KR100925828B1 - Method of expressing the quality of the sound in vehicle as the quantitive equation and device thereof - Google Patents

Abstract

The present invention relates to a method and apparatus for quantitative derivation of sound quality of a vehicle sound. More particularly, the sound quality of a sound output through a speaker of a vehicle is quantitatively derived using various physical variables and is felt by a passenger. The sound quality of the vehicle can be predicted to optimize the sound of the vehicle.

Deep Bass, Bass, Treble, Sum of Absolute Deviation (SAD), Dynamic, Stage, Ambience, Clarity, Orthogonal Array, Regression Equation, Regression Model

Description

Method of expressing the quality of the sound in vehicle as the quantitive equation and device approximately}

The present invention relates to a method and apparatus for quantitative derivation of sound quality. More particularly, the present invention relates to a method for quantitatively deriving sound quality of a sound output through an in-vehicle speaker.

Recently, the interest in the technology for the audio device provided in the vehicle is increasing further. This is the reason that the stability of the vehicle has been the main criteria for vehicle selection in the past, but in recent years, the various conveniences provided to the vehicle occupant along with the stability of the vehicle have become the criteria for vehicle selection.

More specifically, technologies for optimizing various sounds transmitted from the audio device provided in the vehicle to the occupant are further developed. However, there is a limit in expressing quantitatively because the subjective aspect of the occupant is emphasized about the sound quality of the sound transmitted to the occupant through the speaker of the vehicle.

In particular, since physical variables that can represent sound (hereinafter, referred to as 'acoustic variables', for example, bass and treble, etc.) are variously related to one acoustic variable representing sound, In order to evaluate the sound quality, a method of deciding the best point by fixing the remaining sound variables and adjusting only the sound variables repeatedly was studied.

However, this method does not objectively express the sound quality of the sound in the vehicle and has a problem of ignoring the influence of the remaining acoustic variables.

In another prior art, a single expert made detailed evaluation of the sound quality of each acoustic variable and proposed a method of expressing their weighted average as an overall average, but this should be expressed only by the opinion of one expert. There is a problem.

In order to solve the above problems, the present invention outputs various sound samples in sequence with respect to the sound output through the speaker of the vehicle and collects the evaluations from a plurality of evaluators, and then uses the results to determine the sound quality of the vehicle sound. We will prepare objective indicators and derive quantitative regression equations.

A quantitative derivation method for sound quality of a vehicle sound according to the present invention comprises the steps of: extracting a plurality of sample sound sources with a combination of acoustic variables which are physical variables representing the characteristics of the sound; Outputting a sound corresponding to each sample sound source from the plurality of sample sound sources; Receiving a detailed evaluation score for each acoustic variable with respect to the output sound; And

And deriving a relational expression such as an evaluation score for each received acoustic variable and a physical value of the acoustic variable.

The acoustic variables include a deep bass representing the lowest sound of the sound, a bass representing the bass of the sound, a treble representing the high sound of the sound, and a sum of absolute deviation from the reference sound source. ), A stage representing a stage effect, an ambience representing a space in which sound sounds, a dynamic representing a dynamicity and a degree of distortion, and a clarity representing a clarity of the sound.

Further, the relational expression may be expressed as a first or second function of each detailed evaluation score for the physical value of each acoustic variable, and the detailed evaluation score may be an integer value of -3 to 3.

A quantitative derivation apparatus for sound quality of a vehicle sound according to the present invention includes a data processor, a speaker, and an evaluation score input unit, and the characteristics of each component are as follows. The data processor extracts a sample sound source and transmits the output signal to the speaker, and the physical value and the physical value of the acoustic variable based on the detailed evaluation score for each acoustic variable and the total evaluation score for the sound transmitted from the evaluation score input unit. Deriving a detailed evaluation model indicating the relationship between the detailed evaluation score, a total evaluation model showing the relationship between the detailed evaluation score and the total evaluation score, and the speaker outputs the sound to the outside according to the output signal transmitted from the data processor The evaluation score input unit may transmit the detailed evaluation score of the acoustic variable of each sample sound input by the evaluator and the total evaluation score of each sample sound source to the data processor.

According to the present invention, it is possible to express the sound quality of the vehicle sound more objectively and quantitatively to increase the control reliability of the in-vehicle sound and to predict the optimum sound for the vehicle, thereby increasing the convenience of the vehicle.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a view showing a quantitative derivation method for the sound quality of a vehicle sound according to the present invention and the sequence of the device. More specifically, Figure 1 shows a quantitative derivation procedure for the sound quality of the vehicle sound according to the present invention.

Not only are the types and genres of the sounds (especially music) reproduced through the audio device of the vehicle, but also the sound quality of the sound is evaluated differently according to the subjective viewpoint of the listener who listens to them. That is, there is a problem that an objective basis is not supported in adopting various physical variables (hereinafter, referred to as "acoustic variables") representing the characteristics of sound as subjective evaluation data about sound quality.

Referring to FIG. 1, the present invention is intended to be expressed as an objective equation for evaluating sound quality of a sound using physical values of sound variables.

First, a sound source for the sound to be output through the speaker of the vehicle can be produced (S100). At this time, a total of eight acoustic variables, a sample sound source consisting of a combination of them is produced.

Here, the eight acoustic variables include a deep bass (DB) hereinafter referred to as "DB", a bass representing bass (Bass: B, hereinafter referred to as "B"), and a sound. Treble (T, hereinafter referred to as "T"), a deviation of the reference sound source (SAD, hereinafter referred to as "SAD"), and a stage effect (Stage: S) (Hereinafter referred to as "S"), ambience (A, hereinafter referred to as "A") representing the space in which the sound sounds, and dynamic (D, hereinafter referred to as "D") indicating the dynamics and distortion of the sound. And Clarity (C, hereinafter referred to as "C") indicating the clarity of sound.

 Subsequently, a detailed evaluation model showing a correlation between the acoustic variable and the individual detailed evaluation scores through the evaluation scores of the plurality of evaluators for the sound composed of the sample sound source is derived (S110).

Thereafter, a total evaluation model showing a correlation between the detailed evaluation score of each acoustic variable and the total evaluation score of the sound for the sound is derived (S120).

2 is a diagram showing an orthogonal array of sample sound sources according to the present invention. Referring to FIG. 2, since the number of sounds that can be combined using eight acoustic variables increases exponentially, it is necessary to extract a sample sound source composed of some of them. As a result, 16 sample sound sources of various sound combinations are selected and expressed as orthogonal array method as shown in FIG. 2. In the present invention, although the sample sound source extracts 16 sample sound sources for each genre, the number of sample sound sources can be changed.

In addition, by separating the genre of sound to separate the sound source (POP) and classic (CLASSIC) (sample). Here, each variable has a number of 0 or 1, and thus a predetermined physical value is output through the speaker. In more detail, it is shown in Table 1.

Acoustic variables 0 One DB +5 [dB] +10 [dB] B -6 [dB] +6 {dB] T -6 {dB] +6 [dB] SAD 0 [dB] +20 {dB] S 0.6 msec 1.8 msec A 0% 100% D 0% 5% C 0% 5%

Referring to Table 1, each acoustic variable has a different physical value, but is represented by 0 or 1 in the form of the same number in FIG. 2. For example, 0 of DB is substantially +5 [dB], and 1 is +10 [dB]. As in the example of DB, the physical values corresponding to 0 and 1 of each acoustic variable shown in the orthogonal array can be seen.

The sample sound source configured as described above is output through the speaker of the vehicle and the evaluation score is expressed as the physical value of the acoustic variable. This will be described in detail with reference to FIG. 3.

3 is a diagram showing an acoustic evaluation table according to the present invention. More specifically, FIG. 3 shows an evaluation table of 16 sample sound sources for each genre output according to the rule of FIG. 2.

Referring to FIG. 3, the evaluation table evaluates the detailed evaluation score of each acoustic variable in seven levels from -3 to +3 points for the sample sound source, and the overall total evaluation score is -3 to +3 points. It is configured to be evaluated in stages. That is, by outputting 16 different sample sound sources for the same music according to Figure 2, the detailed evaluation score for each acoustic variable and the total evaluation score for the corresponding sound source can be evaluated. Evaluation of this is evaluated by a plurality of evaluators. Of course, the above detailed evaluation score can be input by the evaluator's evaluation apparatus and

Here, the physical value of each acoustic variable is represented by x, each detailed evaluation score is y, and the total evaluation score is z. It is an object of the present invention to quantitatively derive the relationship of x, y and z.

4 is a chart showing the relationship between the physical value of each variable and the qualitative evaluation value for the sound output according to the present invention. Referring to FIG. 4, the relationship between the physical value and the evaluation score for each acoustic variable is summarized in one diagram. That is, by standardizing the detailed evaluation scores evaluated by each evaluator, it is possible to express a correlation between the physical value of each acoustic variable and the evaluation score.

For example, in the case of DB, the standard evaluation score corresponding to the physical value +5 [dB] is -0.5417 and the standard evaluation score corresponding to the physical value +10 [dB] is -0.556. Similarly, in the case of SAD, the standard evaluation score corresponding to the physical value 0 [dB] is -0.6528 and the standard evaluation score corresponding to the physical value +20 [dB] is -1.1806.

FIG. 5A is a diagram illustrating a relationship between a physical value and an evaluation score of each variable for sound using the results of FIG. 4. 5B is a graph showing the relationship between the physical value of each acoustic variable and the evaluation score when the genre is pop. 5C is a graph showing the relationship between the physical value and the evaluation score of each acoustic variable when the genre is classic.

 Referring to FIGS. 5A, 5B, and 5C, a relationship between a physical value of each acoustic variable and a standardized detailed evaluation score may be seen. For example, if the genre is Pop and the acoustic variable is DB, the relation is approximated by a quadratic equation of sy = -0.01 (x-8) ^ 2. It passes through (+5, -.05417) and (+10, -0.556) and is configured to have an inflection point between x = + 5 and x = + 10. In this way, the relationship between the physical value for each acoustic variable and the standard detailed evaluation score corresponding to the physical value can be derived.

Of course, such a relationship may be expressed as a linear function or a quadratic function according to the policy. In the present invention, DB, B, T, and S are quadratic functions, and in the case of SAD, A, D, and C, Expressed as a linear function

Of course, if the genre is Classic, it is obvious that the relation for each acoustic variable is expressed differently. As described above, the change in the feeling of sound according to the change of each acoustic variable is quantified through the relation between the physical value of the corresponding acoustic variable and the evaluation score.

6 is a histogram showing the relationship between the detailed evaluation score and the total evaluation score of each acoustic variable for the sound. Referring to FIG. 6, the detailed evaluation score and the total evaluation score and the relationship are calculated through linear regression analysis.

More specifically, the detailed evaluation scores of the acoustic variables (DB, B, T, S, SAD, A, D, and C) and the estimated total evaluation scores for each output sound are stored in a predetermined number or more. Deduce the relationship between the total score and the score.

For example, if the distribution of evaluation scores for the sound output based on various sample sources is listed below, the detailed and total evaluation scores of DB, B, T, S, SAD, A, D, and C are listed as follows:

specimen DB B T S SAD A D C Total Score One -2 One -One -One 2 2 3 -One 2 2 -One 2 -One -3 -3 2 One One -One .. .. .. .. .. .. .. .. .. .. n 3 -One One 2 -3 -2 One 3 3

DB correlation (-2 points for 2 points, -1 points for -1 points, 3 points for 3 points) for the total evaluation score, and B correlation (1 point for 2 points, -1 for 2 points) By combining information such as 2 points when the point is -1 point when the point is 3) and regression analysis, the total scores for each detailed score can be calculated.

The correlation between the detailed evaluation score and the total evaluation score calculated through an embodiment of the present invention is the same as that of Equation 1 (when the genre is 'pop') and Equation 2 (when the genre is 'classic').

As in Equation 1 and Equation 2, the detailed evaluation score and the total evaluation score for the sound have separate correlations according to each genre. Here, Ztotal represents a total evaluation score, y1 is a DB, y2 is B, y3 is T, y4 is S, y5 is SAD, y6 is A, y7 is D, and y8 is a variable representing a detailed evaluation score of C.

7 is a conceptual diagram of a quantitative derivation apparatus for sound quality of vehicle sound according to the present invention. Referring to FIG. 7, the apparatus for quantitatively deriving sound quality of a vehicle sound may include a data processor 10, a speaker 20, and an evaluation score input unit 30. Since the detailed description according to the present invention has been made, the description of the apparatus will be simplified.

The data processor 10 extracts a sample sound source and transmits an output signal thereof to the speaker 20, and based on the detailed evaluation score for each acoustic variable transmitted from the evaluation score input unit 30 and the total evaluation score for the sound. Derived detailed evaluation model showing the relationship between the physical value of the acoustic variable and the detailed evaluation score, and a total evaluation model showing the relationship between the detailed evaluation score and the total evaluation score.

The speaker outputs a corresponding sound to the outside according to the output signal transmitted from the data processor 20.

The evaluation score input unit 30 transmits the detailed evaluation score of the acoustic variables of each sample sound input by the evaluator and the total evaluation score of each sample sound source to the data processor 10.

1 is a view showing a quantitative derivation method for the sound quality of a vehicle sound according to the present invention and the sequence of the device.

2 is a diagram showing an orthogonal array of sample sound sources according to the present invention.

3 is a diagram showing an acoustic evaluation table according to the present invention.

4 is a chart showing the relationship between the physical value of each variable and the qualitative evaluation value for the sound output according to the present invention.

FIG. 5A is a diagram illustrating a relationship between a physical value and an evaluation value of each variable for sound using the result of FIG. 4.

5B is a graph showing the relationship between the physical value of each acoustic variable and the evaluation score when the genre is pop.

5C is a graph illustrating a relationship between a physical value and an evaluation score of each acoustic variable when the genre is classic.

6 is a histogram showing the relationship between the detailed evaluation score and the total evaluation score of each acoustic variable for the sound.

7 is a conceptual diagram of a quantitative derivation apparatus for sound quality of vehicle sound according to the present invention.

Claims (6)

Extracting a plurality of sample sound sources in which physical values of sound variables, which are variables representing sound characteristics, are set different from each other; Outputting sound for the plurality of sample sound sources; Receiving a detailed evaluation score which is an evaluation score for each acoustic variable for each sample sound source for the output sound; And And collecting a detailed evaluation score for each sample sound source to derive a relationship between a physical value of the acoustic variable set in the sample sound source and the detailed evaluation score. The method of claim 1, The acoustic variables include a deep bass representing the lowest sound of the sound, a bass representing the bass of the sound, a treble representing the high sound of the sound, and a sum of absolute deviation from the reference sound source. ), A stage that represents a stage effect, an ambience that represents a space where sound sounds, a dynamic that represents the dynamics and distortion of the sound, and a clarity that represents the clarity of the sound. Quantitative derivation method for sound quality of vehicle sound The method of claim 1, And when a plurality of detailed evaluation scores are received for a specific physical value of the acoustic variable, a standard detailed evaluation score, which is an average value of the plurality of detailed evaluation scores, is used. The method of claim 1, The detailed evaluation score is a quantitative derivation method for the sound quality of the vehicle sound, characterized in that consisting of an integer value of -3 to 3. The method according to claim 1, wherein Receiving a total evaluation score for each sample sound source for the output sound; Collecting correlation information of the acoustic variable with respect to the total evaluation score for each sample sound source and regressing it; And And deriving a relationship between a detailed evaluation score and the total evaluation score for each acoustic variable of the output sound. In the quantitative derivation device for the sound quality of the vehicle sound comprising a data processor 10, a speaker 20 and the evaluation score input unit 30, The data processor 10 extracts a sample sound source and transmits an output signal thereof to the speaker 20, and calculates the detailed evaluation score for each acoustic variable and the total evaluation score for the sound from the evaluation score input unit 30. On the basis of this, a detailed evaluation model representing the relationship between the physical value of the acoustic variable and the detailed evaluation score is derived, a total evaluation model representing the relationship between the detailed evaluation score and the total evaluation score is derived, and the speaker 20 is a data processor. According to the output signal transmitted from the 20 outputs the sound to the outside, the evaluation score input unit 30 is a detailed evaluation score for the acoustic variable of each sample sound input by the evaluator and the total evaluation score of each sample sound source Quantitative derivation apparatus for the sound quality of the vehicle sound, characterized in that for transmitting to the data processing unit (10).

Images