JPH0454439B2 - - Google Patents

Description

[Detailed description of the invention]

This invention focuses on converting the acoustic characteristics of a sound field for listening to music into the form of physical and psychological quantities that are easy to evaluate, and relates to an audio device that can vary physical quantities to enable more preferable music listening. It is something. Research has been conducted on desirable sounds and sound fields, but the research conducted this time revealed that the following parameters are necessary to determine a good sound field in a concert hall, which is the standard for the sound reproduced by acoustic equipment. It gradually became clear that there was something to show for it. In other words, there are four important objective parameters that represent the characteristics of binaural sound: audible sound pressure, delay time of the first reflected sound, reverberation time of subsequent reverberant sound, and interaural cross-correlation coefficient. That is, the preference (auditory comfort) in a series of simulated sound fields.
This was clarified by a test. Next, the above four elements will be explained in detail. First, Figure 1 shows the relationship between a sound source and a person's head in a space with reflective walls. In the figure, 11 is a human head, 12 is a sound source, and 13 and 14 are reflecting walls. Here, the sound source signal is p(t), and the impulse responses from the sound source to the left ear and right ear are hl(t) and hr(t), respectively.
Then, the left ear and right ear signals fl(t) and fr
(t) is fl(t)=∫o ^t p(ν)hl(t-ν)dν=
p
(t)*hl(t) ...(Formula 1a) fr(t)=∫o ^t p(ν)hr(t-ν)dν=
It is expressed as p(t)*hr(t) (Formula 1b). The * mark in the above formula indicates composition. In the figure, n=0 of hnm indicates that the sound enters the ear directly, n=1 indicates that the sound reflected by the reflective wall 13 enters the ear, and n=2 indicates that the sound reflected by the reflective wall 14 enters the ear. m=1 indicates the case where the sound enters the left ear, and m=r indicates the case where the sound enters the right ear. By the way, in FIG. 1, the reflecting walls 13, 14
The number of sounds that enter the ear after being reflected by the walls is shown only two for one reflecting wall, but assuming that many such reflections occur, the impulse response when reflected at the reflecting walls 13 and 14 is written as W _o (t ), the impulse responses reaching the left and right ears hl(t), hr
(t) are respectively hl(t)=〓 〓 ⁿ⁼⁰ G _o W _o (t−Δt _o ) * hnl(t) ...(Formula 2a) hr(t)=〓 〓 ⁿ⁼⁰ G _o W _o It can be expressed as (t-Δt _o )*hnr(t) (Formula 2b). This (Equation 2a) and (Equation 2b)
Using (Equation 1a) and (Equation 1b) above, fl (t) = 〓 〓 ⁿ⁼⁰ p (t) * G _o W _n (t - Δt _o ) * hnl ... (Equation 3a) fr (t)=〓 〓 ⁿ⁼⁰ p(t)*G _o W _n (t-Δt _o )*hnr ... (Formula 3b). Here, if the sound source p(t) does not have uniform radiation characteristics, p(t) can be replaced by pn(t), taking into consideration the radiation pattern for each direction. By the way, the information in the acoustic signal that enters both ears includes independent and objective acoustic parameters, and the first parameter is the sound source signal p.
(t) can be given. Using the sound source signal, calculate the long-term autocorrelation function Φp(τ) as follows: Φp(τ) = lim T→∞1/2T∫ ^T _-T p′(t) p′(t+τ)dt... 4) It can be expressed as Here p'(t)=p(t)
*s(t), where s(t) corresponds to the sensitivity of the ear,
Theoretically, it is expressed by the characteristics of the middle ear and the outer ear, but in practice it can be expressed as the impulse response of a G filter, which is well known as an approximation of auditory characteristics. Naturally, the power Φp of p′(t)
By dividing (Equation 4) by (o), the normalized autocorrelation function φp (τ) can be expressed as φp (τ) = Φp (τ) / Φp (o) ... (Equation 5) . FIGS. 2a and 2b illustrate the measured values of the normalized autocorrelation corresponding to the above equation (5).
FIG. 2a shows the measured values corresponding to the music "Royal Pavane" by Gibbons, and this music will be referred to as music A hereinafter. Figure 2b is the music “Sinho Nietsta” by Arnold, Op.
48: Movement, Allegro con brio (Synfornietta, Opus 48: Movement,
This music is hereinafter referred to as Music B. Next, the second objective parameter is the impulse generated by reflection at boundaries such as walls. This is related to the initial time delay between the direct sound and the first reflected sound, as well as the initial reflected sound, subsequent reverberation, and changes in the spectrum due to reflection. The objective parameters of 3 include the impulse responses hnl(t) and hnr(t) to the left and right ears.These impulse responses play an important role in sound localization, and each has an independent relationship with the other. This is clear from the fact that in the case of a centrally localized signal, hnl(t)hnr(t). Next, the two impulse responses hnl(t) and
Derive the mutual dependence between hnr(t). First, the long-term interaural cross-correlation function Φlr(τ) between the binaural signals fl(t) and fr(t) is expressed as fl(t), fl
(t), Φlr(τ)=lim T→∞f(t)fr(t+τ) dt, |τ|≦1ms (Formula 6). By the way, a sense of spaciousness or a sense of no direction occurs when the interaural cross-correlation is small, and in the case of only signals from a fixed direction, when |τ| The cross-correlation function has a large peak. The reason why |τ|<1ms is set here is that the time difference between the binaural signals fl(t) and fr(t) is usually
This is because it cannot exceed 1ms. First, the interaural cross-correlation function Φlr (τ) with only direct sound is fl (t) = p (t) * hol (t), fl (
t)
It can be obtained by substituting =p(t)*hol(t) into (Equation 6). By the way, the normalized autocorrelation function φlr(τ) for only the direct sound is It can be expressed as Note that φlr(τ) is temporarily hol
(t)hor(t), it becomes approximately 1. Here, Φll(o) and Φrr(o) represent the autocorrelation functions of the signals at the left and right ears, respectively, with τ=0. Next, if another reflected sound is added to the direct sound after the time when the autocorrelation function of the direct wave becomes small, then the normalized autocorrelation function φlr(τ) will be If it is equal to the function δ(t), then It is expressed as Here, Φlr(τ) is the interaural cross-correlation function of the nth reflection, and Φllll(o), Φrrr(o) are the self-correlation functions at τ=0 of the nth reflection of the left and right ears, respectively. It shows the correlation function. Note that when the sound source is in front of the user in a normal room, the maximum value of the interaural cross-correlation function is obtained extremely close to τ=0. By the way, the strength of the normalized interaural cross-correlation is
As IACC IACC=｜φlr(τ)｜max,for｜τ｜＜1ms
(Formula 9) is defined as: In addition, φlr(τ)=φlr(τ)/√()(
), and Φll(o) and Φrr(o) indicate the autocorrelation functions of the signal tones for the left and right ears at τ=0, respectively. The above describes the long-term autocorrelation function of the sound source signal, the multiple impulse responses caused by sound reflecting off walls, etc., and the interaural cross-correlation function that indicates the correlation between binaural signals. However,
Next, from these physical quantities, four elements that are important objective parameters representing the characteristics of the binaural sound signal are determined: hearing sound pressure, delay time of the first reflected sound, reverberation time of the subsequent reverberant sound, and how to find the optimal values of the interaural correlation function values. The results of the preference test conducted using the aforementioned music A and music B are shown in FIGS. 3a and 3b.
The horizontal axis is IACC, and the vertical axis is audible sound pressure (unit, dBA). The listening sound pressure can be considered to be the value of τ=0 of the autocorrelation function, and the value on the horizontal axis of the graph indicates the preference. From Figures 3a and b, it is clear that the optimal listening sound pressure [P]p does not depend much on IACC; it is 77 to 79 dBA for music A with a rather slow tempo, and 77 to 79 dBA for music B with a fast tempo. is found to be 79 to 80 dBA. It can be seen that the optimal value of the listening sound pressure is around 79 dBA in both cases. Next, according to the results of evaluating a synthesized sound field consisting of a direct sound and a single reflected sound when reproduced by a speaker using music and speech, the normalized autocorrelation function φp(τ) of the sound source signal is When the level of the reflected sound is varied by ±6 dB of the direct sound, the optimal delay time for the reflected sound is |Φp(τ)|, which is equivalent to 1/10 of the level G ₁ of the first reflected sound. It became clear that it would be appropriate to respond to that time. There｜
The horizontal axis is the time τd at which Φp(τ) | corresponds to 1/10 of the level G ₁ of the first reflected sound, and the vertical axis is the delay time [Δt ₁ ]p of the single reflected sound at which the preference is maximum. The diagram shown is FIG. 4. The range shown in the figure shows the delay time when it is 0.1 lower than the maximum value of the preference, and the symbol ○ in the figure indicates the first reflected sound level G ₁ = 6 dB, ● indicates the level G ₁ = 0 dB, □ is G ₁
= -6dB is shown. Especially |
If τp is the time when Φp(τ)| becomes 0.1 times Φp(o), then when G ₁ =0 dB, it can be expressed as τd = τp. Note that the τp of the aforementioned music A and B are 127 ms and 35 ms, respectively, as can be inferred from Figure 2 a and b.
As is clear from Figure 4, a straight line can be drawn to approximate each point in the figure, and when looking at both axes corresponding to the straight line, τd is the single reflection with the maximum preference. It can be seen that this almost matches the sound delay time [Δt ₁ ]p. At the same time |Φp(τ)|
This almost coincides with the time τp at which Φp(o) becomes 0.1 times. That is, if this is expressed mathematically, the most preferable delay time [Δt ₁ ]p of the first reflected sound can be expressed as [Δt ₁ ]p=τp (Formula 10). In addition, when τ<τp, |φp(τ)|=≦KG ^c ...(Formula 11), and in this case K=const(=0.1), C=const(=
1.0)

[Formula]. When using the normalized autocorrelation function Φp(τ), the above (Equation 11) becomes |Φp(τ)|≦0.1 when τ<τp
...(Formula 12) Furthermore, the autocorrelation function of the sound source signal has a close relationship with the optimal reverberation time, and this relationship is shown in FIG. The vertical axis is the optimal reverberation time of the subsequent reverberant sound [Tsub]
p, and the horizontal axis is τp. The reverberation time referred to here is not expressed as the time for the direct sound to attenuate by 60 dB, but as the time for the signal in the reverberation section to attenuate by 60 dB. In the figure, music A and music B are the same as those mentioned above, but music E is Mozart.
``Symphony in C major K, 551 Jupiter 4th movement, molto allegro'', and Speech S is ``Symphony in C major K, 551 Jupiter 4th movement, molto allegro'', and Speech S is Doppo Kunikida's ``Sound of the Tone River'', ``I'm surprised at how blue the sky is, how bright and high it is.'' (τp =12ms)
It is. As is clear from the figure, the relationship shown in FIG. 5 can be approximately approximated by the function [Tsub]p≈(23±10)τp. Next, Figure 6 is a graphical representation of the results measured in a synthetic sound field consisting of a direct sound and a single reflected sound. Preference and IACC values are shown. It can be seen from FIG. 6 that the preference increases as the value of IACC decreases. That is, the correlation function between the preference score and the IACC value is negative (-0.76: 1% significance level), and this holds true when IACC takes the maximum value at τ = 0. . It can also be read from the figure that the most effective way to reduce IACC is to make the early reflected sound arrive within a range of ± (55° ± 20°) from the front. Next, a preference measure using the above four parameters will be described. Incidentally, this scale of preference was determined through a comparative test, and each parameter independently influences the scale of preference. As a result, since the principle of superposition can be applied, it becomes possible to generalize sound field preference data obtained in concert halls etc. by normalizing each objective parameter by its optimal value. . The preference scale obtained in the comparative test will be explained below. First, FIG. 7 shows the preference measure S ₁ as a function of listening sound pressure. In this figure, the scale at optimal hearing sound pressure is set to zero. The value of S ₁ is almost symmetrical around the most preferable listening sound pressure (0 dB), but a shift towards a weaker listening sound pressure will result in a slightly lower preference than a shift towards a stronger sound pressure. The scale tends to be good. If this is expressed mathematically, S ₁ (LL)≦S ₁ (−LL) (Formula 13). Here, LL=20 ^log (p/[P]p), p is the audible sound pressure, and [P]p is the optimal audible sound pressure. Note that in the figure, ○ indicates the value of music A, and X indicates the value of music B. Next, the measure S ₂ of preference as a function of the delay time between the direct sound and the first reflected sound (first reflected sound delay time) is shown in FIG. The horizontal axis of this figure is normalized by the most preferred time delay [Δt ₁ ]p. By the way, it is known that the optimal value [Δt ₂ ]p of the delay time of the second reflected sound is [Δt ₂ ]p1.8τp, so when measuring the delay time of the second reflected sound, This optimal value was used as the delay time. Of course, it goes without saying that the condition of (Equation 10) exists. In the figure, ○, a, A, and △ are individual measurement results for music A, and X,
b, B, and △ are individual measurement results for music B, C for music C, D for music D, □ for music E, and ● for speech S. Music A, B, and E have already been mentioned above, but music C is Haydn's Symphony No. 102 in B flat major, second movement Adagio (τp = 65ms), and music D is Wagner's Siegfried. I
dyll) 332 measures (τp = 40ms). Furthermore, the sixth
As in the figure, the optimal delay time scale for the first reflected sound is set to zero. Furthermore, FIG. 9 shows a measure of preference S ₃ as a function of the reverberation time of subsequent reverberant sound. In the figure, the solid line is the total sound pressure of reflected sound. is the measure of preference _S3 as a function of the reverberation time of the subsequent reverberant sound in the case of , and the dashed line is that for the case of G=1.1. In the figure, ○ and a are experimental results for music A, X and b for music B, □ for music E, and ● for speech S, all of which are for the case of G=4.1. G = 4.1 corresponds to a situation where there is a lot of reverberation, such as when you are at the back of a concert hall, and G = 1.1
This corresponds to cases where there is a lot of direct sound, such as in the front seats of a hall. The preference scale at the most preferred reverberation time is set to zero. Next, the preference measure S ₄ as a function of IACC is shown in FIG. In the diagram, ○ is music A
When , X, b are the experimental results for music B. Due to the nature of IACC, if the maximum value is taken at τ=0, the sound image is in the front direction. As is clear from the figure, as IACC approaches 1, the preference measure S ₄ decreases rapidly. Therefore, it is necessary to make IACC smaller than 0.5 as much as possible. So far, four scales of preference S ₁ to
Although up to _S4 has been described, these four scales can be approximated by the approximate expressions shown below. First, the hearing sound pressure preference ^scale S ₁ shown in ^Fig _. 7 is S ₁ = _{-α 1 |} p) ...(Formula 15) In addition, α ₁ = {0.07±0.03, X ₁ >0, 0.04±0.02, X ₁ <
0} ...(Equation 16) Next, the preference measure S ₂ of the delay time of the first reflected sound shown in Fig. 8 is S ₂ = -α ₂ |X ₂ | ^3/2 ...(Equation 17) However _{, it} can _be ^expressed as In addition, α ₂ = {1.42±0.6, X ₂ >0, 1.11±0.5, X ₂ <0}
...(Formula 19). Furthermore, the scale S ₃ of the preference of the reverberation time of _the subsequent reverberant sound shown in FIG. 9 is S ₃ = _{−α 3} ^| Tsub]p) ...(Formula 21) In addition, α ₃ = {(0.74±0.25)G+(0.45±0.15), X ₃ >0
,
−(0.42±0.14)G+(2.36±0.79), X ₃ <0}
(Formula 22) If α ₃ is negative in (Formula 22), α ₃ =0 (Formula 22'). The remaining IACC preference measure S ₄ shown in Figure ₁₀ ^can be expressed as S ₄ = _{−α 4} | Can be done. Note that α ₄ =1.45±0.44 (Formula 25). Based on the principle of superposition, the overall measure S of preference in a concert hall or the like can be expressed as S= ₄ 〓 ⁱ⁼¹ Si (Formula 26). Of course, Si(i=1,
2, 3, and 4) indicate the four preference scales mentioned above. By using the preference scales S and Si (i=1, 2, 3, 4) obtained in this way, it is possible to accurately evaluate the sound field. This invention uses four parameters to measure preference in a concert hall or room: audible sound pressure in the sound field, delay time of the first reflected sound, reverberation time of the subsequent reverberant sound, and interaural cross-correlation coefficient. It is an object of the present invention to provide an acoustic device that can correct the sound field based on each parameter and create a more preferable sound field, focusing on the fact that each preference scale and the overall preference scale can be determined. It is. (1) Sound field evaluation measuring device related to the present invention First, one embodiment of the sound field evaluating measuring device related to the present invention will be described with reference to the drawings. FIGS. 11 and 12 are block diagrams showing the general structure of the sound field evaluation measuring instrument. Figure 12 is 4
FIG. 3 is a block diagram showing each parameter.
In the figure, 1 is a human head or dummy head, 2
1 and 2r are microphones inserted into the entrance of the ear canal, 3, 31, and 3r are preamplifiers, 4 is a physical quantity analyzer, 5 is a comparator, 6 is a psychological quantity converter, 7 is a comprehensive evaluator, 8 is a preference output terminal, 9 is a recorder for recording preferences, 10 is a sound field evaluation measuring device, 40 is an adder, 41 is a listening sound pressure analyzer, 42 is a first reflected sound delay time analyzer, 43
is a subsequent reverberation sound reverberation time analyzer, 44 is an interaural cross-correlation function analyzer, 51 is a listening sound pressure psychological quantity converter,
52 is a first reflected sound delay time psychological quantity converter, 53 is a subsequent reverberation sound reverberation time psychological quantity converter, 54 is an interaural cross-correlation coefficient psychological quantity converter, y, y1, y2, y
3 and y4 are comparison data input terminals. The respective optimum values are input to the comparison data input terminals y, y1, y2, y3, and y4, and the input terminal y1
A signal corresponding to the optimum hearing sound pressure ([P]p) is sent from the input terminal y2, a signal corresponding to the optimum first reflected sound delay time ([△t _{1 ]p) is sent from the input terminal y3, and a signal corresponding to the optimum hearing sound pressure ([△t 1} ]p) is sent from the input terminal y3. A signal corresponding to the subsequent reverberation sound reverberation time (Tsub]p) is input from the input terminal y4, and a signal corresponding to the optimal binaural cross-correlation coefficient is input. Note that the hearing sound pressure analyzer 41, first reflected sound delay time analyzer 42, subsequent reverberant sound reverberation time analyzer 43, and interaural cross-correlation function analyzer 44 shown in FIG.
This corresponds to a physical quantity analysis section corresponding to the physical quantity analyzer 4 in FIG. Listening sound pressure psychological quantity converter 5 in Fig. 12
1. First reflected sound delay time psychological quantity converter 52; Subsequent reverberation sound reverberation time psychological quantity converter 53; Binaural cross-correlation coefficient psychological quantity converter 54 is the comparator 5 and psychological quantity converter in FIG. This corresponds to the comparison and psychological quantity conversion section corresponding to 6. Similarly, the comprehensive evaluator 7 corresponds to a comprehensive evaluation section. First, a simple flow of one embodiment of this sound field evaluation measuring instrument will be explained using FIG. 11. First, place a dummy head or human head in the location where you want to evaluate the sound field, such as a concert hall or listening room.
Set up. The sound pressure signals of both ears are detected by microphones 21 and 2r attached to the left and right ear canal entrances, amplified and added by a preamplifier 3, and then amplified by a physical quantity analyzer 4. Physical quantities, that is, audible sound pressure, first reflected sound delay time, subsequent reverberation sound reverberation time, and interaural cross-correlation coefficient IACC are obtained from the sound pressure signals of both ears, and the respective values are input to the comparator 5 from the input terminal y. The psychological quantity converter 6 compares each optimum value input as comparison data, and then performs processing based on the already explained (Formula 14), (Formula 17), (Formula 20), and (Formula 23). This is converted into a psychological quantity, that is, a quantity corresponding to the preference, and then comprehensively evaluated in the comprehensive evaluator 7, and the preference in the sound field is outputted from the output terminal 8. This measuring instrument can be used to evaluate the sound field of concert halls and listening rooms. Next, the embodiment shown in FIG. 11 will be explained in detail with reference to FIG. 12, which shows the flow of four individual parameters. As mentioned above, the sound pressure signals detected by the microphones 21 and 2r are amplified by the preamplifiers 31 and 3r, and added by the adder 40, and then subjected to audible sound pressure analysis corresponding to the physical quantity analysis section. 41, the first reflected sound delay time analyzer 42, and the subsequent reverberant sound reverberation time analyzer 43, and in each analyzer, the audible sound pressure p, the first reflected sound delay time Δt ₁ , and the subsequent reverberant sound are input to Each reverberation time Tsub is analyzed and measured and output. Also, apart from these three trends,
Two preamplifiers 3 before being added by adder 40
The outputs from 1 and 3r are input to an interaural correlation function analyzer 44 corresponding to a physical quantity analysis section, where the maximum value IACC of the interaural cross-correlation coefficient is analyzed and output. Next, the audible sound pressure p output from each of the four analyzers 41, 42, 43, and 44 corresponding to the physical quantity analysis section, the first reflected sound delay time Δt ₁ , the subsequent reverberant sound reverberation time Tsub, and the maximum Interaural cross-correlation coefficient
The IACC is input to the listening sound pressure psychological quantity converter 51, the first reflected sound delay time psychological quantity converter 52, and the subsequent reverberant sound reverberation time psychological quantity converter 53, which correspond to the comparison and psychological quantity converting section, respectively. The optimum hearing sound pressure [P] p, the optimum first reflected sound delay time [△t ₁ ] p, and the optimum subsequent reverberation sound reverberation time [Tsub] p are input from each of the comparison data input terminals y1, y2, y3, and y4. , compared with the optimal interaural cross-correlation coefficient. The optimal binaural cross-correlation coefficient may be considered to be 0, but in practice there are many cases where it is acceptable to set it to 0.4 to 0.5 or less, so it is set to 0.4 in this embodiment. In this comparison and the calculation formula for psychological quantity conversion in the psychological quantity converter, the hearing sound pressure psychological quantity converter 51
In the first reflected sound delay time psychological quantity converter 52, the above-mentioned (formula 17) to (formula 19) are used, and in the subsequent reverberation sound reverberation time psychological quantity converter 53, are the aforementioned (Equations 20) to (Equations 22′),
In the interaural cross-correlation coefficient psychological quantity converter 54, (Equations 23) to (Equations 25) are as follows, and the preference scale S ₁ which is the output obtained from this comparison and psychological conversion section is S ₂ , S _{3 ,} and S ₄ can be obtained through calculation processing using a microcomputer program or the like. However, if accuracy is not a big issue, a conversion table as shown below may be used.

【table】

[Table] The above conversion tables are examples of conversion tables for the audible sound pressure psychological quantity converter 51 and the first reflected sound delay time psychological quantity converter 52, respectively. This conversion table contains the respective preference scales S 1 corresponding to the values of the listening sound pressure p, the optimum listening sound pressure [ _P ] p, the first reflected sound delay time △t ₁ and the optimum first reflected sound delay time. , S ₂ are stored. Although only the conversion tables for the audible sound pressure psychological quantity converter 51 and the first reflected sound delay time psychological quantity converter 52 are shown here, the following reverberation sound reverberation time psychological quantity converter 53 and the interaural cross-correlation coefficient psychological quantity converter 53 are shown. A similar conversion table can be created for the quantity converter 54 as well. The preference scale S ₁ output from the converters 51 to 54 of the comparison and psychological quantity converter in this way
~ _S4 is input to the comprehensive evaluator 7, and the overall preference measure S obtained as a result of the comprehensive evaluation is output to the output terminal 8. This output is recorded on the recorder 9. However, although the recording device 9 is included in this embodiment, it is not necessarily necessary to include it, and it goes without saying that it may be omitted or may be replaced with a display device. Next, the audible sound pressure analyzer 41 of the physical quantity analysis section used in the sound field evaluation measuring instrument 10 in FIG.
FIG. 13a shows an example of the configuration of the first reflected sound delay time analyzer 42, the subsequent reverberant sound reverberation time analyzer 43, and the interaural cross-correlation function analyzer 44.
This will be explained using ~e. FIG. 13a shows the basic configuration of the first reflected sound delay time analyzer 42 or the subsequent reverberant sound reverberation time analyzer 43 using a square integral type reverberation meter by MR Schroeder, which is a modified correlator. It has the function of the listening sound pressure analyzer 41 that also includes an output corresponding to sound pressure. FIG. 13b shows an example of a priority encoder used to output a code corresponding to the first reflected sound or reverberation time of the first reflected sound delay time analyzer 42 or the subsequent reverberant sound reverberation time analyzer 43. It is a diagram. FIGS. 13c and 13d are diagrams showing other configuration examples of the audible sound pressure analyzer. 1st
FIG. 3e is a diagram showing an example of the configuration of a circuit for determining the interaural cross-correlation coefficient. In Fig. 13a, 400 is an input terminal of a circuit used as a first reflected sound delay time analyzer or a subsequent reverberation sound reverberation time analyzer using a square integral type reverberation meter, and 401, 402, and 403 are delay time analyzers. Circuits, 404-411 are multipliers, 412-415 are integrators, 416-418 are adders, 419-42
1 is a comparator, 422 is an output terminal for the autocorrelation coefficient φ _p , and 423 to 425 are output terminals O ₁ to 425 that output information on the delay time of the first reflected sound or the subsequent reverberation time, respectively.
Output terminal of O _o-1 , 426 is attenuator, Fig. 13b
, 430 is a priority encoder into which O ₁ to _{O o-1} outputted from output terminals 423 to 425 are input, and x ₁ to _{x o-1} correspond to the first reflected sound delay time or the subsequent reverberation sound reverberation time. code output,
In Figures 13c and d, 435 is a rectifier circuit;
36 is a reduction pass filter, 437 is a multiplier, the first
In Figure 3e, 441 is an input terminal into which the left input signal Lin of the circuit for calculating the interaural cross-correlation coefficient is input;
442 is an input terminal into which the right input signal Rin is input, 443 and 444 are adders, 445 to 448 are root mean square circuits (RMS), 449 is an arithmetic circuit, and 45
0 is an output terminal that takes out an output corresponding to the cross-correlation coefficient. Note that in the circuit of FIG. 13a, the delay circuit,
(n-1) adders and comparators each; a multiplier and an integrator to which the output signal from the delay circuit is input;
Each has n multipliers and output terminals into which the output signal from the integrator is input, but in the drawing, (n-1) multipliers are represented by 3, and n multipliers are represented by 4. Only one is drawn, the rest are omitted, and the ones drawn in the figure are given consecutive symbols. Therefore, the (n-1)th delay circuit 403 following the second delay circuit 402, and the third to (n-2)th delay circuits are not labeled for convenience. First, FIG. 13a will be explained. When a signal is applied to the input terminal 400, the input signal is applied to the multipliers 404 to 407, and also to the delay circuit 4.
01 is also given. The output of the delay circuit 401 is given to the multiplier 405 and also to the next delay circuit 4.
The sequentially delayed signals also given to 02 are (n-1)
The output of the second delay circuit 403 is produced. The outputs of the respective delay circuits 401-402 are multiplied by multipliers 405-407. Multiplier 40
4 performs a square calculation of the input signal itself.
Next, the output signals of the multipliers 404-407 are integrated in integrators 412-415, respectively, and the output signals are further squared by multipliers 408-411. Note that the output signal of the integrator 412 is directly input to the terminal 42 in addition to the input to the multiplier 408.
2 is output. This signal has an autocorrelation coefficient φ(o)
In other words, it is the square of the power of the input signal, and includes information corresponding to the sound pressure signal. The processing of squaring the power can also have an equivalent function simply by converting it into an absolute value. By the way, the output signals of the multipliers 408 to 411 are the square of the autocorrelation coefficient φ ₀ ² ,
The signals of φ ₁ ² , ..., φ ² _o-1 are added to adders 416 to 418, and the outputs of adders 416 to 417 are respectively added to the next adder, and the (n-1)th adder 418 will be carried out until By the way, the multiplier 408
The output signals from the comparator circuits 419 to 42 and the output signals from the adders 416 to 417 are respectively output from the comparator circuits 419 to 42
1, the output signal of the last adder 418 is compared with the signal after being attenuated by the attenuator 426,
Output signals to output terminals 423 to 425
Outputs O ₁ ~423 ~ _{O o-1} 425. Note that the attenuation ratio of the attenuator 426 is set to (0.1) ² when determining the first reflected sound delay time, and to (0.001) ² when determining the subsequent reverberation sound reverberation time. This is similar to what has already been known as an optimal delay time detector, and is used to find the delay time at which the value of the autocorrelation function becomes 1/10 or 1/1000. Further, the total delay time from the delay circuit 401 to the delay circuit 403 is required to be about the length of the first reflected sound delay time and the subsequent reverberation sound reverberation time, that is, about 100 to 200 ms and 3 to 5 seconds, respectively.
Naturally, as this time elapses, the integrating circuits 412-415 need to be reset.
However, the reset circuit is not shown. Next, the priority encoder 430 shown in FIG. 13b is connected to the comparator 419 shown in FIG. 13a.
~421 output signal O ₁ 423 ~ _{O o-1} 425 is encoded by the attenuator 426 in FIG. 13a.
Displays the code x ¹ … x _o-1 corresponding to the delay time τ (= τp) (the delay time that first becomes smaller than the damping ratio) corresponding to the damping ratio of (0.1) ₂ or (0.001) ² ,
Or output. The following FIGS. 13c and d are diagrams showing other examples of circuits used as audible sound pressure analyzers, and in this case as well, the output signal φ ₀ is output from the output terminal 422 of FIG. 13a.
You can find out the audible sound pressure in the same way as you can.
This configuration includes a low-pass filter 436 and a rectifier circuit 435 or a square multiplier 4 as shown in FIG. 13c.
This is a circuit that combines 37. A circuit 519 for determining the interaural cross-correlation coefficient shown in FIG. 13e uses an example of a circuit system for determining the normalized correlation coefficient. First, when the input signal Lin is input from the input terminal 441 and the input signal Rin is input from the other input terminal 442, the input signals Lin and Rin are directly input to the root mean square circuit 445,
448, and output the respective root mean square values γ and δ. Further, an adder 443 adds a sum signal between input signals Lin and Rin, and an adder 444 adds a difference signal to the other signal by making one signal negative.
Take out each and use root mean square circuits 446, 447
Similarly, the sum signal and the difference signal are input to the respective root mean square values α and β. The four root mean square signals α, β, γ, and δ obtained in this way are processed in the calculation circuit 449, that is, (α ² −
β ² )/(4γδ) is performed, and an output φlr(o) corresponding to the cross-correlation coefficient, that is, an interaural cross-correlation coefficient (IACC) is obtained at the output terminal 450. However, the first
When using the circuit shown in Figure 3 e, the input signal Lin and
0 before Rin is input to input terminals 441 and 442
If a ~1ms delay circuit is provided, the delay time is varied, and the maximum value is determined, the circuit configuration corresponds to the above (Equation 9), and from that circuit, a value faithful to (Equation 9) can be obtained. You will be asked for it. A delay circuit for this purpose is not shown in FIG. 13e. In addition, to calculate the interaural cross-correlation coefficient (IACC) with higher accuracy, use the above-mentioned (Equation 7).
It is preferable to perform numerical calculations using the formulas up to (Formula 9). In the embodiment described so far, the audible sound pressure analyzer 41, the first reflected sound delay time analyzer 4
2 and the subsequent reverberant sound reverberation time analyzer 43 have been used as the sum signal of two acoustic signals entering the dummy head or the left and right ears of the human head 1, but usually the sum signal between the two signals is The delay time is 1 ms or less, and there is no significant difference in the analysis results of these three analyzers 41 to 43, so it is not possible to obtain sufficient analysis results even if the left or right acoustic signal is used. can. The optimum listening sound pressure [P]p of the sound field evaluation instrument 10, which is an embodiment of the first invention described above, varies depending on the type of sound source, but it is approximately (79±5 )
Since it is about dB, 79 dB is used here as a representative. In addition, the optimal first reflected sound delay time [△t ₁ ] p and the optimal subsequent reverberant sound reverberation time [Tsub] p are determined by the first reflected sound delay time analyzer 42 from the actual sound source itself.
First, find the delay time τp that is smaller than the attenuation ratio and use this as the optimal first reflected sound delay time [△
t ₁ ]p, and multiply it by (23±10) and use it as the optimal subsequent reverberation sound reverberation time [Tsub]p. In this example, [Tsub]p=23τp is used. According to this sound field evaluation measuring instrument, one or two microphones are used, an amplification device is used to amplify the acoustic signal obtained from the microphone, and the audible sound pressure p is measured from this amplified signal. the audible sound pressure measuring means for measuring the audible sound pressure;
a first reflected sound delay time measuring means for measuring the reflected sound delay time Δt ₁ ; a subsequent reverberant sound reverberation time measuring means for similarly measuring the subsequent reverberant sound reverberation time Tsub from the amplification signal; Similarly, a binaural cross-correlation coefficient measuring means for measuring the binaural cross-correlation coefficient IACC from the amplified signal, and the audible sound pressure p
and the optimum listening sound pressure [P] p, and when the above listening sound pressure is larger than the optimum listening sound pressure, -( 0.07
±0.03) is multiplied by -(0.04±0.02) when it is small, and the value is output as the preference (psychological quality of hearing) S _1. In other words, the above (Equations 14) to (Equations 16) a first comparison conversion means that produces a corresponding output; and the absolute value of the logarithm to the base 10 of the ratio between the first reflected sound delay time Δt ₁ and the optimal first reflected sound delay time [Δt ₁ ]p. to the 3/2 power of the above first
When the reflected sound delay time is larger than the optimal first reflected sound delay time, the value obtained by multiplying by -(1.42±0.6), and when it is smaller than the optimum first reflected sound delay time, multiplied by -(1.11±0.5) is output as the preference scale _S2 . (formula
17) A second comparison conversion means that outputs an output corresponding to (Equation 19), and a logarithm to the base 10 of the ratio of the above-mentioned subsequent reverberant sound reverberation time Tsub and the optimal subsequent reverberant sound reverberation time [Tsub] p. When the reverberation time of the subsequent reverberation is larger than the optimal reverberation time of the subsequent reverberation, the sound pressure of the entire reflected sound is -(0.74±0.74±
0.25) times the value and -(0.45±0.15), and when it is small, the total sound pressure of the reflected sound is -(0.42±0.14).
If the value after multiplying the multiplied value by the addition value of -(2.36±0.79) is negative, the value is output as is, and if it is positive, zero is output as the preference scale S _3. In other words, the above (Equation 20 ) to (Equation 22′), and the value of the 3/2 power of the interaural cross-correlation coefficient is −(1.45±0.44).
a fourth comparison conversion means that outputs the value multiplied by , as the preference measure _S4 , that is, output corresponding to the above-mentioned (Equation 23) to (Equation 25); and the first to fourth comparison and conversion means. Since it is equipped with a comprehensive evaluation means that comprehensively evaluates the preference scales S ₁ , S ₂ , S ₃ , and S ₄ output from the above and outputs the overall preference scale S, accurate and faithful sound field evaluation is possible. It can be carried out. (2) Acoustic device related to the present invention Next, based on four physical parameters: the audible sound pressure in the sound field, the delay time of the first reflected sound, the reverberation time of the subsequent reverberant sound, and the interaural cross-correlation coefficient. hand,
DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an acoustic device related to the present invention that can correct a sound field and create a more preferable sound field will be described with reference to the drawings. (2-1) First acoustic device related to the present invention FIG. 14a is a configuration block diagram of an embodiment of the first acoustic device that can correct the sound field and create a more preferable sound field. Fig. 14b shows one example of the sound field expansion device included in the acoustic device of Fig. 14a, and Fig. 14c shows another example of the reverberation device included in the acoustic device of Fig. 14a. It is. First, in Fig. 14a, 501 is an input terminal
IN _R , 502 is the input terminal IN _L , 503a to 503
d is an adder, 504a and 504b are reverberation devices, 5
05a, 505b, 505r, 505l are attenuators, 506a, 506b are delay circuits, 507 is 1
A scaler which is a seed frequency dividing circuit, 508a, 508
b is an up-down counter, 509 is a digital-to-analog converter, 510a and 510c are first reflected sound delay time analyzers, 510b is a subsequent reverberation sound reverberation time analyzer, 511a to 511c are encoders,
512a, 512b, 513a, 513b are comparators, 514a, 514b, 514c are smoothing circuits,
515a, 515b are variable resistors for setting target values, 516 is a sound field expansion device, 517r, 5
17l is a power amplifier for driving a speaker, 518r and 518l are speakers, 519 is an interaural cross-correlation coefficient (IACC) calculator, 520 is a multiplier, 531 and 532 are reverberation devices 504a and 50
4b to the input terminal inr to the sound field expansion device 516,
inl, 521 is the CLOCK signal input terminal, 533, 5
34 is an output terminal outr of the sound field expansion device 516;
outl, 535 is an input terminal attm from the comparator 513b to the sound field expansion device 516. In FIG. 14b, 530a to 530d are adders, 536 is an attenuator, 537 is a low pass filter, 538 is a high pass filter, and 539 is a phase shift circuit or delay circuit. In Figure 14c, 504c is
Reverberators by MR, Schroeder and BS Atal, 540a to 540i are adders, 541a and 451b are phase inversion circuits, 54
2a, 542b are voltage followers, 543a
543f is a delay circuit, 544a to 544f are attenuators, 545 is an input terminal of attenuator 504c, 54
6 is an output terminal of the attenuator 504c. Note that the same reference numerals in the drawings as those in other drawings indicate the same or corresponding parts. Next, the operation of one embodiment of the first acoustic device will be described using FIGS. 14a, 14b, and 14c. First, record, tape, etc. are input to input terminals IN _R 501 and IN _L 502.
An acoustic signal from a microphone, radio, etc. is provided.
These acoustic signals are added by an adder 503c and input to the first reflected sound delay time analyzer 510c. This first reflected sound delay time analyzer 510c is constituted by the circuit 42 in FIG. 12 described above, that is, the same circuit as shown in FIG. The attenuation ratio is set to 1/10 so that time can be measured. This first
The output of the reflected sound delay time analyzer 511 is input to the encoder 511, and from the encoder 511, the optimal delay time of the first reflected sound [△t ₁ ] depending on the sound source is input.
An output ([△t ₁ ]p=τp) obtained by setting p to τp is output. The output from the encoder 511c is input to the comparator 512a and also to the multiplier 520, multiplied by 23 and given to the comparator 512b as the optimum subsequent reverberation sound reverberation time [Tsub]p (=23·τp). It will be done. On the other hand, the signals coming into the dummy head placed in a sound field such as a listening room or the microphones 2r and 2l placed at the left and right ears of the human head 1 are amplified by preamplifiers 3l and 3r. , the outputs are added by an adder 503d, and then input to the same first reflected sound delay time analyzer 510a as 42 in FIG. 12 as before, and also to the subsequent reverberant sound reverberation time analyzer 510b. is input. The configuration of this subsequent reverberation sound reverberation time analyzer 510b is 43 in FIG.
It has the same circuit configuration as that described in . Next, the output of the first reflected sound delay time analyzer 510a is input to the encoder 511a, and a signal corresponding to the delay time Δt ₁ of the first reflected sound after being affected by the sound field is obtained. After this, comparator 512a
is input. Further, the output of the subsequent reverberation sound reverberation time analyzer 510b is input to the encoder 511b, and a signal corresponding to the subsequent reverberation sound reverberation time Tsub after being affected by the sound field is obtained, and this signal is sent to the comparator 512b. is input. By the way, in the comparator 512a, the optimum first reflected sound delay time [Δt ₁ ]p from the encoder 511c and the delay time Δt ₁ of the first reflected sound after being affected by the sound field from the encoder 511a are determined. are compared, and if the first reflected sound delay time △t 1 is larger than the optimal first reflected sound delay time [△t ₁ ]p, the first reflected sound delay time △ _{t 1 is} In order to reduce the comparator 51
A countdown signal CD is applied from 2a to the counter 508a to reduce the content of the counter 508a. Therefore, in the scaler 507 that divides the frequency based on the content of the counter 508a, the value to be divided becomes small, and the value to be divided is reduced from the input terminal 521. of
DELAY obtained by dividing the CLOCK signal
The frequency of the CLOCK signal increases and its DELAY
Reverberation circuits 504a and 504b are activated by the CLOCK signal.
A delay circuit 506a constituted by a BBD (bucket bridge device) etc.
The operation is such that the delay time of 506b is reduced. However, if the first reflected sound delay time Δt ₁ is smaller than the optimal first reflected sound delay time [Δt ₁ ]p, the comparator 512a gives a count-up signal CU to the counter 508a. , the content of the counter 508a becomes larger, the value divided by the scaler 507 becomes larger, the frequency of the DELAY CLOCK signal becomes lower, and the delay time of the delay circuits 506a, 506b becomes larger, which is the opposite operation to the above case. do. Furthermore, in the comparator 512b, the multiplier 52
The optimum subsequent reverberant sound reverberation time [Tsub] p of 0 is compared with the subsequent reverberant sound reverberation time Tsub after being given the influence of the sound field from the encoder 511b, and if the optimum subsequent reverberant sound reverberation time [Tsub] If the subsequent reverberant sound reverberation time Tsub is large with respect to p,
A count-up signal CU is sent to the counter 508b, and the contents of the counter 508b are converted to an analog value by the digital-to-analog converter 509.
Attenuators 505a and 5 of 4a and 504b, respectively.
05b, and by that signal, the attenuator 5
The attenuation values at 05a and 505b become larger,
For this purpose, the subsequent reverberation sound reverberation time Tsub is reduced. However, if the subsequent reverberation sound reverberation time Tsub is smaller than the optimal subsequent reverberation sound reverberation time [Tsub] p, the counter 5
A countdown signal CD was sent on 08b,
The contents of the counter 508b are converted into analog values by a digital-to-analog converter 509, and then passed through a smoothing circuit 514c to attenuators 505a and 505b.
and the signal causes the attenuators 505a,
The attenuation value at 505b becomes smaller, and therefore the reverberation time Tsub is increased, which is the opposite of the above case. By the way, each amplified signal from the preamplifiers 3l and 3r before being added is input to the binaural cross-correlation coefficient calculator 519, and the binaural cross-correlation coefficient
The IACC is determined, converted into an analog signal, and input to the next smoothing circuit 514b. Next, the comparator 513b compares the signal from the smoothing circuit 514b with the target voltage signal from the variable resistor 515b, which sets the target value of the interaural cross-correlation coefficient IACC. If the signal from the smoothing circuit 514b, that is, the signal corresponding to the interaural cross-correlation coefficient IACC, is larger than the target signal from the variable resistor 515b, the output signal from the comparator 513b becomes smaller. , input terminal att in535 of the sound field expansion device 516
is input, and the sound field expansion device 516 operates to further reduce the binaural cross-correlation coefficient IACC. However, if the signal corresponding to the binaural cross-correlation coefficient IACC is smaller than the target signal from the variable resistor 515b, the output signal from the comparator 513b becomes larger and is input to the sound field expansion device. Input to terminal att in535, sound field expansion device 516
operates in the opposite way to the above case, such as increasing the interaural cross-correlation coefficient IACC. By the way, the first reflected sound delay time analyzer 510a receives the added signal added by the adder 503d.
outputs a sound pressure signal φ ₀ corresponding to the listening sound pressure, and after passing this sound pressure signal φ ₀ through a smoothing circuit 514a, it is input to a comparator 513a. In this comparator 513a, a target voltage signal from a variable resistor 515a that sets a target volume value,
Comparing the signal from the smoothing circuit 514a, if the output signal from the smoothing circuit 514a corresponding to the listening sound pressure is larger than the target volume value from the variable resistor 515a, the output signal from the comparator 513a The output of the attenuators 505r and 505l increases.
The attenuation of the sound increases, lowering the audible sound pressure. Furthermore, if the signal corresponding to the listening sound pressure is smaller than the target volume value, the output from the comparator 513a becomes smaller, and the attenuation of the attenuators 505r and 505l becomes smaller, causing the listening sound pressure to decrease. The operation is the opposite of the above case, such as raising it. Note that the smoothing circuit 514 shown in FIG. 14a
a, 514b, and 514c are all provided with delay circuits 506a and 506a in the reverberation circuits 504a and 504b in order to prevent the generation of noise due to sudden signal changes.
The delay time 506b is also used to prevent sudden changes. Next, the operation of the sound field enlarging device 516 will be explained using the sound field enlarging device shown in FIG. 14b. First, the input terminals from the reverberation devices 504a and 504b
The difference component between the two signals input to inr 531 and inl 532 is obtained by adder 530a, passes through attenuator 536, and passes through low-pass filter 537. Then, the adders 530b and 530c input the input terminal
It is added to the main paths from inr 531 to output terminal outr 533 and from input terminal inl 532 to output terminal outl 534, respectively. Apart from that, the output of the low-pass filter 537 further passes through a high-pass filter 538 and is given a delay or a phase-shift rotation in a phase-shift or delay circuit 539, and then the crosstalking phase is also removed. Adders 530d and 530e add the signals to the respective main paths described above with opposite phases, and the signals after the addition are output to output terminals outr533 and outl534, respectively. By changing the attenuator 536 of the sound field expansion device 516, the interaural cross-correlation coefficient IACC can be changed. Further, this attenuator 536 is set so that the attenuation increases as the value of the signal from the input terminal att in 535, that is, the output signal from the comparator 513b increases. Furthermore, what is shown in FIG. 14c is a reverberator 504c made by MR Schroeder and BS Atal, and this reverberator 504c can be used in place of the reverberation devices 504a and 504b shown in FIG. 14a. A more natural reverberation effect can be obtained. Note that in the above embodiment, the microphone 2
It has been described that r and 2l are used to obtain the sound pressure at the entrance of the ear canal of a dummy head or head 1, but if it is not necessary to obtain the value of the interaural cross-correlation coefficient IACC, a single microphone may be used. Binaural cross-correlation coefficient
Even when determining IACC, it may be possible to do so using two microphones that are not attached to a dummy head or the like. Although the acoustic signal is shown as a two-channel stereo signal, it goes without saying that there are cases in which a two-channel signal is not necessarily required. As mentioned above, in the first acoustic device,
Optimal first reflected sound delay time measuring means for measuring the optimal first reflected sound delay time [△t ₁ ] p from the input terminals IN _R and IN _L and the acoustic signals input from the input terminals IN _R and IN _L ; Its optimal first reflected sound delay time [△t ₁ ]
Optimum subsequent reverberation sound reverberation time output means for outputting an optimum subsequent reverberation sound reverberation time [Tsub] p corresponding to (23±10) times p; listening sound pressure measuring means for measuring the hearing sound pressure φ 0 at φ ₀ ; first reflected sound delay time measuring means for measuring the first reflected sound delay time Δt ₁ from the sound field signal; and subsequent reverberation from the sound field signal. Subsequent reverberation sound reverberation time measuring means for measuring sound reverberation time Tsub and interaural cross-correlation coefficient from the sound field signal.
The interaural cross-correlation coefficient measurement means for measuring IACC compares the above-mentioned optimal first reflected sound delay time [△t ₁ ]p and the above-mentioned first reflected sound delay time △t ₁ , and calculates the value according to the difference. a first comparison means that outputs a signal; a second comparison means that compares the optimum subsequent reverberation sound reverberation time [Tsub]p and the abovementioned subsequent reverberation sound reverberation time Tsub, and outputs a signal according to the difference; The delay time changes depending on the output signal of the first comparison means, the reverberation time changes depending on the output signal of the second comparison means, and reverberation sound is added to the acoustic signal input from the input terminal. a binaural cross-correlation coefficient setting means that can set a target binaural cross-correlation coefficient value in advance, and a binaural cross-correlation coefficient IACC and a binaural cross-correlation coefficient a third comparing means for comparing the set value of the relational coefficient setting means and outputting a signal according to the difference; , a sound field expanding means having an output capable of changing the interaural cross-correlation coefficient IACC of the sound field, an audible sound pressure setting means capable of setting a target audible sound pressure value in advance, and the above-mentioned sound field. a fourth comparing means that compares the audible sound pressure φ ₀ of the audible sound pressure with the set value of the audible sound pressure setting means and outputs a signal according to the difference; Since it is equipped with an attenuation means that can change the attenuation rate according to the output signal of the comparison means 4, and an electroacoustic conversion means that amplifies the signal from the attenuation means and radiates an acoustic signal into the space, the sound Not only can a more favorable sound field be created by correcting the field, but also the optimal first reflected sound delay time [△t ₁ ] p and optimal reverberation time [Tsub] for the sound source within this first acoustic device.
p is obtained, and the values of the listening sound pressure and the interaural cross-correlation coefficient IACC can be changed according to the user's preference. (2-2) Second acoustic device related to the present invention Next, in a sound field device for a sound field such as a Japanese-style room with relatively few reflections or a vehicle interior where there are many reflections but the reverberation time is extremely short, , the audible sound pressure in the sound field, the first reflected sound delay time, and the subsequent reverberation sound reverberation time, by correcting the sound field based on only three physical parameters, it is sufficient to create a more preferable sound field. It goes without saying that the above-mentioned first acoustic device can be used as an acoustic device in the sound field targeted here, but the sound field used is a small room like this, a Japanese-style room with relatively little reflection, or a car. When limited to a room, the acoustic device of the present invention not only can obtain the same effect as the first acoustic device described above without using the parameter of the interaural cross-correlation coefficient IACC, but also It is more compact and inexpensive, and has a simpler circuit configuration than the first acoustic device, so it is easier to work with and has better durability than the first acoustic device. It can be effective. Therefore, the acoustic device of the present invention is capable of creating a more preferable sound field by correcting the sound field using three physical parameters: the audible sound pressure in the sound field, the first reflected sound delay time, and the subsequent reverberation sound reverberation time. An example of this will be described with reference to the drawings. FIG. 15 is a block diagram of an embodiment of the acoustic device of the present invention. In the figure, 451 to 453 are absolute value conversion circuits, 5
11d is an encoder, 551 is an input terminal IN _R , 5
52 is an input terminal _INL , and 553 is an integral reset signal input terminal. 554 is the first reflected sound delay time analyzer, which is shown in the circuit diagram of FIG. 13a and FIG. 14a described above.
First reflected sound delay time analyzer 510a, 510
c. The internal configuration is slightly different from that of the subsequent reverberation time analyzer 510b, but the functionality is almost the same, and the attenuator 42 is added to the configuration of a normal autocorrelator.
6 and a comparator 419 are added. Even in the case of such a configuration, it operates in the same manner as the first reflected sound delay time analyzer 510a, 510c or the subsequent reverberant sound reverberation time analyzer 510b described above. Note that the same reference numerals as those in FIG. 13a and FIG. 14a indicate the same or equivalent parts. First, audio signals are input from the input terminals IN _R 551 and IN _L 552, and the two signals are sent to the adder 5.
03c and input to the first reflected sound delay time analyzer 544. This first reflected sound delay time analyzer 544 performs the same operation as the first reflected sound delay time analyzers 510a and 510c in FIG. 14a, and the output therefrom is the same as in the case of FIG. 14a described above. is input to the encoder 511d, and the optimum first reflected sound delay time [Δt ₁ ]p for the acoustic signal is output. A signal from the encoder 511d is given to the scaler 507 so that a delay time corresponding to this optimal first reflected sound delay time [Δt ₁ ]p is obtained. The scaler 507 divides the frequency of the CLOCK signal from the input terminal 521 and outputs the DELAY signal.
Create a CLOCK signal and reverberation circuits 504a, 504
DELAY to delay circuits 506a and 506b of
CLOCK signal is sent. Audio signal is input terminal
Reverberation circuit 5 from IN _R 551 and IN _L 552 respectively
The signals are inputted to 04a and 504b, given reverberant sound, and sent to attenuators 505r and 505l. On the other hand, the signal φ ₀ corresponding to the sound pressure of the first reflected sound delay time analyzer 554 is passed through the smoothing circuit 514a to the comparator 5.
13a and is compared with the voltage of a variable resistor 515a which sets a target value of sound pressure. If the sound pressure is smaller than the target value, the comparator 51
The output of 3a becomes small, and the attenuators 505r and 50
5l operates to reduce the amount of attenuation by the signal sent from the comparator 513a. However, if the sound pressure is larger than its target value, the output of the comparator 513a will be larger,
The attenuators 505r and 505l are connected to the comparator 513a.
The operation is opposite to the above case, in that the amount of attenuation is increased by the signal sent from the receiver.
Attenuators 505r and 505 thus obtained
The output of 1 passes through power amplifiers 517r and 517l and is reproduced by speakers 518r and 518l. At this time, the optimal subsequent reverberation sound reverberation time [Tsub]p, which is the optimal value of the reverberation time, is (23±10) of the optimal first reflected sound delay time [△t ₁ ]p, as mentioned earlier.
It is desirable to double the amount. Reverberation circuit 504a, 5
When the reverberator is configured as shown in 04b, the subsequent reverberant sound reverberation time Tsub=-ε・△t ₁ /
It is known that log(g). Note that g is the attenuation rate. Here, by using the relationship Tsub=(23±10) _Δt1 , it can be seen that the attenuation rate g must be in the range of 0.588 to 0.811. In addition, when the subsequent reverberant sound reverberation time Tsub=23△t ₁ , the attenuation rate g=
It is 0.741. That is, by setting the attenuation rate g in this way, the optimal first reflected sound reverberation time [Δt ₁ ]p and the optimal subsequent reverberation sound reverberation time [Tsub]p can be obtained.
Of course, even with such processing, effects in the sound field will be added, but even in spaces with relatively little reflection, such as a Japanese-style room, or with many reflections, such as the interior of a car,
This is very effective in spaces where the reverberation time is extremely short. By the way, in the acoustic device of the present invention, in a sound field such as a small room, a Japanese-style room with relatively little reflection, or a car interior, it is possible to provide a simpler acoustic device that has the same or better effect as the above-mentioned acoustic device. However, when the acoustic device of the present invention is used in a sound field other than a small room or a vehicle interior, the performance is slightly lower than that of the above-mentioned acoustic device.
However, since the sound field is corrected using three physical parameters: audible sound pressure, first reflected sound delay time, and subsequent reverberation sound reverberation time, it is possible to create a much more favorable sound field than conventional acoustic equipment. Moreover, the acoustic device of the present invention has advantages over the above-mentioned acoustic devices in that it is compact, inexpensive, has a simple circuit configuration, has good workability, and has good durability. As described above, in the audio device of the present invention, the input terminals _INR , _INL , and the input terminals INR, _INL ,
Optimal first reflected sound delay time measuring means for measuring the optimal first reflected sound delay time from the acoustic signal coming in from IN _L ; and audible sound pressure corresponding signal measurement for measuring a signal corresponding to the audible sound pressure from the acoustic signal. means, receiving a signal corresponding to the optimum first reflected sound delay time as a control signal, inputting the acoustic signal, adding a delay time corresponding to the optimum first reflected sound delay time to the acoustic signal, and The subsequent reverberant sound reverberation time is (23±10) of the above optimal first reflected sound delay time.
a reverberation means set to double, a listening sound pressure setting means that can set a target listening sound pressure value in advance, and a listening sound pressure that is the output of the listening sound pressure corresponding signal measuring means and the listening sound pressure setting means. and a comparison means that outputs a signal corresponding to the difference, and an output signal of the reverberation means is inputted, and the input sound from the reverberation means is Since it is equipped with attenuation means that can change the attenuation rate given to the signal, and electroacoustic conversion means that amplifies the signal from the attenuation means and radiates the acoustic signal into space, it corrects the sound field, Not only can a more favorable sound field be created, but also an optimal first reflected sound delay time and an optimal subsequent reverberation sound reverberation time can be obtained within the acoustic device of the present invention, and the listening sound pressure can be changed according to the user's preference. This has the effect of making it possible to As described above, according to the present invention, the acoustic characteristics of the sound field in which music is listened to are measured, and the physical and psychological quantities of the listening sound pressure, the first reflected sound delay time, the subsequent reverberant sound reverberation time, and the like are measured. Based on the evaluation using parameters such as the interaural cross-correlation coefficient, it is possible to provide an acoustic device that changes the physical quantity and enables more preferable music listening by measuring this physical quantity. . In addition, the comparators 512 and 41 in the above-mentioned audio device
It is configured to obtain an output of 9,420 and create a DELAY CLOCK by a scaler 507, and the first reflected sound delay time can be controlled by this DELAY CLOCK. This DELAY
By making the value of CLOCK manually controllable from the outside, it is possible to configure it so that it can be changed to correspond to various kinds _of music. This is especially useful in cases where there is a problem, and it has the advantage that it can be manufactured at a low cost.

[Brief explanation of the drawing]

Figure 1 shows the relationship between the reflective wall, the sound source, and the human head. Figure 2 a, b shows an example of the normalized autocorrelation function measured using actual music. Figure 3 a, b shows the relationship between the reflective wall, the sound source, and the human head. A diagram showing the results of a preference test conducted using music A and music B. Figure 4 shows τd and [△t ₁ ] p
Figure 5 is a diagram showing the relationship between τd and [Tsub]p, Figure 6 is a diagram showing the relationship between normalized preference and IACC value, and Figure 7 is relative audible sound. Figure 8 shows the relationship between the pressure and the preference measure S _1. Figure 8 shows the relationship between the delay time between the direct sound and the first reflected sound and the preference measure S _{2. Figure 9 shows the relationship between the preference measure S 2} and the subsequent reverberant sound. Figure 10 is a diagram showing the relationship between the reverberation time and the preference scale _S3 .
Figures 11 and 12 are diagrams showing the relationship between IACC and preference scale _S4 ; Figures 11 and 12 are block diagrams showing a schematic configuration of an embodiment of a sound field evaluation instrument related to the present invention; and Figure 13a. , b is a diagram showing the basic configuration of a first reflected sound delay time analyzer or a subsequent reverberant sound reverberation time analyzer using a square integral type reverberation meter, and Figures 13c and d are diagrams showing the audible sound pressure analyzer and other components FIG. 13e is a diagram showing an example of the configuration of a circuit for calculating the interaural cross-correlation coefficient, and FIG. 14a is the configuration of an embodiment of the first acoustic device related to the present invention. FIGS. 14b and 14c are block diagrams of a sound field expansion device and a reverberator, and FIG. 15 is a block diagram of an embodiment of a second acoustic device related to the present invention. In the figure, 1 is a human head or dummy head, 2
r, 2l are microphones, 3, 3r, 3l are preamplifiers, 4 is a physical quantity analyzer, 5 is a comparator, 6 is a psychological quantity converter, 7 is a comprehensive evaluator, 8 is an output terminal,
9 is a recorder, 10 is a sound field evaluation measuring device, 41 is a listening sound pressure analyzer, 42, 510a, 510c, 554
5 is a first reflected sound delay time analyzer, 43, 510b is a subsequent reverberation sound reverberation time analyzer, 44 is an interaural cross-correlation function analyzer, 51 is an audible sound pressure psychological quantity converter, 5
2 is a first reflected sound delay time psychological quantity converter; 53 is a subsequent reverberation sound reverberation time psychological quantity converter; 54 is a binaural cross-correlation coefficient psychological quantity converter; 451 to 453 are absolute value conversion circuits; 501; 502,521,531,5
32, 535, 551, 552 are input terminals, 50
3a to 503d are adders, 504a, 504b are reverberation devices, 505a, 505b, 505r, 50
5l is an attenuator, 506a and 506b are delay circuits,
507 is a scaler, 508a and 508b are up-down counters, 509 is a digital-to-analog converter, 511a to 511d are encoders, 51
2a, 512b, 513a, 513b are comparators,
514a to 514c are smoothing circuits, 515a, 51
5b is a variable resistor, 516 is a sound field expansion device, 51
7r, 517l are power amplifiers, 518r, 51
8l is a speaker, 519 is an interaural cross-correlation coefficient calculator, 520 is a multiplier, and 533 and 534 are output terminals. In addition, in the figures, the same reference numerals indicate the same or corresponding parts.

Claims (1)

[Claims] 1. Measure the optimal first reflected sound delay time from the input terminal and the acoustic signal coming from the input terminal,
an optimum first reflected sound delay time measuring means for outputting, an audible sound pressure corresponding signal measuring means for measuring a signal corresponding to the audible sound pressure from the acoustic signal, and a signal corresponding to the optimum first reflected sound delay time. is received as a control signal, and the acoustic signal is inputted, and a delay signal corresponding to the optimum first reflected sound delay time is added to the acoustic signal, and the subsequent reverberation sound reverberation time is set to the optimum first reflected sound delay time. 1 reflection sound delay time (23
±10) reverberation means set to double, audible sound pressure setting means that can set the target audible sound pressure value in advance, and audible sound pressure that is the output of the audible sound pressure corresponding signal measuring means and the above audible sound pressure. A comparison means for comparing the set value of the sound pressure setting means and outputting a signal corresponding to the difference, and an output signal of the reverberation means are input, and a signal is input from the reverberation means in accordance with the output signal of the comparison means. an acoustic device comprising attenuation means capable of changing the attenuation rate given to an acoustic signal; and electroacoustic conversion means for amplifying the signal from the attenuation means and radiating the acoustic signal into space. .