JPH03274998A - Sound field reproducing method - Google Patents

Abstract

PURPOSE:To facilitate a system in which a listener has an experience sound in almost the same state as if the listener were resident in a room to be realized by using a transfer function for each speaker and a signal from a sound source and generating a signal to be reproduced from a relevant speaker. CONSTITUTION:When an arrangement beta of the system comprising N channels is given, a signal relating to direction information is distributed to a speaker whose reproduction area is represented in the direction information to generate a transfer function of the speaker. The transfer function of each speaker is generated from the transfer function of a direction transmission number generating circuit 7 including the direction information of a specific sound field depending on the number and arrangement of speakers 3-1 to 3-N of the reproduction system. Even when the number and arrangement of the speakers of the reproduction system are not match with the number of sound incoming directions and how to decide the area, the listener can has an experience in the same state as if the listener were resident in a room condition to be realized.

Description

[Detailed Description of the Invention] "Industrial Application Field" This invention is mainly used in audio playback and voice communication audio reproduction, in which a sound source is reproduced using a plurality of speakers to simulate the sound field of a specific room. This relates to a sound field reproduction method that reproduces a sound field.

``Prior art'' When listening to music through audio reproduction, to get the feeling of being in an actual concert hall, or when listening to sounds from a remote location through communication,
In order to create an atmosphere as if you were in an actual conference room at a remote location, it is necessary to reproduce those indoor sound fields with the sound field where the listeners are actually present. To achieve this, a technique is required that allows signals that arrive at the listener to be delayed due to the echoes in the room to be radiated from the direction of arrival within the room where the listener is actually present. The means to achieve this is the multi-channel reproduction method.

FIG. 6 shows the configuration of an apparatus for a conventional sound field reproduction method.

The sound source signal from the sound source 1 is passed through N convolution operators 2-1 to
2-N, and the respective outputs of these computing units 2-1 to 2-N are supplied to speakers 3-1 to 3-N, respectively. The speakers 3-1 to 3-N are arranged in the reproduction chamber 4, and the reproduction chamber 4
Listener 5 is inside.

The sound source 1 outputs as a signal a dry source that does not include the reflected sound of the indoor sound field to be reproduced. Based on the number N of speakers 3-1 to 3-N given to the reproducer 4 and their arrangement, the structure of the reflected sound of the indoor sound field to be reproduced is allocated to the N speakers. Specifically, the time-domain transfer function of the indoor sound field to be reproduced for each direction of arrival up to the listening point is stored in the convolution calculator 2-n (n = 1 +...N), and this time-domain transfer function is performed. By reproducing the output signal obtained by convolving the function and the output signal of the sound source l through the corresponding speakers 3-n (n-1,...N), the listener can hear the sound in the room to be reproduced. You can experience sound almost as if you were there. The output signal of the sound source 1 is x(L), and the time domain transfer function to the listening point of the indoor sound field to be reproduced for each direction of arrival is h(n, L) (n=1.
...N, N: number of speakers, t=0. ..., ↑-t
, 'r: sampled delay time length), the signal applied to the speaker 3-n (n = 1 + . . . N>) is expressed as follows.

Strictly speaking, the reproduction room 4 itself must be an anechoic room with no reflected sound, and the speakers 3 must be placed across all directions of arrival of the reflected sound in the room to be reproduced, but it is possible to approximate it to some extent. Hopefully it will be effective.

[Problem to be solved by the invention] In this type of system, a common source called a dry source can be used, but the time domain transfer function to the listening point of the indoor sound field to be reproduced for each direction of arrival depends on the arrangement of the speakers. It is not practical because it has to be created according to the size and number of pieces.

The purpose of this invention is to provide a sound field reproduction method that uses multiple speakers to reproduce a sound source and reproduce a sound field. An object of the present invention is to provide a sound field reproduction method that allows you to experience sound in almost the same state as if you were in the room.

"Means for Solving the Problem" According to the present invention, direction information indicating the arrival direction of the sound is provided for each time step (sample point) of the transfer function of a unique sound field, and the direction information is used as a reproduction area. A transfer function is created for each speaker by distributing the value (sound volume) of each time step of the transfer function to each speaker, and the sound is reproduced from the corresponding speaker using the transfer function for each speaker and the sound source signal. Generate the power signal.

``Example'' Once the sound source placed in the sound field to be reproduced and the position of the listener are determined, the sound field is determined by the transfer function from the sound source to the listener through various routes, and from which direction of the listener each route is. The transfer function in the 0-time domain, which can be described by direction information indicating whether the sound is arriving, is treated as a sample value system, so h(t)
When expressed as (t=o, . . . , ?-1), there are T values, each of which can be interpreted as the loudness of the sound arriving at time (time step) t. Figure 1A shows the transfer function h<t> when T=8 (when T=0, h (0)
= 0).

The arrival direction of the sound arriving at time (time step)
f1-1. ..., by dividing L into regions labeled with L numbers, the transfer function h(L) is divided from 1 to L for each time (time step) of the sound loudness h(t).
Direction information I that has any value up to I! , (t) 0 For example, as shown in FIG. 2, the arrival direction of the sound in the sound field to be reproduced (at the listener's position) is determined by (j!
=) 1.2, 3.4, and the transfer function h(t) of the sound field is h+(t), h as shown in the figure for each of the four directions.
If it is divided into □(t), bs(t), and ha(t), the direction information i for the transfer function h(L) in Figure 1A is
<t> can be expressed as shown in FIG. 1B. In other words, h(1) is ff1(1) = 1, and h(2) is 1
(2)=2, h(3) is ff1(3)=3, h
(4) is 1(4)-1, and h(5) is ll(5)=2
and h(6) is 1(6)=4, and h(7) is ff1
(7) Set to -3. By representing the loudness h of the sound with X bits and the direction information l with X bits, the transfer function including the direction information is represented by the value h'(L) of x+y bits including the loudness and direction.

In reality, there are cases where arrivals come from two or more directions at the same time at a certain time (time step) t, but this is rare, and if there is, the direction information of the largest one of each h(t) is collected. give. Or before and after that h(t±
If 1) is zero, it is okay to shift one sample (time step) back and forth and give one set of h(t) and 2([) so that multiple directions do not overlap at the same time. By defining the number and region of L directions in this manner, a transfer function h'(t) containing directional information about a sound field to be reproduced is described.

Note that the period of time step t is 2 times the highest frequency in the signal.
It is made equal to the period of twice the frequency, and the number of T is made equal to, for example, the reverberation time.

FIG. 3 shows an embodiment of the present invention, and parts corresponding to those in FIG. 6 are given the same reference numerals. A transfer function h'(
L), for example, the functions h(t) and f indicating the first V4A, B
f1(t) is stored, and the transfer function h(t) is distributed to the speakers whose reproduction area is in that direction according to the direction information X<t> in the direction-specific transfer function creation circuit 7, and is Create a transfer function 0 For example, the number and area of arrival directions of sound in a sound field to be reproduced is divided into a total of 8 directions at 45 degrees in the horizontal direction as shown in Figure 4, and the transfer function h(t ) is given information that the direction of arrival of the sound at each time step (time) t is one of eight directions ff1-1 to ff8.

On the other hand, the conditions for the number of channels (number of speakers) of the reproduction system and the arrangement of the speakers in the actual reproduction sound field differ depending on the reproduction system, and a specific arrangement (β) − (1, ..., N) is given, then the transfer function for that speaker is calculated by distributing the transfer function h([) corresponding to the direction information to the speaker whose reproduction area is one or more of the direction information. Create. For this reason, for example, direction information j! is obtained from the transfer function h'(t) containing direction information stored in the transfer function storage circuit 6. (t), and the correspondence table ((α)
→(β)), and the direction-by-direction transfer function creation circuit 7 creates a direction-by-direction (by-speaker) transfer function for each channel of the reproduction system. This table is specific to playback systems.

To execute this method, h'(0),..., h'(T-1) stored in the transfer function storage circuit 6 are converted to h(0),...h(τ-1). 1 (0) ,...
, IT-1) and the correspondence table ((α) → (β))
Accordingly, the direction-specific transfer function creation circuit 7 generates the direction-specific (speaker-specific) transfer function h(n, k) (n=1.=4.
k=0. ..., T-1) and convolution operator 2
-n, the convolution operation with the output signal X(1) of the sound source l is performed, and the result is reproduced by the speaker 3-n.

Figure 5 shows an example of the configuration of a playback system with a listener, where the number of channels (number of speakers) in the playback sound field is 4, the speakers are placed front, back, left and right, and areas of 90 degrees each are defined. It shows. Assuming that the normalized direction information follows the example in Figure 4, 1(t) is 1.2, then n=1,
3.4 has n=2, 5.6 has n=3, 7.8 has n=4
Create a correspondence table (α1→(β)) corresponding to each of . The transfer function h (1, k) for speaker 3-1 is h (I!
−1°k) and h (1=2.k), and the amount of transmission 1ikh(2,k) to the speaker 3-2 is h
(j!=3.k) and h (Il=4.k). The same applies below.

In this way, by creating a transfer function for each speaker according to the number and arrangement of speakers in the reproduction system from a transfer function that includes specific sound field direction information, the arrangement and number of speakers in the reproduction system can be adjusted. It is possible to provide a system that allows a listener to experience sound in almost the same state as if he/she were in the room to be reproduced, even if the number of directions of arrival of sound in the sound field to be reproduced and the method of determining the area do not match.

"Effects of the Invention" As described above, according to the present invention, directional information is given to the transfer function of the sound field to be reproduced, and based on the directional information, the direction (speaker) is A method for creating a transfer function that gives directional information ( By standardizing the spatial division, number of divisions, etc.), even if the arrangement and number of speakers in the reproduction system do not match the standard, the standardized direction information held by the reproduction system and the arrangement and number of speakers in the reproduction system can be used. By associating the number of speakers with each other, it is possible to generate a playback signal for each speaker that corresponds to the playback system, making it easy to provide a system that allows listeners to experience sound in almost the same conditions as they would in the room that is to be reproduced. It has the advantage of

[Brief explanation of drawings]

Fig. 1A shows the transfer function h(t), Fig. 1B shows an example of the direction information ff1(t), Fig. 2 shows an example of the transfer function for each direction of arrival, and Fig. 3 FIG. 4 is a block diagram showing the configuration of an embodiment of the present invention. FIG. 4 is a diagram showing an example of the number of arrival directions and regions of sound in the sound field to be reproduced. FIG. 5 is a diagram showing the configuration of the reproduction system where there is a listener. FIG. 6, which shows one example, is a block diagram showing the configuration of a device for a conventional sound field reproduction method. 9

Claims (1)

[Claims] (1) In a sound field reproduction method that uses multiple speakers to reproduce a sound source and reproduce a sound field, directional information indicating the direction of arrival of the sound is given at each time step of the transfer function of the sound field to be reproduced. , distribute the value of each time step of the transfer function to the speakers whose reproduction area is the directional information, create a transfer function for each speaker, and respond using the transfer function for each speaker and the signal of the sound source. A sound field reproduction method characterized by generating a signal to be reproduced from a speaker.