JP2003279400A - Tone quality estimating equipment and method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide tone quality estimating equipment and a tone quality estimating method which change a tone to be estimated to a physical quantity so as to fit to feeling in human acoustic sense, and enable tone quality estimation. <P>SOLUTION: The tone quality estimating equipment is constituted of a noise level calculating means 3 for measuring a noise level NLi of the tone to be estimated; an accumulated variation level calculating means 2 for measuring variation with time of a tone, summing a peak difference dBn of the variation level, and calculating an accumulated variation level CFLi; and an estimating means 4 for showing a relation of estimation points of a previously set noise level and the accumulated variation level. The accumulated variation level calculating means 2 consists of a converting means 1, an envelope converting means 11, a primary delay system response means 12, a variation level calculating means 13, and a variation level accumulating means 14. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound quality evaluation apparatus that collects a sound to be evaluated and evaluates the sound quality from the noise level of the sound and a cumulative fluctuation level calculated based on a time fluctuation component and a frequency component.

[0002]

2. Description of the Related Art Conventionally, in evaluating the sound quality of a sound emitted from a product, the noise level dB (A) of the sound to be evaluated is sometimes measured by a sound level meter to evaluate the noise level. It was not possible to obtain an evaluation that was sufficiently suitable for the feeling at. Furthermore, when a feeling evaluation (impression evaluation) for evaluating the quality of sound produced by a product by a subject's hearing and an evaluation by SD method are used, these evaluation methods can directly evaluate the likes and dislikes of humans, but the physical quantity and There is no direct consistency. Therefore, as a sound quality improvement measure in product development,
It was unsuitable for use as an index for finding improvement points. Therefore, in evaluating the quality of the sound emitted by a product, frequency analysis of the sound to be evaluated is performed, the information of the frequency band necessary for improving the sound quality is quantified as a physical quantity, and the sound quality is evaluated according to this physical quantity. Has been done. An example of a conventional sound quality evaluation apparatus is disclosed in Japanese Patent Laid-Open No. 10-2677.
No. 43 and Japanese Patent Laid-Open No. 07-306087.

Here, the human sense of hearing will be briefly described with reference to FIG. Generally, the auditory mechanism inputs sound waves from the outer ear into the ossicles as vibrations via the eardrum of the inner ear, transmits the vibrations of the ossicles to the basement membrane in the cochlea, and in this basement membrane, the positions of the front side to the back side are varied. Vibrations of lower frequencies than high frequencies are sequentially discriminated and transmitted to hair cells, and each hair cell converts the vibration of each frequency into an electric signal and transmits it to the brain via nerve fibers. It is known that the hair cells here generate a time delay due to a chemical change when converting vibration into an electric signal.

[0004]

By the way, in the above-mentioned respective conventional examples, in order to evaluate the sound quality, after the frequency analysis of the sound to be evaluated, the sound quality is evaluated mainly by quantifying each frequency region. . Therefore, these conventional examples are insufficient to evaluate the temporal variation and smoothness, which are temporal sound quality factors. That is, as outlined with reference to FIG. 13, the human auditory mechanism includes a time delay in addition to the frequency discrimination of the vibration of the sound to be evaluated. It is presumed that it is necessary to quantify and evaluate the quality of the sound in consideration of such aural characteristics, so as to suit the human hearing feeling such as the timbre and sound quality of the sound.

Therefore, for example, in the evaluation of diesel noise of a diesel engine, a time-varying sound, that is, a rattling sound that tends to be annoying and an intermittent sound called a "crispy sound" When evaluating sound quality, conventional sound level evaluation alone or frequency analysis to quantify sound is similar to the human hearing mechanism that includes a time delay, that is, human hearing. There was no evaluation of the tone and sound quality of the sound that would be adequately matched.

For this reason, it is presumed that it is effective to consider the temporal variation element as in the case of the human auditory mechanism including the time delay, in order to make the sound quality into a physical quantity. An object of the present invention is to provide a sound quality evaluation apparatus and a sound quality evaluation method capable of performing sound quality evaluation by physically quantifying a sound to be evaluated so as to match a human hearing feeling based on the above problems. And

[0007]

According to a first aspect of the present invention, there is provided noise level calculation means for measuring the noise level of a sound to be evaluated,
A cumulative fluctuation level calculating means for measuring the time fluctuation of the above sound, obtaining the fluctuation level of the sound intensity corresponding to the time fluctuation of the same sound, calculating the cumulative fluctuation level by summing the peak differences of the same fluctuation level, and preset. It is characterized in that it is constituted by an evaluation means showing the relationship between the evaluated points of the noise level and the cumulative fluctuation level. The sound to be evaluated in this way is obtained by analyzing the cumulative fluctuation level indicating the temporal fluctuation feeling and the noise absolute value, and the evaluation unit can perform sound quality evaluation according to the magnitude of these two physical quantities.

According to a second aspect of the present invention, in the sound quality evaluation apparatus according to the first aspect, the cumulative fluctuation level calculating means converts the sound into an electric signal, and the signal from the converting means for each frequency band. To calculate the sound pressure time signal, to extract the envelope representing the strength of the sound using the sound pressure time signal, and the first order to process the envelope as a response of the first-order lag system The delay system response means, the fluctuation level calculation means that performs logarithmic display processing of the envelope of the sound intensity processed as the response of the first-order delay system, and the peak difference of the fluctuation levels calculated by the fluctuation level calculation means are totaled. And a fluctuation level accumulating means for calculating a cumulative fluctuation level. As described above, since the cumulative fluctuation level calculating means is composed of the converting means, the envelope converting means, the first-order lag system responding means, the fluctuation level calculating means and the fluctuation level accumulating means, the temporal fluctuation feeling is felt by human hearing. Therefore, the physical quantity can be surely met so that the sound quality can be evaluated.

According to a third aspect of the invention, in the sound quality evaluation apparatus according to the second aspect, the first-order lag system response means performs the first-order lag process of the envelope of the sound intensity when the envelope amplitude increases. It is characterized in that the processing is performed with a constant such that the time constant at the time of the decrease of the amplitude is larger than the time constant. In this way, since the first-order lag system processing is processed with a constant whose time constant when the envelope amplitude is decreasing is larger than the time constant when the envelope amplitude is increasing, it is possible to contribute to the calculation of the cumulative fluctuation level suitable for human hearing.

According to a fourth aspect of the present invention, in the sound quality evaluation apparatus according to the first aspect, the noise evaluation is performed based on the overall value of the cumulative fluctuation level and the overall value of the noise level. In this way, since the noise evaluation is performed based on the overall value of the cumulative variation level and the overall value of the noise level, the sound quality can be appropriately physicalized and the sound quality can be evaluated.

According to a fifth aspect of the present invention, a sound to be evaluated is sampled and converted into an electric signal, a noise level is calculated from the electric signal, and a time variation of the sound is measured from the electric signal.
It is characterized in that the cumulative difference level is calculated by summing the peak differences of the fluctuation levels of the sound intensity corresponding to the time fluctuation of the same sound, and the sound is evaluated by the noise level and the cumulative fluctuation level. In this way, the noise level is calculated after converting the sound to be evaluated into an electric signal, and the time fluctuation of the sound is measured from the electric signal, and the peak difference of the fluctuation level of the sound intensity corresponding to the time fluctuation of the same sound is calculated. It is possible to calculate the cumulative variation level by summing and perform sound quality evaluation based on the noise level and the cumulative variation level.

According to a sixth aspect of the present invention, in the sound quality evaluation method according to the fifth aspect, the step of calculating the fluctuation level obtains an envelope curve representing the loudness of the sound from the temporal fluctuation of the sound, It is characterized by including a step of calculating an envelope curve of the sound intensity from an envelope curve representing a loudness.
In this way, the envelope representing the loudness of the sound is obtained from the temporal variation of the sound, and the envelope of the sound intensity is calculated from the same envelope, which can contribute to the calculation of the cumulative variation level suitable for human hearing. .

According to a seventh aspect of the present invention, in the sound quality evaluation method according to the sixth aspect, the step of calculating the fluctuation level is performed by first transforming the envelope of the sound intensity from the envelope representing the loudness of the sound. It is characterized in that it has a step of processing as a delay system response. In this way, since the envelope curve of the sound intensity is processed as the response of the first-order lag system from the envelope curve representing the loudness of the sound, it is possible to contribute to the calculation of the cumulative variation level adapted to the human auditory sense.

According to an eighth aspect of the present invention, in the sound quality evaluation method according to the seventh aspect, the step of calculating the fluctuation level is performed by first-order lag processing of the envelope of the sound intensity when the envelope amplitude increases. It is characterized in that the processing is performed with a constant such that the time constant at the time of the decrease of the amplitude is larger than the time constant. In this way, the first-order lag processing of the envelope of the sound intensity is processed by a constant whose time constant when the envelope amplitude is decreasing is larger than the time constant when the envelope amplitude is increasing. It can contribute to the calculation of a suitable cumulative variation level.

[0015]

BEST MODE FOR CARRYING OUT THE INVENTION A sound quality evaluation apparatus M as an embodiment of the present invention will be described below in which an overall value is applied to the evaluation of a diesel engine, for example. The sound quality evaluation apparatus M shown in FIG. 1 includes a cumulative fluctuation level calculation unit 2 including a conversion unit 1, a noise level calculation unit 3, and an evaluation unit 4, and these respective functions are the human hearing functions described above (see FIG. 13). Reference)). The cumulative fluctuation level calculation means 2 is in addition to the conversion means 1,
An envelope conversion unit 11, a first-order lag system response unit 12, a fluctuation level calculation unit 13, and a fluctuation level accumulation unit 14 are provided.

The conversion means 1 includes a microphone 7 for converting a sound wave (sound pressure) of a diesel engine (hereinafter simply referred to as engine 6) as a sound to be evaluated into an analog signal, and an A / D conversion for converting the analog signal into a digital signal. And a container 8. The microphone 7 is provided at a position separated by a distance L to be evaluated with respect to the main body of the engine 6 which is the DUT, and is attached to the head 901 of the dummy 9 here. The engine 6 is operated at low idle, for example, and the noise at that time is sampled by the microphone 7 as a sound to be evaluated, converted into an electric signal, and input to the A / D converter 8 and the noise level calculation means 3, respectively. There is. The noise level calculating means 3 may be configured to separately employ a dedicated microphone (not shown) different from the microphone 7 to detect the noise signal (sound pressure Pa waveform).

The A / D converter 8 digitizes the noise signal from the microphone 7, and the data recording section 5 takes in the digital signal at a predetermined data sampling cycle and stores it. Noise signal (sound pressure Pa
Waveform) is shown in FIG. The noise signal of the data recording unit 5 is input to the envelope conversion unit 11. The envelope converting means 11 includes a frequency discriminating unit 11a that separates the noise signal from the data recording unit 5 into frequency bands, and a sound pressure time signal calculating unit 11b that calculates the sound pressure time signal Pi (Pa) in each frequency band. When,
Envelope P _ei representing the strength of sound using the sound pressure time signal Pi
And an envelope calculation unit 11c for extracting

The process of separating the noise by the frequency discriminating unit 11a for each frequency band is adopted as adapted to the frequency discriminating function of the basilar membrane in human hearing described above. A signal obtained as a result of such frequency discrimination processing, for example, a sound pressure time signal Pi (1 kHz) in a band having a center frequency of 1 kHz is shown in FIG. The sound pressure signal Pi separated for each band by the sound pressure time signal calculation unit 11b
Is the envelope P _ei for the sound pressure signal in the envelope calculation unit 11c.
Used to calculate. As shown in FIG. 4, this process is similar to the process of sequentially connecting the peaks in the sound pressure waveform of the solid line to obtain the sound pressure envelope shown by the broken line.

Next, the first-order lag system response means 12 converts the envelope P _ei representing the loudness (sound pressure) of the sound into the envelope I _ei (W / m ² ) of the sound intensity (energy signal). The energy signal conversion unit 12a and the first-order lag processing unit 12b that processes the envelope of the sound intensity as a response of the first-order lag system in order to simulate the characteristics of human hearing. The sound pressure is obtained by adding a sound pressure fluctuation component to the atmospheric pressure, and the energy signal conversion unit 12a extracts the change component of the sound intensity when the sound pressure is generated. The sound pressure envelope P _ei (Pa) is converted into an envelope of sound intensity, in other words, unit conversion is performed, and is calculated by the formula (1). For example, the envelope I _ei of the sound intensity in the band with the center frequency of 1 kHz.
(W / m ² ) is shown in FIG.

[0020]

[Equation 1]

Here, I _{ei is} the envelope of sound intensity W / m ² ρ is air density kg / m ³ c is the velocity of sound m / sec.

Next, the first-order delay processing section 12b, as shown in FIG. 6, shows the amplitude (mW / mW) of the sound intensity indicated by the broken line.
m ² ) The waveform (input signal) is processed as a first-order lag response and converted into an output signal shown by a solid line, that is, the rise of the amplitude is accelerated and the fall is delayed, so that the amplitude of the sound has a droop. Converted to a time-delayed signal. In order to perform such response processing of the first-order lag system, the impulse response W (t) of the first-order lag system represented by the equation (2) is derived. That is, the envelope I _ei (W / m ² ) of the intensity which is the input signal is converted into the envelope I _ei ′ (mW / m) of the intensity of the sound which has been subjected to the time delay processing which is the output signal.
² ) and shown in FIG. 7.

[0023]

[Equation 2]

[0024] Where W (t): Impulse response of first-order lag system t: time T: Time constant.

In the response processing of the first-order lag system, the input signal (envelope amplitude I _ei ) is x (n), the output signal (time lag signal I _ei ') is y (n), and the previous value is y (n-). 1) and the time constant Tup when rising (x (n) ≧ y (n−1))
(For example, 10 mmsec), when descending (x (n) <y (n
−1)) time constant Tdown (for example, 20 mmsec) and time step width Δt, the response value y (n) of the first-order lag system at the time of rising is expressed by equation (3a), and the response value of the first-order lag system at the time of falling. y (n) is formula (3b) and n is 2, 3, 4
.. can be calculated according to the change over time.

[0026]

[Equation 3]

As described above, by setting the first-order lag system processing so that the time constant Tdown at the time of falling is larger than the time constant Tup at the time of increasing the envelope amplitude, vibration of the basilar membrane by hair cells in human hearing. Of the sound intensity of the sound after the time delay processing shown in FIG. 7 can be simulated, and the envelope of the sound intensity I _ei '(mW / m ² ) shown in FIG. 7 can be simulated.
Is the envelope curve I of the strength of the same sound
_ei '(shown as a time-delayed signal indicated by a solid line) covers a waveform portion having a relatively small peak located therebelow, and in this respect also, a situation similar to human hearing can be reproduced. The fluctuation level calculation means 13 uses the envelope curve I _ei ′ of the sound intensity that has been subjected to the time delay processing, as shown in Expression (4).
(MW / ^{m 2)} was normalized by dividing the effective value ^{_{I RMSi (W / m 2)}} , was treated in the logarithmic values to suit the sensitivity of the human ear, obtains this as fluctuation level _FL i (dB) .

[0028]

[Equation 4]

Here, FL _i : fluctuation level (F1uctuat
ion Revell) dB I _ei ′: envelope of sound intensity in consideration of ear reaction delay W / m ² I _RMSi : effective value W / m ^{2 of} sound intensity.

In this way, an example in which the envelope I _ei '(mW / m ² ) of the sound intensity after the time delay processing is converted into the decibel display as the fluctuation level FL _i (dB) in the band with the center frequency of 1 kHz is shown. , Shown in FIG. In the fluctuation level accumulating means 14, the peak difference dBk between the maximum value hn and the minimum value ln occurring during a certain evaluation time (for example, one engine cycle) at the fluctuation level FL _i calculated by the fluctuation level calculating means 13, that is, (dB1 + dB2 · .. + d
Sum the Bn). Moreover, the weight of the loudness of each band (overall value L _{OA of the} noise level)
And a weight W _{i in} consideration of the human frequency preference pattern shown in FIG. 9 are added to digitize as a cumulative fluctuation level CFL _i . Specifically, it is executed as an arithmetic process of Expression (5). .

[0031]

[Equation 5]

Here, CFL _i : cumulative fluctuation level (Cum
u1ative F1uctuation Leve1) dB i: frequency band number dBk: peak difference in fluctuation level generated during one engine cycle dB N: cumulative number of times (number of peak difference in fluctuation level during one cycle) Li: noise level in each frequency band dB _LOA : overall value of noise level dB Wi: weight considering human frequency preference pattern dB dB _ref : dB reference dB.

Here, in the extraction of the peak difference dBk shown in FIG. 8, the data of the candidates of the maximum value and the minimum value are taken in advance and the filtering process is performed. That is, the fluctuation level FL _i (dB) here repeats the local maximum value hn and the local minimum value ln, but the local maximum value and the local minimum value of a relatively small level are generated between them. Therefore, each time the maximum value after the minimum value is determined, a threshold is set on the x-axis (time axis) and the y-axis (variation level) from the point of the minimum value, and the maximum and minimum values within that threshold range are set. The filtering process of eliminating will be executed. next,
The setting map of the weight W _{i in} consideration of the human frequency preference pattern shown in FIG. 9 is provided as a trapezoidal shape in which the low frequency side and the high frequency side are raised. After this, the obtained cumulative fluctuation level C
The overall value CFL of FL _i is calculated using the equation (6), and is optimized as the physical quantity for noise determination.

[0034]

[Equation 6]

Here, CFL: cumulative variation level overall dBi: frequency band number.

In this way, the cumulative variation level overall CFL representing the temporal variation of noise is calculated as one of the physical quantities for evaluation judgment, and its distribution is shown in FIG.
As shown in 0. On the other hand, the noise level calculation means 3 converts the noise signal (sound pressure Pa waveform) from the microphone 7 into a decibel value dB (A) corrected by using the A characteristic filter and takes in the decibel value NLidB (for each data sampling cycle. A) Overall value NL (dB
(A)) is calculated as a noise level, and this is used as one of the physical quantities for evaluation and determination.

Next, the evaluation means 4 determines the cumulative fluctuation level (d
B) and the sound level (dB (A)) are two physical quantities, and the sound quality is evaluated in 10 levels. Here, the evaluation means 4 adopts the sound quality evaluation processing map m1 shown in FIG. In the case of overall evaluation, the sound quality evaluation processing map m1 has a horizontal axis representing the cumulative fluctuation level CFL (dB) and a vertical axis representing the noise level NL (d).
B (A)) is taken, and ten evaluation lines (1 to 10) for the noise level perceived by humans are appropriately set.
In this case, each evaluation line is set as a straight line having an inclination of 1 passing through the evaluation standard, and divides the evaluation area. Here, No.
It is set to determine that the area where the noise level value is smaller than the evaluation line of 7 (lower left side) is good in sound quality control, and the area where the noise level value is large (upper right side) is poor sound quality control.

Here, as each sound quality evaluation processing map m1, it is also possible to adopt one in which the evaluation level is different for each model of the engine to be tested. A sound quality evaluation method for performing engine noise evaluation by using such a sound quality evaluation device M using overall values will be sequentially described.

First, the sound quality evaluation device M is set on the engine 6 before noise reduction. Next, the engine 6 is operated at low idle, for example, and the noise at that time is collected by the microphone 7 and converted into an electric signal (sound pressure Pa signal). The sound pressure Pa signal from the microphone 7 is input to the A / D converter 8 and the noise level calculation means 3, respectively. The noise level calculation means 3 converts the noise signal (sound pressure Pa waveform) into a decibel value dB (A) and derives the overall value NL (dB (A)) for each data sampling period as a noise level. This noise level NL (dB
(A)) is adopted in the step of evaluation by the evaluation means 4.

On the other hand, the cumulative fluctuation level calculating means 2 measures the temporal fluctuation of the sound and sums the peak differences dBk of the fluctuation levels FL _i (dB) to calculate the cumulative fluctuation level CFL _i . Specifically, first, the sound pressure signal is converted into a digital signal in the A / D converter 8 and is input to the data recording unit 5 for storage processing.

Next, in the envelope conversion means 11, the frequency discriminating unit 11a performs a step of separating the sound pressure signal for each frequency band, and the sound pressure time signal calculating unit 11b separates the sound pressure signal Pi for each band. Is the envelope calculation unit 11
The step of extracting the envelope P _ei representing the loudness of the sound in which the respective peaks are sequentially connected by c (see FIG. 4) is performed. Next, the energy signal converter 12 of the first-order lag system response means 12
In a, a step of calculating the envelope curve I _ei (W / m ² ) of the sound intensity from the envelope curve P _ei representing the loudness of the sound is performed.

Next, in the first-order delay processing section 12b, the rising time constant T is increased with respect to the envelope I _ei (W / m ² ) of the sound intensity.
The Stay up-, using respectively the down time constant Tdown performs processing to have a "Who", performs the step of extracting the intensity of the envelope _{I ei '(mW / m 2} ) of the time delay processed sound. This adapts to the above-mentioned function of delaying the ear response in human hearing. Next, in the fluctuation level calculating means 13, the envelope I _{e i} ′ of the sound intensity after the time delay processing is divided by the effective value I _RMSi (W / m ² ) and the decibel conversion matching the sensitivity of the human ear is performed. Fluctuation level FL _i (dB)
(See FIG. 8).

The fluctuation level accumulating means 14 serves as the accumulative fluctuation level calculating section 14a as indicated by the equation (4).
The peak difference dB of the fluctuation level FL _i generated during the evaluation time
The step of obtaining the cumulative fluctuation level CFL _i by adding various additional values such as the weight W _{i in} consideration of the human frequency preference pattern to the total value of k is performed. Further, as the overall value calculation unit 14b, the cumulative fluctuation level CFL
The cumulative fluctuation level overall value CFL of _i is calculated by the equation (5).
Ask as shown in. After this, the evaluation means 4
Using the sound quality evaluation processing map shown in FIG. 11, evaluation values corresponding to the cumulative fluctuation level CFL (dB) and the noise level NL (dB (A)) are derived from 10 levels.

At this time, for example, in No. Area where the cumulative fluctuation level value and noise level value are larger than the evaluation line of 7 (upper right side)
It is evaluated that the engine 6 needs a sound quality countermeasure if it is in the position before the countermeasure indicated by the symbol. Based on this evaluation result, for example, countermeasures are taken for the engine 6 such as ignition timing adjustment processing and a shield plate is newly installed around the body of the engine 6, resulting in intermittent noise that fluctuates temporally due to diesel noise. It is assumed that the diffusion of noise such as "sound" and "crispy sound" is suppressed.

Then, the sound quality evaluation device M is used again to calculate the cumulative fluctuation level CFL (dB) and the noise level NL (dB (A)) of the engine 6 again under the same conditions as before the sound quality countermeasure, and the evaluation means 4 is calculated. Using the sound quality evaluation processing map, the cumulative fluctuation level CFL (dB) and the noise level NL (dB) after the measures are taken.
The evaluation value after the countermeasure corresponding to (A)) is derived. As a result, the evaluation value after the countermeasure is, for example, No. Assuming that the position after the countermeasure indicated by the symbol ○ in the area (lower left side) where the cumulative fluctuation level value and the noise level value are smaller than the evaluation line of 7 has been reached, the fact that this countermeasure was effective depends on the evaluation value that is a physical quantity. It was confirmed accurately. In addition, in FIG. 12, the fluctuation level (dB) of the engine before the countermeasure is shown by a broken line, and the fluctuation level (dB) after the countermeasure is shown by a solid line. The fluctuation level (dB) decreases after the countermeasure, and the countermeasure is appropriate. I can guess that it was.

By using the sound quality evaluation apparatus M as described above, it is possible to derive an evaluation value suitable for the human auditory mechanism that the human auditory sensation reacts with a time delay. Noise can be accurately confirmed by the evaluation value that quantifies the quality. Although the device under test is described as the engine 6 in the above description, the present invention is applicable not only to various types of engines but also to various types of noise generated by other industrial equipment, and the accumulated fluctuation level (dB) and noise level (dB). (A)) is calculated, an evaluation value which is a physical quantity is calculated by the evaluation means 4 using a sound quality evaluation processing map (not shown) set according to the device under test, and it can be effectively used as a noise countermeasure.

[0047]

As described above, according to the first aspect of the invention, the sound to be evaluated is obtained by analyzing the cumulative fluctuation level and the absolute noise value, which represent a feeling of temporal fluctuation, and is determined according to the magnitude of these two physical quantities. The evaluation means can perform sound quality evaluation.

According to the second aspect of the present invention, since the cumulative fluctuation level calculating means includes the converting means, the envelope converting means, the first-order lag system responding means, the fluctuation level calculating means, and the fluctuation level accumulating means, the temporal fluctuation feeling is generated. The physical quantity can be surely made so as to match the feeling of human hearing, and the sound quality can be evaluated.

According to the third aspect of the present invention, since the first-order lag system processing is performed with a constant having a larger time constant when the envelope amplitude rises than the time constant when the envelope amplitude rises, a cumulative fluctuation level suitable for human hearing is obtained. Can contribute to calculation.

According to the fourth aspect of the present invention, since the noise evaluation is performed based on the overall value of the cumulative variation level and the overall value of the noise level, the sound quality can be appropriately converted into a physical quantity and the sound quality can be evaluated.

According to a fifth aspect of the present invention, the sound level to be evaluated is converted into an electric signal, the noise level is calculated, and the time fluctuation of the sound is measured from the electric signal to determine the sound intensity corresponding to the time fluctuation of the same sound. The cumulative difference level can be calculated by summing the peak differences of the fluctuation levels, and the sound quality can be evaluated based on the noise level and the cumulative fluctuation level.

According to the sixth aspect of the present invention, the envelope representing the loudness of the sound is obtained from the temporal variation of the sound, and the envelope of the sound intensity is calculated from the same envelope. It can contribute to the calculation of the fluctuation level.

According to the seventh aspect of the present invention, the envelope of the sound intensity is processed as the response of the first-order lag system from the envelope representing the loudness of the sound, so that the cumulative fluctuation level suitable for human hearing can be calculated. Can contribute.

According to the eighth aspect of the present invention, the first-order lag processing of the envelope of the sound intensity is processed with a constant such that the time constant when the envelope amplitude rises is larger than the time constant when the envelope amplitude rises. Therefore, it can contribute to the calculation of the cumulative variation level that is suitable for human hearing.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a sound quality evaluation apparatus as an embodiment of the present invention.

FIG. 2 is a sound pressure waveform diagram when the sound quality evaluation apparatus of FIG. 1 measures a sound pressure waveform of an engine.

FIG. 3 is a sound pressure waveform diagram after discrimination when the sound quality evaluation apparatus of FIG. 1 frequency-discriminates the sound pressure waveform of the engine.

FIG. 4 is a sound pressure waveform diagram illustrating a sound pressure envelope calculation function of the sound quality evaluation apparatus of FIG.

5 is a waveform diagram of a sound intensity envelope I _ei calculated by the sound quality evaluation apparatus of FIG. 1 based on the sound pressure waveform of FIG.

6 is a sound intensity waveform diagram for explaining the time delay processing function of the sound pressure waveform included in the sound quality evaluation apparatus of FIG. 1. FIG.

FIG. 7 is a waveform diagram after the time delay processing calculated by the sound quality evaluation apparatus of FIG. 1 based on the sound intensity envelope I _ei of FIG.

8 is a fluctuation level waveform chart in a state in which the sound quality evaluation apparatus of FIG. 1 converts the waveform chart calculated based on the envelope after time delay processing of FIG. 7 into a fluctuation level FL _i .

9 is a characteristic diagram of a weight map in consideration of a human frequency preference pattern adopted in the calculation of the cumulative variation level included in the sound quality evaluation apparatus of FIG.

10 is a diagram showing, for each band, a value of a cumulative fluctuation level calculated by the sound quality evaluation apparatus of FIG. 1 based on the sound pressure envelope of FIG.

11 is a characteristic diagram of an engine sound quality evaluation processing map used by the sound quality evaluation apparatus of FIG. 1;

FIG. 12 is an explanatory diagram of an evaluation example of a fluctuation level waveform diagram calculated based on an envelope curve after time delay processing performed before and after noise countermeasures using the sound quality evaluation apparatus of FIG. 1;

FIG. 13 is a functional explanatory block diagram of each functional unit of human hearing.

[Explanation of symbols]

1 conversion means 2 Data recording section 3 Cumulative fluctuation level calculation means 4 Noise level calculation means 5 Evaluation means 7 microphone 8 A / D converter 11 Envelope conversion means 12 Primary delay system response means 13 Fluctuation level calculation means 14 Fluctuation level accumulation means M engine sound quality evaluation device

Claims (8)

[Claims] 1. A noise level calculation means for measuring a noise level of a sound to be evaluated, a time fluctuation of the sound is measured, a fluctuation level of a sound intensity corresponding to the time fluctuation of the same sound is obtained, and a peak of the fluctuation level. Sound quality evaluation characterized by comprising a cumulative fluctuation level calculating means for calculating the cumulative fluctuation level by summing the differences, and an evaluating means for indicating the relationship between preset noise level and cumulative fluctuation level evaluation points. apparatus. 2. The sound quality evaluation apparatus according to claim 1, wherein the cumulative fluctuation level calculation means separates the sound signal into an electric signal and the electric signal from the conversion means for each frequency band. Envelope conversion means for calculating a sound pressure time signal and extracting an envelope representing the strength of sound using the sound pressure time signal, and first-order lag system response means for processing the envelope as a first-order lag system response And a fluctuation level calculation means that performs logarithmic display processing of the envelope of the sound intensity processed as the response of the first-order lag system, and the peak difference of the fluctuation levels calculated by the fluctuation level calculation means are summed to obtain a cumulative fluctuation level. And a fluctuation level accumulating means for calculating. 3. The sound quality evaluation apparatus according to claim 2, wherein the first-order delay system response means performs the first-order system delay processing of the envelope of the sound intensity more than the time constant when the envelope amplitude increases. A sound quality evaluation device characterized by being processed with a constant such that the time constant when the amplitude falls is large. 4. The sound quality evaluation apparatus according to claim 1, wherein noise evaluation is performed based on the overall value of the cumulative fluctuation level and the overall value of the noise level. 5. A sound to be evaluated is collected and converted into an electric signal,
The noise level is calculated from the electric signal, and the time variation of the sound is measured from the electric signal, and the cumulative variation level is calculated by summing the peak differences of the variation levels of the sound intensity corresponding to the time variation of the same sound, A sound quality evaluation method, characterized in that the sound is evaluated based on the noise level and the cumulative variation level. 6. The sound quality evaluation method according to claim 5, wherein in the step of calculating the fluctuation level, an envelope curve representing the loudness of the sound is obtained from the temporal fluctuation of the sound, and the envelope curve representing the loudness of the sound. From the above, there is a step of calculating an envelope curve of the above-mentioned sound intensity. 7. The sound quality evaluation method according to claim 6, wherein in the step of calculating the fluctuation level, an envelope curve representing the loudness of the sound is processed as a response of a first-order lag system from an envelope curve representing the loudness of the sound. A sound quality evaluation method, comprising: 8. The sound quality evaluation method according to claim 7, wherein in the step of calculating the fluctuation level, the first-order lag processing of the envelope of the sound intensity is performed more than the time constant when the envelope amplitude increases. A sound quality evaluation method, characterized in that the sound quality is evaluated by a constant having a large time constant when the amplitude decreases.