 10b is canceled in advance. After that, the signal is converted into an analog sound signal in the D / A converter 12 and supplied to the amplifier 13. In the amplifier 13, the acoustic signal is subjected to predetermined amplification and output to left and right speakers (not shown). As described above, the ratio of the localization volume and reverberation data of the distance D1, the amplitude difference and the delay time difference between the localization of the horizontal angle [psi ₁ right channel component, the localization of the vertical angle theta ₁ is a waveform from the waveform memory 1' This is performed by reading data and weighting.

In the above-described embodiment, all the waveform data of seven sequences are read at the same time, but only two types of waveforms necessary for interpolation may be read. In such a case, when another waveform data is required due to the movement of the designated sound image position, interpolation may be performed by reading the waveform data from an intermediate address. Further, a well-known loop process may be performed on the waveform data stored in the waveform memory 1 '. This is particularly effective when the capacity of the waveform memory 1 'is small and a sufficiently long sampling time cannot be secured.

Further, although the waveform data corresponding to only the front vertical direction is used, the waveform data corresponding to positions finely distributed over the whole sky may be used. Moreover, when making it correspond to a perpendicular direction, it does not necessarily need to correspond to a front perpendicular direction. When storing a waveform corresponding to a vertical direction deviated from the front, the delay difference and the amplitude difference given by the horizontal angle may be corrected.
Further, since an example in which a speaker is used as a device for converting an acoustic signal into sound has been shown, the crosstalk canceller 11 has been described.
However, if listening using headphones, the crosstalk canceller 11 is not required.

[0030]

As described in detail above, claim 1 is as follows.
According to the invention described in (1), the waveform data is stored in correspondence with a plurality of predetermined sound image positions in the space, and the waveform data is selectively generated in accordance with the designated space position. As a result, a clear localized sound image can be obtained with a small-scale configuration.

According to the second aspect of the present invention, the waveform data is stored corresponding to a plurality of predetermined discrete sound image positions in the space, and the waveform data is stored corresponding to the designated space position. Since a plurality of data are read out and weighted and added, waveform data corresponding to positions between discrete sound image positions can be obtained. By doing so, the sound image position can be moved smoothly by smoothly changing the weighting value given to the plurality of waveform data.

Further, according to the invention described in claim 4, stores the waveform data corresponding to a predetermined plurality of vertical angle directions in space, in accordance with the vertical angle direction component for the specified spatial position The waveform data is selectively generated, and the delay time difference and the amplitude difference between the left and right channels are given according to the horizontal angle direction component of the designated space position. The amount of data to be stored can be smaller than when storing waveform data corresponding to sound sources at all positions in the space.

Since the positions in the vertical direction are symmetrical with respect to both ears, it is necessary to be quite strict to obtain a clear sense of localization. Therefore, the waveforms are directly matched. On the other hand, in the horizontal angle direction, waveforms are omitted because a sufficient localization feeling can be obtained only by providing a signal characteristic difference of a delay difference and an amplitude difference to both ears. That is, it is possible to reduce the amount of data to be stored while maintaining a sufficient sense of localization in both the vertical and horizontal directions.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a localized sound image generation device according to an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a schematic configuration of a localization control device 2.

FIG. 3 is a diagram illustrating a relationship between a vertical angle θ and multiplication coefficients G _{A to} G _G.

FIG. 4 is a diagram showing a relationship between a distance D and multiplication coefficients a and b.

FIG. 5 is a diagram for explaining a method of collecting waveform data from the waveform memory 1;

[Explanation of symbols]

1 ': waveform memory, 2: localization control device, 4a to 4g
... Multiplier, 5 ... Adder, 6a, 6b ... Multiplier, 7
... Delay unit, 8a, 8b multiplier, 9 reverberation generation unit.

Claims (4)

(57) [Claims] 1. A storage means for storing a plurality of waveform data corresponding to a plurality of predetermined sound image positions in a space, a position specifying means for specifying a position in a space, and reading the waveform data from the storage means. hand,
A localized sound image generating apparatus, comprising: waveform data generating means for selectively generating waveform data corresponding to a position specified by the position specifying means. 2. A storage means for storing a plurality of waveform data corresponding to a plurality of predetermined discrete sound image positions in a space; a position designation means for designating a position in a space; and a means for reading waveform data from the storage means. And
A localized sound image generating apparatus, comprising: weighted addition means for weighting and adding a plurality of waveform data read according to the position designated by the position designation means. 3. The storage device according to claim 2, wherein the storage unit stores the plurality of sound image positions.
Waveform data corresponding to a plurality of predetermined positions in the space
Remembered 3. The method according to claim 1, wherein
Tone image generator. 4. A storage means for storing a plurality of waveform data corresponding to a plurality of positions in a vertical direction of a space, a position designation means for designating a position in a space, and reading the waveform data from the storage means. So,
A waveform data generating means for selectively generating waveform data according to a vertical component of the position specified by the position specifying means; and a horizontal angular component of the position specified by the position specifying means. And a conversion means for providing a delay time difference and an amplitude difference between a right channel component and a left channel component of the waveform data generated by the waveform data generation means.