JPH0157880B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in a reverberation device that produces reverberation that is more faithful to the original sound field.

In general, in order to obtain natural reverberant sound in an electric circuit, (1) the reverberation time of a specific frequency is not extremely long, (2) the reverberation process decays almost exponentially over time, and (3) the reverberation The following conditions must be met: the time density of reflected sound in the pattern increases in proportion to the square of the time interval from the direct sound. However, in order to achieve the condition (3) using an electric circuit, there are problems with the circuit construction method and the relationship with the peak and dip of the frequency characteristic, making it impossible to obtain natural reverberant sound.

FIG. 1 shows a basic block diagram of a conventional reverberation device. The signal from the signal source 1 passes through an adder 3 and is applied to a delay device 4 with a delay time τ, and the output of the delay device 4 is fed back to the adder 3 via a feedback amplifier 5. At signal source 1
is added to the signal from With this feedback configuration,
The output of the delay device 4 can be evaluated as an indirect sound signal 8, and the indirect sound signal 8 is added to the direct sound signal 2 from the signal source 1 in an adder 6, and the output terminal 7 receives the direct sound signal 8. A reverberant sound signal is formed by synthesizing the and indirect sounds. However, as can be seen from FIG. 2, which shows the impulse response of the reverberant sound signal at the output terminal 7, in this conventional reverberation device, the temporal density of the indirect sound signal 8 is coarse, so that the reverberation is still natural and pleasant. It was not possible to realize sound.

SUMMARY OF THE INVENTION An object of the present invention is to overcome the drawbacks of the conventional apparatus and to provide a reverberation apparatus that can provide a more natural reverberation sound by increasing the temporal density of indirect sound signals.

In order to achieve this objective, the present invention focuses on a transversal filter used as a means for simulating early reflections of a sound field.
It is proposed that the reverberation device be configured by being divided into a circuit that forms a spectral component of early reflected sound among the reverberant sound in the original sound field, and a circuit that forms a higher-order reflected sound spectral component for the early reflected sound. However, as can be seen from the following explanation, the early reflected sound forming circuit of the reverberation device according to the present invention is not limited to the transversal filter in any way, and it goes without saying that other convolution and integration circuits can be used. . That is, the present invention provides a late reflection sound formation circuit formed by connecting in parallel an early reflection sound formation circuit represented by the transversal filter, and a plurality of delay devices having different delay times each provided with a feedback circuit between input and output. Natural reverberation sound is achieved by vertically connecting the two channels and providing two channels, one for the right ear and one for the left ear.

The present invention will be described in detail below with reference to embodiments shown in the drawings.

FIG. 3 shows an example in which the reverberation device according to the present invention is configured with a digital circuit, and FIG. 4 shows an example in which it is configured with an analog circuit. In FIG. 3, 10 is an A/D converter, which converts the analog signal applied to the input terminal 9 into a digital signal. The output of this A/D converter 10 is transmitted to two early reflected sound forming circuits 11a and 11.
b. Early reflected sound forming circuit 11a, 1
1b is a circuit that forms early reflected sound by performing convolution integration, and is typically composed of a transversal filter. Each output of the early reflected sound forming circuits 11a and 11b is connected to a late reflected sound forming circuit 1 that produces an attenuation trail portion of the early reflected sound.
2a, 12b, and each output of the late reflected sound forming circuit 12a, 12b is applied to the D/A converter 13.
a, 13b into analog signals, which are supplied to subsequent circuits from output terminals 7a, 7b. The embodiment shown in FIG. 4 is the same as the circuit configuration shown in FIG. 3 except that the A/D converter 10 and D/A converters 13a and 13b are not provided and an analog circuit is used.

FIG. 5 shows early reflected sound forming circuits 11a and 11b.
A specific example of a transversal filter that constitutes the following is shown below. The transversal filter includes delay elements 14-1 and 14- having the same delay time.
2, ......, 14-m connected in series, and a certain constant coefficient C ₁ , C ₂ , at the output of each delay element.
......, coefficient unit 15-1, 15- that multiplies Cm
2,......,15-m and each coefficient unit 15-1,1
5-2, . . . , 15-m. 17 is an input terminal, the first
A signal is applied to the delay element 14-1 in the second stage. The output of the adder 16 is supplied from an output terminal 18 to a subsequent circuit.

Coefficient unit 15-1, 15-2, ......, 15-m
The coefficient values C ₁ , C ₂ , ......, Cm are derived from the desired impulse response of the transversal filter; below, the coefficient values C ₁ , C ₂ , ......, Cm
Explain how to determine. Figure 6A shows the original sound field,
FIG. 6B shows a reproduction sound field. In the original sound field 20, 21 is a sound source, which outputs an acoustic signal i _s .
22 is a listener, and assuming that the impulse response from the sound source 21 to the ear canal of the right ear of the listener 22 is h ₃ and the impulse response to the ear canal of the left ear is h ₄ , the sound pressure signal P _R of the right ear and the left ear The sound pressure signal P _L to the ear is given by the following equation.

P _R =h ₃ *i _s P _L =h ₄ *i _s (1) Here, * represents convolution.

In FIG. 6B, speakers 28a and 28b are arranged on both sides in the reproduction sound field 26, and the listener 24 is located at a position approximately equidistant from both speakers 28a and 28b. The audio electrical signal is transmitted from the input terminal 30 to a transversal filter 32a,
The signal is supplied to speakers 28a and 28b via 32b. Transversal filter 32a, 32
Let the impulse responses of b be a and b, respectively, the impulse response from the speaker 28a to the external auditory canal of the left ear of the listener 24 be _h1 , and the impulse response to the external auditory canal of the right ear be _h2 , then the impulse response from the speaker 28b to the external auditory canal of the listener 24 is The impulse responses to the external auditory canals of the left and right ears of the listener 24 can be considered to be equal to h ₂ and h ₁ , respectively, and the sound pressure signal received by the listener 24 is expressed by the following equation.

g _R = (h ₁ * b + h ₂ * a) * i _s g _L = (h ₂ * b + h ₁ * a) * i _s (2) Here, g _R is the sound pressure signal of the right ear, and g _L is This is the sound pressure signal from the left ear. Note that when the speakers 28a and 28b are directly driven by the electrical signal applied to the input terminal 30, the acoustic output in the original sound field 20 is
Let i be equal to _s . This can be achieved by using speaker systems with the same characteristics in the original sound field 20 and the reproduced sound field 26.

In equations (2) and (3), if the characteristics of a and b are set so that P _R ≒g _R P _L ≒g _L (3), the sound pressure to the listener 22 in the original sound field 20 is The signal can be reproduced to the listener 24 by loudspeakers 28a, 28b in the reproduction sound field 26. The conditions are h ₃ = h ₁ * b + h ₂ * a h ₄ = h ₂ * b + h ₁ * a (4) from equations (1), (2), and (3). More specifically, the evaluation function δ is the sum of the squares of the differences on both sides of equation (3), and a,
If b is determined, it becomes the desired impulse response of the transversal filters 32a and 32b. As a result, the transversal filter performs a convolution integral on the input electrical signal corresponding to the impulse response characteristic of the original sound field. δ is given by the following equation.

δ=Σ _o+n-1 〓 ⁱ⁼¹ {(g _Ri −P _Ri ) ² +(g _Li −P _Li ) ² } (5) Here, n is the number of time series samples of the impulse response, and m is Transversal filter 32
This is the number of taps a and 32b.

If the number m of taps of the transversal filters 32a and 32b is made as large as possible, reverberant sound can be obtained with the configuration shown in FIG. 6B, and any impulse response of the original sound field can be reproduced. However, in reality, it is difficult to increase the number of taps m beyond a certain level, and as a result, the time length of the early reflected sound by the transversal filter is also limited. This time length t is the delay element 14-1, 14-
2,......,14-m delay time τ and number of taps m
Therefore, t = m x τ. If the frequency band is 10 KHz, sampling is performed at 20 KHz, and the number of tap stages is 500, then t = 500 x 1/20 = 25 (msec), and it cannot be called reverberant sound yet. .

Therefore, in the present invention, the early reflected sound forming circuit 11
Late reflected sound forming circuits 12a and 12b are connected to the rear stage of the transversal filters a and 11b. A specific configuration of the late reflected sound forming circuits 12a and 12b is shown in FIG. 40 in Figure 7
are input terminals that receive the outputs of the early reflected sound forming circuits 11a and 11b. The input signal of the input terminal 40 is input to a plurality of parallel-connected adders 42-1, 42-2,
......, 42-n, delay devices 44-1, 44-2, ......, 44 each having a different delay time.
−n. Each delay device 44-1, 44-
2,......, 44-n are connected to an adder 48, and associated feedback circuits 46-1, 46
-2,......, 46-n to the linked adder 4
2-1, 42-2, ......, 42-n. The adders 42-1, 42-2, ......, 4
2-n, the input signal from the input terminal 40 and the feedback circuit 46-1, 46-2,......, 46-n
and the feedback signal from Although a variable resistor is illustrated as the feedback circuit, it is of course possible to select other appropriate elements. The output of adder 48 is connected to output terminal 50.

To explain the operation of the circuit shown in FIG. 7, each delay device 44-1, 44-2, . . . , 44-n is provided with an individual feedback path. The response shown in FIG. 2 to the input impulse is shown. However, in the circuit shown in FIG. 7, a plurality of delay devices with different delay times are connected in parallel, and the sum is calculated at the subsequent stage, so that the signal at the output terminal 50 has a unit delay time (i.e., , τ ₁ , τ ₂ , ......, τ _o ) are superimposed. Therefore, τ ₁ , τ ₂ ,
By appropriately selecting a delay time such _as . These delay times τ ₁ , τ ₂ ,
………, the relationship between τ _o is 1, 0.9 (±0.02), 0.8 (±
0.02), 0.7 (±0.02), 0.6 (±0.02), 0.5 (±0.02
)...
The sequence of numbers is optimal.

FIG. 8 schematically shows the impulse response of the reverberation device according to the present invention for one channel, with the horizontal axis representing time and the vertical axis representing signal level. Area A is the early reflected sound forming circuit 11a, 11
The area B exclusively shows the effect of the late reflected sound forming circuits 12a and 12b following the area A, and the curve that attenuates in this area B shows the effect of the late reflected sound forming circuits 12a and 12b. Circuits 12a, 12
b shows the envelope due to impulses formed more closely together. Even in the region A, there are coarse impulses that can be expressed by the attenuation curve as shown in the region B. Further, although the impulse responses of other channels may have different temporal relationships from the pattern shown in FIG. 8, they show qualitatively similar patterns.

In the above embodiment, the case where the early reflected sound forming circuit is constructed using a transversal filter has been explained, but this early reflected sound forming circuit may also be constructed using a delay device with a tap as shown in FIG. It may be configured as follows. In FIG. 9, an input signal applied to an input terminal 51 is transmitted to a plurality of delay elements 52-1, 52-2 of a tapped delay device 52.
...... 52-n, each delay element 52-n is delayed by a different delay time set by each delay element 52-n.
1, 52-2......It is output independently from 52-n. The output of each delay element is a variable resistor 53-1, 5
3-2...53-n sets the level to a predetermined level, and the adder 54 adds the total to the output terminal 55.
is output from. In this case, the tap delay device 5
_The delay times τ ₁₁ , τ ₁₂ . Each delay time of the delay elements 52-1, 52-2, . . . 52-n and variable resistors 53-1, 53-2, . . .
The level setting by 53-n can be determined by the same method as explained with reference to FIGS. 6A and 6B. Specifically, the tapped delay device is configured with a RAM (random access memory) in a digital circuit configuration, and the delay time is determined by the read address. The analog circuit configuration consists of a BBD (bucket bridge device) with a tap, and the delay time is determined by the clock frequency.

In addition, the late reflected sound forming circuit includes delay units 44-1, 44-2...44- as shown in FIG.
Feedback circuits 46-1, 46-2... between the input and output of n.
...46-n are provided with delay devices 44-1', 44-2'...44-n, respectively.
Feedback circuits 46-1', 46-2'... between input and output of n'
...Feedback circuits provided with 46-n' are connected cascaded, and the outputs of each of the cascaded circuits are added together in an adder 48, or a feedback circuit is connected between the input and output of the delay device. It may also be configured by cascade-connecting circuits that add up the outputs of each of a plurality of feedback circuits provided with a plurality of feedback circuits. In this case, the relationship between the delay times of the delay devices of each channel constituting the cascaded feedback circuits may be set so that large peaks and dips do not occur on the frequency characteristics. When the feedback circuit including the delay device is configured in multiple stages in this manner, the time density of reflected sound can be further increased, and a more natural reverberation feeling can be given to the listener.

As can be understood from the above explanation, the reverberation device according to the present invention employs a configuration that faithfully reproduces the early reflected sound of the original sound field using the early reflected sound forming circuit, so that the sense of distance and spatiality of the original sound field is information can be accurately provided to the listener, and by adding a late reflected sound forming circuit, it is possible to provide the listener with a natural sense of reverberation that cannot be obtained by only reproducing early reflected sounds.

[Brief explanation of drawings]

Figure 1 is a block diagram of a conventional reverberation device.
FIG. 2 shows the impulse response of the device shown in FIG. FIG. 3 shows a block diagram of an embodiment in which the reverberation device according to the present invention is configured with a digital circuit.
FIG. 4 shows a block diagram of an embodiment in which the reverberation device according to the present invention is constructed from analog circuits. FIG. 5 shows the configuration of a transversal filter used in the reverberation device according to the present invention. Figures 6A and 6
Figure B is a diagram explaining the theory of determining the coefficients of the transversal filter, Figure 6A shows the original sound field, and Figure 6B shows the reproduced sound field. FIG. 7 shows a specific example of a late reflected sound forming circuit. FIG. 8 schematically shows the impulse response of the reverberation device according to the present invention for one channel. FIG. 9 shows another specific example of the early reflected sound forming circuit, and FIG. 10 shows another specific example of the late reflected sound forming circuit. 1... Signal source, 3, 6... Adder, 4... Delay device, 5... Feedback amplifier, 7... Output terminal, 9...
Input terminal, 10...A/D converter, 11a, 11
b...Early reflected sound forming circuit, 12a, 12b...
Late reflected sound formation circuit, 13a, 13b...D/A
Converter, 14-1, 14-2, ......, 14-m
...Delay element, 15-1, 15-2, ......, 1
5-m... Coefficient unit, 16... Adder, 17... Input terminal, 18... Output terminal, 20... Original sound field, 2
1... Sound source, 22, 24... Listener, 26... Playback sound field, 28a, 28b... Speaker, 30...
Input terminal, 32a, 32b... Transversal filter, 40... Input terminal, 42-1, 4
2-2,......,42-n...adder, 44-
1,44-2,......,44-n...delay device, 4
6-1, 46-2,......, 46-n... Feedback circuit, 48... Adder, 50... Output terminal.

Claims (1)

[Scope of Claims] 1. A plurality of delay devices each having a feedback circuit and having different delay times are connected in parallel, and are provided with an early reflection sound forming circuit that forms an early reflection sound of the original sound field by performing convolution integration. It is equipped with two channels consisting of a late reflected sound forming circuit that outputs the sum of the outputs, and the convolution integral in the early reflected sound forming circuit is the listener's right ear impulse response in the original sound field and the listener's right ear impulse response in the reproduced sound field. The sum of the squared difference between the right ear impulse response and the squared difference between the listener's left ear impulse response in the original sound field and the listener's left ear impulse response in the reproduced sound field is minimum A reverberation device characterized in that the reverberation is performed based on a coefficient value determined so that 2. The early reflected sound forming circuit is a transversal filter in which delay elements having the same delay time are connected in series, and the output of each delay element is multiplied by the determined coefficient value and then added. A reverberation device according to claim 1. 3. The early reflected sound forming circuit supplies the input signal to a plurality of delay devices having different delay times under the predetermined coefficient value, and adds up the outputs of the plurality of delay devices to produce an output. A reverberation device according to claim 1. 4. The reverberation device according to claim 1, 2, or 3, wherein the early reflected sound forming circuit and the late reflected sound forming circuit are configured with digital circuits. 5. The reverberation device according to claim 1, 2, or 3, wherein the early reflected sound forming circuit and the late reflected sound forming circuit are configured with analog circuits. 6. Claim 1, wherein the early reflected sound forming circuit is configured with a digital circuit, and the late reflected sound forming circuit is configured with an analog circuit.
The reverberation device according to item 1, 2 or 3.