JPH0228810B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION This invention focuses on the overall reverberation (attenuation) waveform for an indoor space, particularly focusing on the time change of the total reverberant energy existing in the room, and uses the "impulse response square integral method" and digital processing technology. The present invention relates to a method and apparatus for measuring using.

In general, it is often necessary to carry out targeted characteristic measurements in order to learn various characteristics of a particular acoustic room from an acoustic point of view. For example, when a lecture is given in one hall, the acoustic effects of the speech or music become important.
In this case, if matters such as the short sound attenuation characteristics of the entire room for signals from one signal source or the reverberation time of the entire room are clearly understood, it is possible to determine the structure of the venue where the event will be held. It is possible to aim for high acoustic effects due to its acoustic characteristics.

Among various acoustic characteristics, for example, transient characteristics, especially the reverberation (attenuation) waveform necessary to calculate the reverberation time for one sound receiving point in the room, can be measured by cutting off steady-state band noise. There is a method of measuring the attenuation characteristics of a sound receiving point at a certain time. However, in such a method, the accuracy of measuring reverberation time is limited by random changes in the decay curve (transient curve). To minimize the effect of variations in the attenuation curve on the measured reverberation time values, it is necessary to repeat the reverberation measurement many times and average the approximately obtained reverberation times, e.g. by hitting a straight line. be. However, this method is not only inefficient, but also does not provide the true characteristics of the damping, ie, the essential transient response curve of the system to stationary noise.

By the way, MR Schroeder's so-called impulse response square integral method is known as a method for measuring the essential transient characteristics of one sound receiving point. The principle is that when the sound source band noise is cut off from the steady state, the transient characteristic of the sound receiving point, that is, the essential transient response curve of the sound receiving point corresponding to the average of ∞ times of the reverberation (attenuation) waveform <S ² ( t)> is calculated from the impulse response r(x) between the sound source and the sound receiving point.According to this, the sound pressure level S(t) at a certain point in time t in the transient characteristic is expressed as follows. It will be done.

<S ² (t)>=N∫ ^∞ _t r ² (x) dx ………(1) where N: power of sound source band noise, r(x):
impulse response.

Therefore, by squaring and integrating the impulse response r(x) in the integration interval [t, +∞], the collective average of the infinite number of squares of the sound pressure level S(t) at time t can be found. . Methods of actually obtaining transient characteristics, such as reverberation waveforms, of a certain chamber based on the above principle can be roughly divided into the following two methods. (1) is the so-called double impulse method, and (2) is the so-called backward integration method.

The principle of the double impulse method (1) is carried out by expanding the principle equation (1) above as follows.

<S ² (t)>=N∫ ^∞ _t r ² (x)dx=N∫ ^∞ ₀ r ² (x
) dx−N∫ ^t ₀ r ² (x) dx……(2) Here, we first obtain the square integral of the impulse response r(x) for the integral interval [0, +∞], and from this we sometimes By sequentially subtracting the square integral of the impulse response r(x) in the integral interval [0, t] that changes from moment to moment, the transient characteristic < S ²
(t)> is obtained.

FIG. 1 shows the configuration of a conventional device for measuring reverberant waveforms according to this principle.

In this figure, the first impulse from an impulse source 2' (for example, a paper pistol) is picked up by a microphone 3' in a measurement chamber 1', and the output is sent to a squaring circuit 4' and then squared. ,
The signal is sent to the integrating circuit 6' via the switch SW, where the signal is integrated and displayed on the display device 7. When generating the second impulse, if you change the switch SW and connect the inverting circuit 5' to invert it and then display its output via the integrating circuit 6', the latter can be subtracted from the former. As the process progresses, transient characteristics such as reverberant waveforms are obtained.

However, in this double impulse method, it is necessary to generate two identical pulses with high accuracy as a sound source, or to use a tape recorder etc. to generate two identical pulses.
It is technically almost impossible to reproduce the impulse response of the first time in exactly the same way as the first time. Therefore, since the generated impulse waves are different, there is a drawback that the measurement error becomes large especially at the end of the integration, that is, at the point where the level is low at the end of the reverberant waveform.

The principle of the backward integration method in (2) is to convert the principle equation (1) into <S ² (t)>=N∫ ^∞ _t r ² (t)dx = N∫ ^-t _-∞ r ² ( -x) dx ......The method used is to integrate in the opposite direction as shown in (3). In other words, the integral interval [t, ∞] is changed to [-
∞, -t] to obtain the transient characteristic <S ² (t)>.

FIG. 2 shows the configuration of a conventional device for implementing this method.

In this figure, first, the impulse response r(x) from the picked-up microphone 3'' is once recorded on the tape recorder 4'', and then the tape recorder 4'' is reversed to reproduce the impulse response in reverse. The output is supplied to a display device 7'' via a squaring circuit 5'' and an integrating circuit 6''. However, this method has the disadvantage that real-time processing is not possible because the collected impulse response is recorded once on a tape recorder, etc., and the impulse response r(x)
There is a drawback that a tape recorder or the like is required to record the data. Therefore, it lacks immediacy.

According to MR Schroeder's impulse square integral method, one measurement yields the result of measuring an infinite number of times at exactly the same location and taking the average, that is,
Although a collective average can be obtained, this does not immediately represent the true reverberation characteristics of the indoor space. Even when using the square integral method, the characteristics change depending on which point in the room is measured. In other words, the acoustic energy present in the room must be observed as a whole. Therefore, it can be said that the average that takes into account the spatial extent, that is, the collective average that also takes into account the spatial average, should be regarded as the true reverberation characteristic.

This invention was made in view of the above circumstances, and its purpose is to generate a large number of impulse responses ri(x) (i=1,
2...n) is converted into a digital signal and then processed,
It is an object of the present invention to provide a method and apparatus for determining a comprehensive reverberation waveform (a spatial collective average of reverberation waveforms) of a room. That is, the method according to the present invention is a method for measuring a reverberation waveform by analyzing impulse responses ri(x) (i=1, 2...n). 2...
n) sequentially, and the collected impulse response ri(x) is calculated based on the basic clock rate fS.
(i=1, 2...n) into a digital signal,
The impulse response ri (x) (i = 1, 2...n) converted into this digital signal is squared, and this squared value ri ² (x) (i = 1, 2... n) is repeatedly accumulated, and Based on the clock rate fL (fL=fS×2 ^-k ) (where k is a natural number) lower than the basic clock rate fs, the iterative accumulation result is read and the predetermined integration period [0, tj (j=1 , 2...m)]
Find each integral value ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x)dx,
Sequentially _add these integral values∫ ^tj(j=1,2 … ^m) ₀ ri ² (x)dxo 〓 ⁱ⁼¹
∫ ^tj(j=1,2, … ^m) ₀ ri ² (x)dx, and based on this calculated result, _o 〓 ⁱ⁼¹ ∫ ^tm ₀ ri ²
(x)dx− _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² Calculate (x)dx and display the reverberant waveform over time based on this calculation result/
It is characterized by being able to record.

Further, the apparatus according to the present invention is an apparatus for measuring a reverberant waveform by analyzing impulse responses ri(x) (i=1, 2...n). 2...n),
An analog/digital conversion circuit that converts the digital signal to a digital signal based on the basic clock rate fS, a squaring circuit that squares the digital signal from this conversion circuit based on the basic clock rate fS, and an output of this squaring circuit based on the basic clock rate fS. An accumulation circuit that repeatedly accumulates data, and a clock rate fL (fL=fS×2 ^-k ) lower than the basic clock rate fS (however,
(k is a natural number), the values of the accumulation circuit are read out, and these values are integrated into a predetermined integration interval [0,
=1,2...m)] ∫ ^tj(j=1,2 ... ^m) ₀ ri ²
(x) A holding circuit that holds dx and each integral value from this holding circuit ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) dx is stored at a predetermined storage location (i ,
j) and overwrite _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x)dx
A storage device ^that stores ^, and ^the contents stored ⁱⁿ _this storage ^device _{are read} out ^{. )} ₀ ri ² (x) dx, and a device that displays/records a reverberant waveform over time based on the result of this subtraction.

Embodiments of the present invention will be described below with reference to the drawings. FIG. 3 shows the configuration of an embodiment of the present invention. In this figure, the output of a short sound generator 1 that generates impulses (filtered impulses) for each frequency band is input to a sound source 3, for example, a speaker, installed in a room 2 in which the reverberation waveform is measured. It's getting old. There are many microphones 4i (i=1, 2...n) in the same room 2.
Microphone 4i (i
=1, 2...n) are each contact 5 of the changeover switch 5
i (i=1, 2...n), and the common terminal 5-COM of this changeover switch 5 is connected to the input terminal of an amplifier circuit 6, and the output of this amplifier circuit 6 is connected to an analog/digital conversion circuit ( The signal is inputted to an accumulator 9 via an A/D conversion circuit (hereinafter referred to as an A/D conversion circuit) 7 and a squaring circuit (multiplier) 8.
This accumulator 9 has a built-in addition register, a full adder, etc., and its output is inputted to an addition circuit 11 via a register 10, and at the other input terminal of the accumulator 9. It is now entered into . The addition circuit 11
is for adding the output of register 10 and the output of register 12, and the output is added to register 1.
The output of the RAM 14 is input to the register 12 and to the subtraction input terminal of the subtraction circuit 15. , the output of the register 13 is inputted to the subtracted input terminal of the subtraction circuit 15 via the latch circuit 16. The output of the subtraction circuit 15 is input to a display/recording device 19 via a read-only memory (hereinafter referred to as ROM) 17 and an interface circuit 18, where it is displayed/recorded.

In this figure, the broken line indicates the timing control circuit 2.
A control signal of 0 is shown. The signal C-1 is generated according to the basic clock rate fS and is input to the A/D conversion circuit 7 and the accumulator 9, respectively.
Furthermore, the signals C-2 and C-3 have a clock rate fL lower than the basic clock rate fS (fL=fS×
2 ^-K ) (where k is a natural number),
Signal C-2 is input to register 12, and signal C-2 is input to register 12.
3 is input to register 10, register 13, RAM 14 and latch circuit 16, respectively. Further, the signal C-4 is input to the RAM 14 as an address.

Next, the operation of the above embodiment will be explained with reference to the time chart shown in FIG. First, before the first impulse is generated from the sound source 3, connect the common terminal 5-COM of the changeover switch 5 and the contact 5-1,
After clearing the storage unit 9, RAM 14 and register 12, the first
The impulse is collected by the microphone 4-1. This collected impulse response r ₁ (x)
is amplified by the amplifier circuit 6 and supplied to the A/D conversion circuit 7. A signal C-1 is input to the A/D conversion circuit 7 from the timing control circuit 20. This signal C-1 is used for sampling A/D in the A/D conversion circuit as shown in FIG. 4A.
This is a basic clock pulse used during D conversion operation, and it samples the analog signal input to the A/D conversion circuit 7, and converts this sampled value into the A/D conversion circuit 7.
The conversion circuit 7 converts it into a digital signal. This A/D converted digital signal is squared (multiplied) by a squaring circuit 8, and this squared value is input to an accumulator 9, which converts the input digital signal into the Accumulation is performed sequentially at the timing of signal C-1. This accumulated value is sequentially input to the register 10 at the timing of the signal C-3 from the timing control circuit 20 (shown in FIG. 4B).
The contents of are sequentially added to the contents of the register 12 by the adder circuit 11 at the timing of the signal C-3, and the result of this addition is stored in a predetermined location of the RAM 14 and input to the register 13.
It is assumed that the RAM 14 has an address structure as shown in FIG.
This is done by the signal C-4 (shown in FIG. 4C).

Now, from the measurement start time (t=0), the contents of the accumulator 9 (S _1-1 ) are determined by the first pulse of signal C-3 (the pulse at time t _1-1 shown in FIG. 4 (b)). ) is input to register 10, then
At the same time, the contents of address 1 of the RAM 14 are input to the register 12 by the signal C-2 (shown in FIG. 4D) from the timing control circuit 20 (since the RAM 14 is initially cleared, the contents of address 1 are "0"). ). The input contents of the register 12 (“0”) and the contents of the register 10 (S _1-1 ) are added by the adder circuit 11 [S _1-1 +0], and the RAM 1
It is written again to address 1 of 4 and is also input to register 13. Next, the time shown in Figure 4
At t _1-2 , the contents of accumulator 9 (S _1-2
) is input to register 10, at the same time
The contents of address 2 of the RAM 14 (which is also "0") are input to the register 12, the adder circuit 11 adds [S _1-2 +0], and this addition result (S _1-2 ) is sent to the RAM 14 again. The data is written to address 2 and input to the register 13. Thereafter, writing to the RAM 14 and inputting to the register 13 are sequentially performed in the same manner until the accumulation end time t1 _-n . That is, when the writing at time t _1-n is completed, each address of the RAM 14 has address 1→S _1-1 ,
Address 2 → S _1-2 , address 3 → S _1-3 , ...address m → S _1-n are written, and the register 13 has S _1-n.
is entered.

In this way, when the data collection of the impulse response r ₁ (x) at the microphone 4-1 is completed, the common terminal 5-
After the contact 5-2 is connected to COM and the accumulator 9 is cleared, the second impulse from the sound source 3 is collected by the microphone 4-2. The impulse response r ₂ (x) is sequentially accumulated in the accumulator 9 via the amplifier circuit 6, A/D conversion circuit 7, and squaring circuit 8, as described above.

Now, at time t _2-1 (see Figure ₄ ), the first accumulated value (S _2-1
) is input to register 10, at the same time
Contents of address 1 of RAM14 (through the process described above)
S _1-1 is stored) is input to the register 12, and the addition circuit 11 calculates [S _2-1 +S _1-1 ], and the result of this operation is [S _2-1 +S _1-1] . ] is RAM
The signal is inputted again to address 1 of 14, and is also inputted to register 13. In the same manner, the cumulative value of the impulse response r ₂ (x) is sequentially added to the contents of the RAM 14, and at the end of the accumulation time t _{2 -n} , each address of the RAM 14 has an address 1→[S _{1 -1} +
S _2-1 ], 2nd address → [S _1-2 +S _2-2 ], 3rd address → [S _1-3
+S _2-3 ]...Each data from address m → [S _1-n +S _2-n ] is written, and [S _1-n +S _2-n ] is written to register 13.
The value is entered. Next, Microphone 4-
3, 4-4, . . . 4-n, impulse responses r ₃ (x), _{r 4} (x) . That is, n impulse responses r ₁ (x), r ₂ (x)
…r _o When all the processing of (x) is completed,
The contents of RAM14 are address 1 → [S _1-1 +S _2-1 +...
+S _o-1 =TS ₁ ], 2nd address → [S _1-2 +S _2-2 +…+S _o-2
=TS ₂ ],...m address → [S _1-n +S _2-n +...+S _on =
TS _n ], and this TS _n is stored in register 13.
The value is entered.

Thus, the contents TS _n of the register 13 are input to the subtracted input terminal of the subtraction circuit 15 via the latch circuit 16, while the contents of the RAM 14 are sequentially read out from address 1 and input to the subtraction input terminal of the subtraction circuit 15. The subtraction circuit 15 sequentially performs the following calculations. That is, [TS _n −TS ₁ =Z ₁ ], [TS _n −TS ₂ =Z ₂ ]…[TS _n
−TS _n =Z _n ] is given. These calculation results Z ₁ to _{Z n} are stored in the ROM17
The data is logarithmically compressed using the logarithmic operation (10logX) function of the computer, and displayed/recorded over time by the display/recording device 19 via the interface circuit 18. Note that output waveforms from each part of the above device are shown in FIG. In this figure, g is the impulse response r _i (x) (i=
1, 2...n), H indicates the output of the squaring circuit 8, R indicates the output of the register 10 in analog display for convenience, and N indicates n impulse responses r _i (x). For convenience, the contents stored in the RAM 14 at the time when all measurements of (i=1, 2...n) are completed are shown in an analog display, and the signal C-3 of the timing generator is shown again. In the figure, O shows the output of the subtraction circuit 15 in an analog display for convenience, and W shows an example of display/recording in the display/recording device 19.

Therefore, each of the above accumulated values (S _1-1 ~ _{S on} , TS ₁
˜TS _n , Z ₁ ˜Z _n ) are expressed as integral expressions as follows. That is, S _ij =∫ ^tj ₀ ri ² (x)dx TS _j = _o 〓 ⁱ⁼¹ ∫ ^tj ₀ ri ² (x)dx Z _j =TS _n −TS _j = _o 〓 ⁱ⁼¹ ∫ ^tm ri ² ( x) dx− _o 〓 ⁱ⁼¹ ∫ _tj ri ² (x) dx. _{Therefore ,} ^the _above ^device _produces ^{reverberation by the relational expression :} This achieves spatial averaging of the waveform.

Now, in addition to the above-mentioned functions, the above-mentioned apparatus has a function of displaying the intermediate progress of each measurement, so that the measurement can be performed while referring to the intermediate progress. Next, this function will be explained. Assume that a new accumulated value is written to address l of the RAM 14 by the signal C-3 of the timing control circuit 20, and this accumulated value is input to the register 13. Next, the RAM address signal C-4 of the timing control circuit 20 changes the address from address 1 to address m until the next pulse of the signal C-3 arrives, and the contents of the RAM 14 are sequentially subtracted by the subtraction circuit 15. Supplied to the input end. On the other hand, the accumulated value input to the register 13 is input to the subtracted input terminal of the subtraction circuit 15 via the latch circuit 16, and in the subtraction circuit 15, [(contents of register 13) - (contents of RAM address 1)] (Contents of register 13) - (Contents of RAM2 address)] 〓 [(Contents of register 13) - (Contents of RAM1 address)] are sequentially subtracted, and the result of this subtraction, that is, the reverberation waveform at that point, is displayed on the display device 19. are displayed sequentially. That is, the above calculation process is performed every cycle of the C-1 pulse, and the progress of the final reverberant waveform of the measurement cycle is displayed. Changes in the address in the above process are shown in FIG. 4C, and FIG. 7 shows an example of the display in this case.

In this way, the above device achieves spatial collective averaging of reverberant waveforms using a large number of microphones, but as shown in Figure 3, by sequentially moving one microphone, each point It is also possible to obtain a similar effect by collecting the impulse response r _i (x) of . In addition, when averaging the reverberation waveform at a certain point in a reverberation chamber with rotary blades (the curve is different each time depending on the state of the rotor blade), or by changing the sound source position instead of the sound receiving position, the reverberation waveform can be It can also be used for extended use, such as when obtaining spatial collective averaging of , or averaging on the frequency axis (for example, averaging of 100Hz and 200Hz waveforms), or when obtaining composite characteristics of each of the above cases. Can be done.

As explained above, the method according to the present invention
Impulse response ri(x) (i=1, 2...n)
In the method of analyzing the reverberation waveform and measuring the reverberation waveform,
The plurality of impulse responses ri(x)(i=
1, 2...n) in sequence, and the basic clock rate.
The collected impulse response ri(x) (i=1, 2...n) is converted into a digital signal based on fS, and the impulse response ri(x) (i=1, 2...n) converted to the digital signal is converted into a digital signal. ...n), and repeatedly accumulates this squared value ri ² (x) (i=1, 2...n), and the basic clock rate fs
Based on the lower clock rate fL (fL=fS×2 ^-k ) (where k is a natural number), the iterative accumulation result is read and the predetermined integration period [0, tj (j=1,
2...m)] ∫ ^tj(j=1,2 ... ^m) ₀ ri ² (x)
dx
Find each integral value ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) dx and add them sequentially _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) dx, and according to this calculated result, _o 〓 ⁱ⁼¹ ∫ ^tm ₀ ri ²
(x)dx− _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² Calculate (x)dx and display the reverberant waveform over time based on this calculation result/
It is to be recorded. Further, the device according to the present invention has an impulse response ri(x) (i=1, 2...
n) to measure the reverberant waveform, the large number of impulse responses ri(x)(i
=1, 2...n) into a digital signal based on the basic clock rate fS, a squaring circuit that squares the digital signal from this conversion circuit based on the basic clock rate fS, and this squaring circuit. an accumulation circuit that iteratively accumulates the output of the clock based on the basic clock rate fS, and a clock rate fL (fL
= fS×2 ^-k ) (where k is a natural number), read out the values of the accumulation circuit, and calculate these values for each value in a predetermined integration interval [0, tj (j=1, 2...m)]. Integral value ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) Holding circuit to hold as dx and each integral value from this holding circuit ∫ ^tj(j=1,2 … ^m) ₀ ri ² ( x) Add and overwrite dx to a predetermined storage location (i, j) based on the clock rate fL and store _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) dx _Read out ^the storage device ^{to be} used and _the ^memory contents stored ⁱⁿ _{this storage} device ^. ri ² (x)
It is equipped with a subtraction circuit that performs subtraction of dx, and a device that displays/records a reverberation waveform over time based on the result of this subtraction.

Therefore, according to the present invention, compared to the case where reverberant waveforms are obtained by analog processing, only one system is required, and the configuration can be extremely simple and inexpensive, and the problem of variation in accuracy at each measurement point is completely eliminated. There are no advantages. On the other hand, this invention provides a large number of impulse responses r _i (x) (i=1, 2...n)
Since the reverberant waveform is obtained by squaring the squared value and integrating this squared value over a number of preset integration intervals, the spatial collective average of the reverberant waveform can be calculated by taking advantage of the advantages of Mr. MR Schroeder's impulse response square integration method. can be converted. In other words, the impulse response square integral method can be extended from finding the reverberation waveform specific to a certain point in a room to finding the reverberation waveform specific to the room, and it can be used to compare reverberation waveforms between rooms or to determine the sound absorption coefficient in a reverberant room. This has the effect of allowing accurate measurements.

Therefore, by making considerations such as aligning the integration start time for each impulse and performing the above averaging process limited to an appropriate spatial area in the room, D can be calculated from the spatial collective average of the obtained reverberant waveform.
It is also possible to accurately determine the spatial average value of so-called early reflection sound indicators such as ts, EDT (initial decay time), and _ts (time center of gravity).

Furthermore, according to the present invention, the squaring operation and the integral operation do not use stored data respectively, but directly perform squaring and integration processing on the digital conversion data of the input signal by a single processing system, Since the data after this integral operation is taken into memory discretely, the advantage is that the required memory capacity can be extremely reduced without having any real effect on the processing accuracy of input data. Achieved. That is, compared to the well-known direct storage method in which an analog signal from a microphone or the like is digitized, sampled and stored without any arithmetic processing, the required memory capacity is drastically reduced. Spatial averaging can be performed with high precision using only an extremely small amount of memory.

[Brief explanation of drawings]

FIG. 1 shows a diagram explaining the double impulse method according to the prior art and its waveform, FIG. 2 shows a diagram explaining the backward integration method according to the prior art and its waveform, and FIG. 3 shows the diagram according to the present invention. 4 is a diagram showing a timing chart for explaining the operation of the embodiment shown in FIG. 3, FIG. 5 is a diagram showing the address structure of RAM, and FIG.
Figure 7 shows the waveforms of each part of the example shown in the figure.
FIG. 6 is a diagram showing an example of display for displaying the measurement progress in the embodiment device shown in the figure. 7...Analog/digital conversion circuit, 8...
Square circuit, 9...accumulation circuit (accumulator),
11...Addition circuit, 14...Storage device (RAM),
15... Subtraction circuit, 19... Display/recording device.

Claims (1)

[Claims] 1 Impulse response ri(x) (i=1, 2...
n) to measure the reverberant waveform by analyzing the plurality of impulse responses ri(x)(i
= 1, 2...n) is sequentially collected, the collected impulse response ri(x) (i = 1, 2...n) is converted into a digital signal based on the basic clock rate fS, and this digital signal is The impulse response ri(x) (i= ¹ , 2...n) converted to crate
Based on the clock rate fL (fL=fS×2 ^-k ) (where k is a natural number) lower than fs, the iterative accumulation result is read and a predetermined integration period [0, tj (j=1,
2...m)] ∫ ^tj(j=1,2 ... ^m) ₀ ri ² (x)
dx
Find this integral value∫ ^tj(j=1,2 … ^m) ₀ ri ² (x)dx and sequentially add _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x)dx And, based on this calculated result, _o 〓 ⁱ⁼¹ ∫ ^tm ₀ ri ²
(x)dx− _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² Calculate (x)dx and display the reverberant waveform over time based on this calculation result/
A method for spatial collective averaging of reverberant waveforms, characterized in that the reverberant waveforms are recorded. 2 Impulse response ri(x) (i=1, 2...
n) to measure the reverberant waveform, the device analyzes the large number of impulse responses ri(x)(i
=1, 2...n) into a digital signal based on the basic clock rate fS, a squaring circuit that squares the digital signal from this conversion circuit based on the basic clock rate fS, and this squaring circuit. an accumulation circuit that iteratively accumulates the output of the clock based on the basic clock rate fS, and a clock rate fL (fL
= fS×2 ^-k ) (where k is a natural number), read out the values of the accumulation circuit, and calculate these values for each value in a predetermined integration interval [0, tj (j=1, 2...m)]. Integral value ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) Holding circuit to hold as dx and each integral value from this holding circuit ∫ ^tj(j=1,2 … ^m) ₀ ri ² ( x) Add and overwrite dx to a predetermined storage location (i, j) based on the clock rate fL and store _o 〓 ⁱ⁼¹ ∫ ^tj(j=1,2 … ^m) ₀ ri ² (x) dx _Read out ^the storage device ^{to be} used and _the ^memory contents stored ⁱⁿ _{this storage} device ^. ri ² (x)
1. A spatial collective averaging device for reverberant waveforms, comprising a subtraction circuit that performs subtraction of dx, and a device that displays/records reverberation waveforms over time based on the subtraction results.