JP4595934B2 - Voice evaluation apparatus and voice evaluation method - Google Patents

Description

  The present invention relates to a technique for evaluating speech.

Various karaoke apparatuses having a song evaluation function for evaluating the skill of a user's song are provided. As this kind of karaoke apparatus, for example, in Patent Document 1, a user's singing position is detected and compared with each music element of the singing voice and each music element of the reference frame of the reference voice corresponding to the singing position. An apparatus for evaluating the pitch, volume, and voice quality of a singing voice by evaluating a song based on the comparison result has been proposed.
JP 2001-117568 A

By the way, in addition to the evaluation of skill by the pitch and volume of the singing voice, the evaluation of the skill of the singing by the listener may be influenced by the clarity of the voice. For example, some listeners feel that the evaluation is higher as the voice is clearer. It is preferable if the user can grasp such clarity of speech.
The present invention has been made under the above-described background, and an object of the present invention is to provide a technology that allows a user to grasp the intelligibility of speech.

The speech evaluation apparatus according to a preferred aspect of the present invention analyzes a speech signal representing speech for each of a plurality of predetermined frequency components and calculates a frequency spectrum, and the level calculation unit calculates the frequency spectrum. From the frequency spectrum, a predetermined level of the nth formant is specified as the first level, and the level of the valley position between the nth formant and the (n + 1) th formant is set as the second level. Level specifying means for specifying, and calculation means for calculating the articulation level of the voice based on the first level and the second level specified by the level specifying means, wherein the first level relative to the second level is calculated. The intelligibility calculation means for calculating the intelligibility so that the intelligibility becomes higher as the ratio of the levels of It is characterized by comprising the intelligibility information output means for outputting the clarity information indicating.

Further, the speech evaluation apparatus according to another preferred aspect of the present invention analyzes a speech signal representing speech for each of a plurality of predetermined frequency components, and calculates a level for each frequency component; When the level for each frequency component calculated by the level calculating means is arranged in descending order of the level , the level of the first order predetermined in the arranged level column is specified as the first level . , a level specifying means for specifying the level of the second order defined lower advance than the first rank as the second level in the aligned level columns of the first to the level specifying means is identified A calculating means for calculating the intelligibility of the voice based on a level and a second level, wherein the intelligibility increases as the ratio of the first level to the second level increases; And clarity calculating means for calculating the intelligibility to so that is characterized by comprising the intelligibility information output means for outputting the clarity information indicating the clarity of the clarity calculation means has calculated.

The aspect mentioned above WHEREIN: You may provide the clarity notification means which alert | reports the clarity shown by the said clarity information based on the clarity information which the said clarity information output means output.
In the aspect described above, the second level calculating means for calculating the level of the audio signal, the third level calculating means for calculating the level of a predetermined frequency band in the audio signal, and the second level calculating means. The level calculated by the second level calculation unit and the third level calculation unit so that the evaluation becomes higher as the ratio of the level calculated by the third level calculation unit to the level calculated by the level calculation unit increases. May further comprise voice quality evaluation means for evaluating the voice quality of the voice and voice quality evaluation information output means for outputting voice quality evaluation information indicating an evaluation result of the voice quality evaluation means.
Moreover, the aspect mentioned above WHEREIN: You may provide the voice quality evaluation result alerting | reporting means which alert | reports the voice quality evaluation result which the said voice quality evaluation information shows based on the voice quality evaluation information which the said voice quality evaluation information output means output.

  Further, in the above-described aspect, the voice quality evaluation information output unit includes the voice quality evaluation correction unit that corrects the evaluation result of the voice quality evaluation unit according to the intelligibility calculated by the intelligibility calculation unit. You may output the evaluation result information which shows the evaluation result corrected by the means.

  According to the present invention, a user can grasp the intelligibility of speech.

Next, the best mode for carrying out the present invention will be described.
<A: Configuration>
FIG. 1 is a block diagram showing a hardware configuration of a karaoke apparatus 1 according to an embodiment of the present invention. In the figure, the control unit 11 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory), and reads and executes a computer program stored in the ROM or the storage unit 12. The respective units of the karaoke apparatus 1 are controlled via the bus BUS. The storage unit 12 is a storage unit for storing a program executed by the control unit 11 and data used at the time of execution, and is, for example, a hard disk device. The display unit 13 includes a liquid crystal panel and the like, and displays various screens such as a menu screen for operating the karaoke device 1 and a karaoke screen in which lyrics telop is superimposed on a background image under the control of the control unit 11. The operation unit 14 outputs a signal corresponding to the operation by the user to the control unit 11. The microphone 15 picks up the sound produced by the user and outputs a sound signal (analog signal) representing the picked up sound. The audio processing unit 16 converts an audio signal (analog signal) output from the microphone 15 into digital data. The audio processing unit 16 converts the digital data into an analog signal and outputs the analog signal to the speaker 17. The speaker 17 is a sound emitting unit that emits sound with an intensity corresponding to the sound signal that is converted from the digital data into an analog signal and output by the sound processing unit 16.

  In this embodiment, the case where the microphone 15 and the speaker 17 are included in the karaoke apparatus 1 will be described. However, the audio processing unit 16 is provided with an input terminal and an output terminal, and the input terminal is connected to the input terminal via an audio cable. An external microphone may be connected. Similarly, an external speaker may be connected to the output terminal via an audio cable. In this embodiment, the case where the audio signal input from the microphone 15 to the audio processing unit 16 and the audio signal output from the audio processing unit 16 to the speaker 17 are analog audio signals will be described. You may make it input / output. In such a case, the audio processing unit 16 does not need to perform A / D conversion or D / A conversion.

  As illustrated, the storage unit 12 includes an accompaniment data storage area 121, a background image data storage area 122, a lyrics data storage area 123, and a model voice data storage area 124. In the accompaniment data storage area 121, accompaniment data constituting an accompaniment musical sound of each music in a data format such as MIDI (Musical Instruments Digital Interface) is stored for each music. The background image data storage area 122 stores background image data representing a background image displayed at the time of karaoke accompaniment. The lyric data storage area 123 stores lyric data representing the lyric of the music displayed as the lyrics telop at the time of karaoke accompaniment. In the model voice data storage area 124, voice data in the WAVE format or the like representing voices (hereinafter referred to as “model voices”) as models of music is stored.

<B: Operation>
Next, the flow of processing performed by the karaoke apparatus 1 will be described with reference to the flowchart shown in FIG. First, a user operates the operation part 14 of the karaoke apparatus 1, and performs operation which selects the music to sing. The operation unit 14 outputs a signal corresponding to the operated content to the control unit 11. The control part 11 selects a music according to the operation signal output from the operation part 14 (step S1).
The control unit 11 causes the background image and lyrics telop of the selected music to be displayed on the display unit 13 and starts karaoke accompaniment (step S2). That is, the control unit 11 reads the accompaniment data from the accompaniment data storage area 121 and supplies the accompaniment data to the audio processing unit 16. The audio processing unit 16 converts the accompaniment data into an analog signal and supplies the analog signal to the speaker 17. The speaker 17 emits an accompaniment sound according to the supplied analog signal. Further, the control unit 11 reads out the lyrics data from the lyrics data storage area 123 and also reads out the background image data from the background image data storage area 122 and causes the display unit 13 to display the lyrics telop and the background image.

  The practitioner sings along with the accompaniment emitted from the speaker 17. At this time, the practitioner's voice is picked up by the microphone 15, converted into a voice signal, and output to the voice processing unit 16 (step S3). The audio processing unit 16 converts an audio signal output from the microphone into digital data (hereinafter simply referred to as “audio signal”).

The control unit 11 analyzes the audio signal in frame units of a predetermined time length (for example, “3 msec”), and detects the audio level and spectrum in frame units (step S4). That is, the control unit 11 detects the level of the audio signal in units of frames, separates the audio signal into a plurality of frequency components, and calculates the level for each frequency component (detects the spectrum). In this embodiment, the control unit 11 detects a spectrum from speech using FFT (Fast Fourier Transform).
FIG. 3 is a diagram illustrating a spectrum detection result. In FIG. 3, the horizontal axis indicates the frequency, and the vertical axis indicates the level. FIG. 3 shows the spectra of two audio signals, that is, the audio signal S1 and the audio signal S2.

Next, the control unit 11 specifies a position where the level change appears as a peak in the detected spectrum, specifies a position where the level change appears as a valley, and determines the level of the specified peak position as a formant. While specifying as the level (first level), the level of the specified valley position is specified as the valley level (second level) (step S5).
Here, the level specifying process shown in step S5 performed by the control unit 11 in this embodiment will be described. First, the control unit 11 arranges the level for each frequency component calculated in step S4 in descending order of the level. FIG. 4 is a diagram showing the contents of the spectrum detection results shown in FIG. 3 sorted in descending order of level. In FIG. 4, the horizontal axis indicates the number of elements, and the vertical axis indicates the level. In this embodiment, in the sorted level column, the level of the frequency component located at ¼ from the head is regarded as the peak level, while the level located at 3/4 is regarded as the valley level. That is, the control unit 11 specifies the level located at ¼ from the head as the formant level in the sorted level column. In addition, the control unit 11 specifies the level that is 3 / 4th from the top in the sorted level column as the valley level. In the example shown in FIG. 4, the level L1 is specified as the formant level, and the level L2 is specified as the valley level. In this way, by specifying the ¼th level with respect to the total number from the top and the ¾th level with respect to the total number, it is possible to specify values close to the formant level and the valley level.

Next, the control unit 11 calculates the speech intelligibility based on the formant level and the valley level (step S6). At this time, the control unit 11 performs a clarity calculation process so that the clarity increases as the ratio of the formant level to the valley level increases. In this embodiment, the control unit 11 calculates the clarity C (dB) from the formant level A and the valley level B using the following equation.
C = 20 * log (A / B) (Formula 1)

  Thus, the higher the ratio of the formant level to the valley level, the higher the clarity. Specifically, for example, in the example shown in FIG. 4, the ratio of the formant level to the valley level is higher in the audio signal S1 than in the audio signal S2, so that the clarity of the audio signal S1 is higher than that of the audio signal S2. Higher than degree.

Next, the control unit 11 evaluates the voice quality. First, the control unit 11 calculates the level of the high frequency band (hereinafter, “high frequency level”) from the audio signal (step S7). In this embodiment, the control unit 11 calculates the level of the frequency band of 1 kHz or higher as the high frequency level. Next, the control unit 11 evaluates the voice quality according to the ratio between the level of all bands calculated in step S4 and the high frequency level calculated in step S7 (step S8). At this time, the control unit 11 performs a voice quality evaluation process so that the higher the ratio of the level calculated in step S7 to the level calculated in step S4, the higher the evaluation. In this embodiment, the control unit 11 calculates a voice quality value D by using the following formula, using the high frequency level E calculated in step S7 and the level F of all bands calculated in step S4.
D = 20 * log (E / F) (Formula 2)
Thus, in this embodiment, the control unit 11 increases the evaluation of voice quality as the number of high frequency components included in the voice increases.

  FIG. 5 is a diagram for explaining an example of voice quality evaluation shown in step S8. FIG. 5 is a diagram illustrating a spectrum of an audio signal. In FIG. 5, the horizontal axis indicates the frequency, and the vertical axis indicates the level. FIG. 5 shows two different audio signal spectra of the audio signal S3 and the audio signal S4. In FIG. 5, the ratio of the high frequency level to the overall level is greater in the audio signal S3 than in the audio signal S4. Therefore, when the voice quality values of the audio signal S3 and the audio signal S4 are respectively calculated, the voice quality of the audio signal S3 is calculated. The value is higher than the voice quality value of the audio signal S4.

  Next, the control unit 11 outputs the intelligibility information indicating the intelligibility calculated in step S6 and the voice quality evaluation information indicating the evaluation result of the voice quality calculated in step S8 to the display unit 13. Based on the intelligibility information supplied from the control unit 11, the display unit 13 notifies the user by displaying the intelligibility indicated by the intelligibility information (step S9).

  FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. As illustrated, the display unit 13 displays a lyrics telop A1, a bar graph A2 indicating the degree of intelligibility, and a bar graph A3 indicating the degree of voice quality. The level of clarity is expressed by the length of the bar line A21 of the bar graph A2. Further, the voice quality evaluation result is represented by the length of the bar line A31 of the bar graph A3. The control unit 11 causes the display unit 13 to display a bar line A21 having a length based on the calculated clarity, and causes the display unit 13 to display a bar line A31 having a length based on the calculated voice quality evaluation result.

  The user can grasp the degree of articulation of his / her voice by visually recognizing the bar graph A2 displayed on the display unit 13. Further, the user can grasp the evaluation result of the voice quality by visually recognizing the bar graph A3 displayed on the display unit 13.

  The control unit 11 determines whether or not to end the process (step S10). This determination may be performed, for example, by determining whether or not the accompaniment sound reproduction of the music has ended, or whether or not an operation for interrupting the accompaniment sound reproduction has been performed by the user. When it is determined that the process is to be ended (step S10; YES), the control unit 11 ends the process as it is, and when it is determined that the process is to be continued (step S10; NO), the control unit 11 returns to the process of step S3. Then, the process of collecting the voice and notifying the intelligibility and voice quality based on the collected voice is repeatedly performed (steps S3 to S9). In this embodiment, the control unit 11 calculates the voice quality value after calculating the clarity. However, the order of processing is not limited to this, and the voice quality value may be calculated first. The order is arbitrary.

  By repeating the processing of step S3 to step S9, the length of the bar line A21 and the length of the bar line A31 that are displayed depend on the intelligibility and voice quality of the user's voice as time passes. fluctuate. That is, in this embodiment, while singing, the clarity and voice quality of the singing voice are notified to the user in real time.

Thus, in this embodiment, when a user sings, the karaoke apparatus 1 notifies the user of clarity and voice quality according to the singing voice. Thereby, the user can grasp the intelligibility and voice quality of his / her singing voice.
Moreover, in this embodiment, the karaoke apparatus 1 alert | reports the evaluation result of voice quality according to a user's song voice. Thereby, the user can grasp what his / her voice quality is. In this embodiment, whether or not the user sang with a good voice (voice quality, articulation, etc.) while the user is singing is immediately displayed on the screen. You can be conscious of singing with a good voice.

<C: Modification>
As mentioned above, although embodiment of this invention was described, this invention is not limited to embodiment mentioned above, It can implement with another various form. An example is shown below.
(1) In the above-described embodiment, the audio signal is analyzed for each frequency component by performing FFT on the audio signal and detecting the spectrum. The method of analyzing the audio signal is not limited to this, and for example, the analysis may be performed using outputs of a plurality of bandpass filters. In this case, for example, the level of a predetermined frequency band may be calculated based on the outputs of a plurality of bandpass filters, and the voice quality may be evaluated according to the ratio between the calculated level of the predetermined frequency band and the overall level. .

(2) In the above-described embodiment, when the control unit 11 detects the spectrum and arranges the spectrum in ascending order as shown in FIG. Although it is configured to specify the appearing position, it is not limited to the ascending order but may be arranged in the descending order. Any mode that can identify the position where the level change appears as a mountain and the position where it appears as a valley is acceptable.
In the above-described embodiment, as shown in FIG. 4, the control unit 11 arranges the levels for each frequency component in descending order of levels, but may arrange them in ascending order.

(3) In the above-described embodiment, a plurality of frequency components are sorted (aligned) in descending order of level, and the formant level and valley level are specified using the sorting result. The method for specifying the formant level and the valley level is not limited to this, and any method may be used as long as the formant level and the valley level can be specified.
As an example, a calculation method using formants is shown below. First, the control unit 11 calculates the spectrum by performing FFT on the audio signal. Next, the control unit 11 extracts formant levels corresponding to the first, second, and third formants from the frequency spectrum obtained as a result of the analysis. The formant is a dominant frequency component on the voice spectrum, and is called first formant, second formant, third formant, fourth formant,. After extracting the formant level, the control unit 11 specifies the level at the valley position between the first formant and the second formant as the valley level.

In the above example, the formant level of the first formant and the level at the valley position between the first formant and the second formant are specified as the valley level. However, the present invention is not limited to this, and the nth formant (n is A natural number) formant level and a level at the valley position between the nth formant and the (n + 1) th formant may be specified, and the value of n is not limited to 1. Specifically, for example, the formant level of the second formant and the level of the valley position between the second formant and the third formant may be specified.
Thus, the position of the specified mountain may be the first appearing mountain (first formant), may be the second appearing mountain (second formant), and appears nth. It may be a mountain (nth formant: n is a natural number). Similarly, the position of the valley to be identified may be the position of the valley that appears at the nth (n is a natural number).

(4) In the above-described embodiment, the intelligibility is calculated using (Equation 1), but the intelligibility calculation method is not limited to this. For example, the spectrum is calculated, the position where the level change appears as a peak in the calculated spectrum is specified, the position where the level change appears as a valley is specified, and the difference value between the peak level and the valley level is calculated. The intelligibility may be calculated so that the higher the difference value is, the higher the intelligibility is. In short, any calculation method may be used as long as the ratio of the mountain level to the valley level is larger so that the clarity becomes higher.

  In the above-described embodiment, the voice quality value is calculated using (Equation 2). However, the calculation method of the voice quality value is not limited to this, and in short, the voice quality evaluation method is the level of the entire frequency band. As long as the ratio of the level of the high frequency band with respect to increases, the evaluation becomes higher, so long as the voice quality is evaluated.

(5) In the above-described embodiment, the degree of clarity is notified by displaying a bar graph indicating each degree. The notification mode is not limited to this. For example, as shown in FIG. 7, the clarity may be notified by changing the background image displayed on the display unit 13 according to the degree of clarity. In this case, for example, as the intelligibility is high, the number of images to be displayed is increased as shown in FIG. 7B. On the other hand, as the intelligibility is low, as shown in FIG. You may alert | report by reducing the number of the images to perform.
In addition, for example, the clarity may be notified by changing the background color according to the clarity. In this case, for example, the higher the intelligibility, the brighter the background color as shown in FIG. 8 (b), while the lower the intelligibility, the darker the background color as shown in FIG. 8 (a). May be. Further, for example, as shown in FIG. 9, the meter A5 indicating the degree of intelligibility may be displayed on the display unit 13 for notification. In FIG. 9, the intelligibility is notified by the position of the needle A51 in the meter A5. In addition, the manner of notifying the intelligibility is not limited to display, and for example, the intelligibility may be informed by emitting a voice message representing intelligibility. In short, any mode may be used as long as the degree of clarity calculated by the control unit 11 is notified.

  In addition, the voice quality evaluation notification mode is the same as the above-described clarity notification mode, not only displaying a bar graph, but also reporting the voice quality evaluation result by changing the background color or emitting a voice message. Any form may be used as long as the result of voice quality evaluation by the control unit 11 is notified.

  Further, in the above-described embodiment, the control unit 11 outputs the clarity information indicating the clarity and the voice quality evaluation information indicating the evaluation result of the voice quality to the display unit 13. The output destination of the intelligibility information and the voice quality evaluation result information is not limited to the display unit 13 and may be output by transmitting the intelligibility information or the voice quality evaluation result information through a communication network, or a hard disk or the like. It is also possible to output to the storage means and store it in the storage means. In short, the control unit 11 may output clarity information and voice quality evaluation result information.

(6) In the above-described embodiment, the clarity and the voice quality are calculated for the user's singing voice collected by the microphone 15. In addition to this, for the model voice data stored in the model voice data storage area 124, the clarity and voice quality may be calculated and displayed on the display unit 13 to notify the user. An example of the screen displayed in this case is shown in FIGS. In FIG. 10, the length of the bar line A21 of the bar graph A2 represents the level of clarity of the user, while the length of the bar line A41 of the bar graph A4 represents the level of clarity of the model voice. In FIG. 11, the clarity of the user is represented by the position of the needle A61 in the meter A6, while the clarity of the model voice is represented by the position of the needle A62. In this case, since the user's evaluation result and the model evaluation result are notified together, the user can compare his / her singing with the model singing.

(7) In the above-described embodiment, the user's voice is evaluated, but the voice to be evaluated may be only the model voice. In this case, only the evaluation result of the model voice may be displayed. Further, for example, a plurality of users may sing at the same time, evaluate each user's voice, and display the evaluation results side by side. Thus, one voice may be evaluated and notified, or a plurality of voices may be evaluated in parallel. The voice to be evaluated may be voice collected by the microphone 15 or voice data stored in advance in the storage unit 12. In addition, for example, a communication unit may be provided in the karaoke device 1 and voice data received via the communication unit may be evaluated. The voice to be evaluated is any voice as long as it represents voice. There may be.

(8) In the above-described embodiment, the frequency components are sorted in descending order of level, and the level of the frequency component at the position 1/4 and the level of the frequency component at the position 3/4 from the top in the sorted column. Identified. The position to be identified is not limited to this. For example, the level of the 1 / 5th frequency component and the level of the 4 / 5th frequency component may be identified. In short, any value may be used as long as it calculates a value close to the difference value between the peak and valley of the level, and the level of the frequency component located at the predetermined first position is determined in advance. The difference value from the level of the frequency component located at the second position may be calculated.

(9) In the above-described embodiment, the bar graph indicating the intelligibility and the bar graph indicating the evaluation result of the voice quality are displayed together. However, the present invention is not limited thereto, and only the intelligibility may be notified. It is good also as a structure which alert | reports only an evaluation result.

(10) In the above-described embodiment, the degree of clarity may be related to the evaluation of voice quality. In this case, the control unit 11 may correct the evaluation result of the voice quality evaluation according to the calculated clarity, and output evaluation result information indicating the corrected evaluation result to the display unit 13. Specifically, for example, the control unit 11 may correct the evaluation result of the voice quality evaluation so that the evaluation result becomes lower as the intelligibility becomes lower. For example, when the calculated clarity is equal to or less than a predetermined threshold, the voice quality evaluation result may be set to a predetermined low value (or not evaluated).
For example, suppose that the user sang singing in a serious manner, such as pronouncing a consonant. In the conventional apparatus, the voice quality may be highly evaluated even when the user sings in a serious manner as described above, but in this embodiment, the intelligibility is reflected in the evaluation by reflecting the intelligibility in the evaluation. Singing can be evaluated low, and serious singing can be prevented from being evaluated highly.

(11) In the above-described embodiment, the voice quality evaluation result may be converted into calorie consumption and notified. In this case, the control unit 11 may calculate the consumed calorie H from the voice quality value G using the following equation. In the following formula, k represents a coefficient.
H = k × G (Formula 3)

(12) In the above-described embodiment, the voice quality is evaluated based on the ratio between the level of the audio signal and the level of the frequency band of 1 kHz or higher in the audio signal. The aspect of the evaluation of the voice quality is not limited to this. For example, the voice quality may be evaluated according to the ratio between the level of the voice signal and the level of the frequency band of 2 kHz or higher in the voice signal. Voice quality may be evaluated according to the ratio with the level of the frequency band. In short, the voice quality is evaluated according to the ratio between the overall level of the audio signal and the level of the frequency band predetermined in the audio signal. Any mode may be used.

(13) In the above-described embodiment, the karaoke apparatus 1 is applied as the voice evaluation apparatus according to the present invention. However, the apparatus applied as the voice evaluation apparatus is not limited to the karaoke apparatus, and for example, a server device, a personal computer, or mobile communication. Various devices such as a terminal are applicable as the voice evaluation device according to the present invention.

(14) The program realized by the control unit 11 of the karaoke apparatus 1 described above is provided in a state of being recorded on a recording medium such as a magnetic tape, a magnetic disk, a flexible disk, an optical recording medium, a magneto-optical recording medium, a RAM, or a ROM. Can do. It is also possible to download to the karaoke apparatus 1 via a network such as the Internet.

1 is a block diagram illustrating an example of a configuration of a karaoke apparatus 1. FIG. It is a flowchart which shows the flow of a process of the karaoke apparatus. It is a figure which shows an example of the spectrum of an audio | voice. It is a figure which shows the result of having sorted the level for every frequency component of an audio | voice. It is a figure which shows an example of the spectrum of a sound. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG. 6 is a diagram illustrating an example of a screen displayed on the display unit 13. FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Karaoke apparatus, 11 ... Control part, 12 ... Memory | storage part, 13 ... Display part, 14 ... Operation part, 15 ... Microphone, 16 ... Sound processing part, 17 ... Speaker, 121 ... Accompaniment data storage area, 122 ... Background image Data storage area, 123 ... Lyric data storage area, 124 ... Model voice data storage area.

Claims (8)

Level calculation means for analyzing a voice signal representing voice for each of a plurality of predetermined frequency components and calculating a level for each frequency component;
When the level for each frequency component calculated by the level calculation means is arranged in descending order of the level, the first rank level determined in advance in the arranged level column is specified as the first level. And a level specifying means for specifying, as the second level, a predetermined second rank level lower than the first rank in the aligned level column;
The calculation means for calculating the intelligibility of the sound based on the first level and the second level specified by the level specifying means, wherein the ratio of the first level to the second level increases. Clarity calculating means for calculating the intelligibility so as to increase the intelligibility;
A speech evaluation apparatus comprising: clarity information output means for outputting clarity information indicating the clarity calculated by the clarity calculation means. Level calculation means for analyzing a voice signal representing voice for each of a plurality of predetermined frequency components and calculating a frequency spectrum;
The level of one predetermined nth formant is specified as the first level from the frequency spectrum calculated by the level calculating means, and the valley position between the nth formant and the (n + 1) th formant Level specifying means for specifying the level of the second level as a second level;
The calculation means for calculating the intelligibility of the sound based on the first level and the second level specified by the level specifying means, wherein the ratio of the first level to the second level increases. Clarity calculating means for calculating the intelligibility so as to increase the intelligibility;
A speech evaluation apparatus comprising: clarity information output means for outputting clarity information indicating the clarity calculated by the clarity calculation means. The speech evaluation apparatus according to claim 1, further comprising: a clarity notifying unit that notifies the clarity indicated by the clarity information based on the clarity information output by the clarity information output unit. Second level calculating means for calculating the level of the audio signal;
Third level calculation means for calculating a level of a predetermined frequency band in the audio signal;
The level calculated by the second level calculation unit and the third level so that the evaluation increases as the ratio of the level calculated by the third level calculation unit to the level calculated by the second level calculation unit increases. Voice quality evaluation means for evaluating the voice quality of the voice based on the level calculated by the level calculation means;
The voice evaluation apparatus according to any one of claims 1 to 3, further comprising voice quality evaluation information output means for outputting voice quality evaluation information indicating an evaluation result of the voice quality evaluation means. The voice evaluation apparatus according to claim 4, further comprising voice quality evaluation result notifying means for notifying a voice quality evaluation result indicated by the voice quality evaluation information based on the voice quality evaluation information output by the voice quality evaluation information output means. Voice quality evaluation correction means for correcting the evaluation result of the voice quality evaluation means according to the clarity calculated by the clarity calculation means;
The voice evaluation apparatus according to claim 5, wherein the voice quality evaluation information output means outputs evaluation result information indicating the evaluation result corrected by the voice quality evaluation correction means. A voice evaluation method of a voice evaluation apparatus provided with a control means,
A first step in which the control means analyzes a sound signal representing sound for each of a plurality of predetermined frequency components, and calculates a level for each frequency component;
When the control means arranges the calculated level for each frequency component in descending order of the level, the control unit specifies a first rank level predetermined in the arranged level column as the first level. And a second step of identifying, as the second level, a predetermined second rank level lower than the first rank in the aligned level column;
The control means calculates speech intelligibility based on the first level and the second level so that the intelligibility increases as the ratio of the first level to the second level increases. A third step of
The control means comprises a fourth step of outputting intelligibility information indicating the calculated intelligibility. A voice evaluation method of a voice evaluation apparatus provided with a control means,
A first step in which the control means analyzes a sound signal representing sound for each of a plurality of predetermined frequency components, and calculates a frequency spectrum;
Position of the valley between the said control means, from the calculated frequency spectrum, as well as identify the level of one of the n formant which is determined in advance as a first level, and the n-th formant and the (n + 1) formant A second step of identifying the level of the second as a second level;
The control means calculates speech intelligibility based on the first level and the second level so that the intelligibility increases as the ratio of the first level to the second level increases. A third step of
The control means comprises a fourth step of outputting intelligibility information indicating the calculated intelligibility.

Images