JP5428459B2 - Singing evaluation device - Google Patents

Description

  The present invention relates to a technique for evaluating a song.

  Various methods for scoring a singer's singing skill in a karaoke apparatus and methods for using the scoring results for various purposes have been proposed. For example, in Patent Document 1, the singing data with the highest scoring in the same section is extracted for each section from the singing data to which scoring results are given for each section posted by a plurality of singers, and they are connected. Thus, a technique for generating exemplary singing data is disclosed.

Japanese Patent Laying-Open No. 2005-241848

By the way, when scoring the skill of a singer's singing, the scoring reference data serving as the reference is data indicating the pitch or beat length indicating the melody of the music. Therefore, if a singer sings according to the melody of the music, the scoring result will get a high score, but the singing technique that sings by intentionally changing the length of the notes, for example, “for” and “shakuri” etc. When singing, the melody of the music does not match the singing, so the scoring result is rather low.
The present invention is made under the background described above, and when evaluating a singer's singing, in addition to the evaluation standard called melody of music, evaluation is performed using an evaluation standard related to the singing skill of the singer. Objective.

The present invention represents the first data representing the melody of the music, the second data representing the singing voice obtained the evaluation result within the predetermined rank in the song of the music , and the evaluation result of the singing by the singing voice. The storage means for storing the third data, the acquisition means for acquiring singing data representing the singing voice to be evaluated, the singing data and the first data are compared, and the singing data and the second data are compared. A comparison means for comparing the second data obtained with the highest evaluation result among the data, a comparison result between the singing data and the first data, and among the singing data and the second data Based on the comparison result with the second data obtained the highest evaluation result, the singing by the singing voice represented by the singing data is evaluated, and an evaluation means for outputting the evaluation result is provided , and is output by the evaluation means. Evaluation result When the third data stored in the storage means is exceeded, the storage means replaces the third data and the second data corresponding to the third data with the evaluation result and the evaluation result. The singing evaluation apparatus characterized by memorizing the singing data corresponding to is provided.

As for the evaluation means, the higher the third data stored in the storage means, the higher the comparison result between the song data and the second data than the comparison result between the song data and the first data. Evaluation may be made with a high specific gravity.

Sound emission means for emitting sound according to the second data stored in the storage means, and operation for third data which is an evaluation result of the second data emitted from the sound emission means You may make it have an operation means to receive.

  Further, the present invention provides a computer with storage means for storing first data representing the melody of music and second data representing the singing voice of the music, and singing data representing the singing voice to be evaluated. The acquisition means for acquiring, the singing data and the first data are compared, the comparing means for comparing the singing data and the second data, and the comparison result between the singing data and the first data And a program for evaluating the singing by the singing voice represented by the singing data based on the comparison result between the singing data and the second data, and realizing the evaluation means for outputting the evaluation result. .

  According to the present invention, when evaluating a singer's singing, the evaluation can be performed using an evaluation criterion related to the singing skill of the singer in addition to the evaluation criterion called melody of music.

1 is a block diagram illustrating a configuration of a karaoke apparatus 1. FIG. It is a figure which shows the structure of the music data 2. FIG. It is a figure which shows the structure of scoring reference data. It is a figure which shows an example of the content of the scoring specific gravity table. It is a flowchart which shows the flow of a scoring process. 2 is a block diagram showing a hardware configuration of a karaoke system 6. FIG.

Embodiments of the present invention will be described below with reference to the drawings.
(1) Configuration FIG. 1 is a block diagram illustrating a configuration of the karaoke apparatus 1.
The karaoke apparatus 1 includes a control unit 11, a storage unit 12, a display unit 13, an operation unit 14, a voice input unit 15, a voice processing unit 16, and a voice output unit 17, and singing evaluation for evaluating a singer's singing. Functions as a device. 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. Thus, each part of the karaoke apparatus 1 is controlled via the bus. The display unit 13 includes a VRAM (Video RAM), a liquid crystal display, and a liquid crystal driving circuit, and displays various images under the control of the control unit 11. The operation unit 14 includes various keys for selecting a music or selecting an operation mode (hereinafter referred to as a scoring mode) for singing a song. A signal corresponding to the operation is output to the control unit 11. The audio input unit 15 is, for example, a microphone, and collects sound and outputs an audio signal (analog signal) representing the collected sound. The audio processing unit 16 converts the analog signal input from the audio input unit 15 into digital data, or converts the digital data supplied from the control unit 11 into an analog signal and outputs the analog signal to the audio output unit 17. Moreover, the audio | voice processing part 16 extracts the information required for the scoring process of the control part 11 from the voice of the singer converted into digital data. The audio output unit 17 is, for example, a speaker, and emits a sound corresponding to the audio signal that is output from the digital data converted into an analog signal by the audio processing unit 16.

  The storage unit 12 is a storage unit for storing a computer program executed by the control unit 11 and data used at the time of execution, and is a non-volatile auxiliary storage device such as an HD (Hard Disk). The storage unit 12 includes a music data storage area 121 and a scoring reference data storage area 122, and stores a scoring program 123 and a scoring specific gravity table 124. The song data storage area 121 stores song data representing accompaniment sounds and lyrics of a plurality of songs. In the scoring reference data storage area 122, data for scoring the skill of singing and scoring reference data related to the scoring standard for scoring the skill of singing are described. The scoring program 123 describes a procedure for calculating the scoring result of the singer's singing skill input from the voice input unit 15 using the scoring reference data, the scoring specific gravity table 124, and the like.

FIG. 2 is a diagram showing the structure of the music data 2 stored in the music data storage area 121.
The music data 2 is composed of a header, accompaniment data, and lyrics data. In the header, metadata relating to music attributes such as a music number that uniquely identifies the music, a music name that indicates the name of the music, a genre that divides the music according to the style and content, and a music time that indicates the performance time of the music are included. include. The accompaniment data is data representing an accompaniment sound by an instrument or the like that is a secondary performance when the singer's singing is a main performance. The accompaniment data is read by the control unit 11 and supplied to the audio processing unit 16, and a sound corresponding to the accompaniment data is emitted from the audio output unit 17. The lyric data is data representing the lyric of the music, and is read by the control unit 11 in synchronization with the reading of the accompaniment data and supplied to the display unit 13. The display unit 13 displays the lyrics represented by the supplied lyrics data.

FIG. 3 is a diagram showing the structure of the scoring reference data 3 stored in the scoring reference data storage area 122.
The scoring reference data 3 includes a header, scoring reference data (first data), singing reference data (second data), and scoring data. In the header, metadata relating to the attributes of the music such as a music number that uniquely identifies the music is described.
The scoring reference data is data indicating the pitch or beat length indicating the melody of the music, and serves as a reference for scoring the song of the singer.
The score data represents the highest score of the scoring results for the song with the song number described in the header.
The singing reference data represents the singing voice of the singer when the highest score is obtained in the singing of the music having the music number described in the header. The singing reference data is used as a reference when the control unit 11 scores the singer's singing together with the scoring reference data. Since the singing reference data is the raw singing voice of the singer who has obtained the highest score, the singing reference data is information including the singing skill of the singer, for example, characteristics such as shabbing and vibrato.

FIG. 4 is a diagram showing an example of the contents of the scoring specific gravity table 124 stored in the storage unit 12.
The scoring specific gravity table 124 describes the specific gravity of scoring reference data and singing reference data at the time of scoring in association with the range of the highest scoring represented by the scoring data of scoring reference data 3. In the case of the example in the figure, for example, when singing a song with a maximum score of 81 to 99 points so far, the scoring result calculated based on the scoring reference data is adopted at a rate of 19%, This means that the scoring result calculated based on the singing standard data is adopted at a rate of 81%. Also, for example, when singing a song with a maximum score of 0 or no description (that is, the song has never been scored in the past), the score was calculated based on the scoring reference data. This means that the scoring result is adopted at a rate of 100%, and the scoring result calculated based on the singing standard data is not adopted at all. Thus, in the scoring specific gravity table 124, information on which of the scoring results calculated based on the scoring reference data and the scoring results calculated based on the singing reference data is to be performed with a specific gravity. It has been described.

(2) Operation A singer selects a song to be sung by inputting a song number or a song name of a song to be sung by operating a plurality of keys provided in the operation unit 14 of the karaoke apparatus 1, and Select the scoring mode as the operation mode. The operation unit 14 receives the operation of the singer and outputs a signal corresponding to the operation to the control unit 11. The control unit 11 reads the music data 2 corresponding to the selected music from the music data storage area 121 using the music number or the music name as a key, and stores it in the RAM. Further, the control unit 11 reads out the scoring reference data 3 from the scoring reference data storage area 122, transfers it to the RAM, and stores it.

First, a process when a singer performs a song according to accompaniment and lyrics guidance by the karaoke apparatus 1 will be described.
The control unit 11 sequentially outputs the accompaniment data stored in the RAM to the sound processing unit 16. The sound processing unit 16 converts the accompaniment data that is digital data into an analog signal and outputs the analog signal to the sound output unit 17, and causes the sound output unit 17 to emit a sound corresponding to the accompaniment data. This emitted sound becomes an accompaniment when the singer sings. In synchronism with the accompaniment output, the control unit 11 sequentially outputs the lyrics data stored in the RAM to the display unit 13. The display unit 13 displays lyrics according to the lyrics data. In this way, the accompaniment represented by the accompaniment data is output from the audio output unit 17, and at the same time, the lyrics represented by the lyric data are displayed on the display unit 13, so that the singer can sing according to the guidance of the accompaniment and the lyrics. Do. This singing sound is picked up by the voice input unit 15 and is emitted from the voice output unit 17 through voice processing such as amplification and echo provision by the voice processing unit 16.

Next, the scoring process performed by the control unit 11 will be described.
The voice input unit 15 collects the voice of the singer, converts the collected voice into an analog audio signal, and outputs the analog audio signal to the voice processing unit 16. The sound processing unit 16 converts the analog audio signal into a digital audio signal, generates song data, and supplies the song data to the control unit 11. The control part 11 acquires this song data. And the control part 11 extracts each pitch and the sound volume for every pitch from the audio | voice waveform which this song data shows. Hereinafter, the pitch extracted from the singing data by the control unit 11 is referred to as “singing pitch data”, and the volume extracted from the singing voice data by the control unit 11 is referred to as “singing volume data”. The control unit 11 performs a singing scoring process based on the singing pitch data and the singing volume data.

FIG. 5 is a flowchart showing a scoring process performed by the control unit 11.
The control unit 11 starts scoring calculation processing according to the procedure described in the scoring program 123 (step S51). In this scoring calculation process, the degree of coincidence is obtained by comparing the singing data and the scoring reference data, the degree of coincidence is obtained by comparing the singing data and the singing reference data, and each degree of coincidence is obtained from the scoring specific gravity table 124. Evaluate in consideration of the contents of, and obtain the scoring results.

  Specifically, first, the control unit 11 extracts each pitch in time series from the scoring reference data and stores it in the RAM, and also extracts the pitch (basic frequency), that is, the pitch in time series from the singing reference data. At the same time, the volume corresponding to the pitch is extracted in time series and stored in the RAM. Next, the control unit 11 obtains the difference between the singing pitch data extracted by the voice processing unit 16 and the pitch of the scoring reference data and stores it in the RAM. Further, the control unit 11 obtains a difference between the singing pitch data extracted by the voice processing unit 16 and the pitch of the singing reference data and stores it in the RAM. Furthermore, the control part 11 calculates | requires the difference of the song volume data extracted by the audio | voice process part 16, and the volume of song reference | standard data for every pitch, and memorize | stores it in RAM. When the singing target song ends, the control unit 11 sums up the contents stored in the RAM, that is, the difference between the singing data and the scoring reference data, and the difference between the singing data and the singing reference data, respectively. To do. The singing score is a value corresponding to the total difference, and the higher the difference value, the lower the scoring result.

  Here, for example, the singer was a singing slightly deviated from the melody stipulated by the scoring standard data, but is stipulated by the singing standard data. Suppose you sing. In this case, since the control unit 11 determines a singing part slightly deviating from the melody as a difference from the scoring reference data in scoring based on the comparison with the scoring reference data, the scoring result is, for example, 70. Calculate with points. On the other hand, in the scoring based on the comparison with the singing reference data, the control unit 11 calculates the scoring result as, for example, 80 points because the singing skill of the singer relatively matches the content of the singing reference data. Contrary to the above, when the singer performs a singing that is relatively suitable for the melody of the music but lacks in terms of singing skill, the control unit 11 compares it with, for example, scoring reference data. The scoring result based on the score is calculated as 80 points, for example, and the scoring result based on the comparison with the singing reference data is calculated as 70 points, for example. In this way, a score difference occurs in the scoring results based on the scoring reference data and the singing reference data. In the following description, a case where the control unit 11 calculates the scoring result based on the comparison with the scoring reference data as 70 points and calculates the scoring result based on the comparison with the singing reference data as 80 points will be described as an example. To do.

  When 70 points are described as the score data in the score reference data 3 of the sung song, the score data 70 corresponds to “61 to 80” in the range of the score data in the score specific gravity table 124. In this case, the control unit 11 obtains a value (25.2 points) obtained by multiplying the scoring result (70 points) calculated based on the scoring reference data by 0.36 and the scoring result calculated based on the singing reference data. A value obtained by multiplying (80 points) by 0.64 (46.2 points) (71.4 points) is calculated as a singer's singing score result.

  Next, the control part 11 determines whether the scoring result calculated by step S51 exceeded the scoring data of the scoring reference data 3 corresponding to the sung music (step S52). And if the control part 11 judges that the singer's song scoring result exceeded the scoring data of the scoring reference data 3 (step S52; YES), the scoring data of the scoring reference data 3 will be used as the singer's singing scoring result. And the singing data to be evaluated is stored in the storage unit 12 as the singing reference data of the scoring reference data 3 (step S53). That is, the singing data and the score data newly evaluated as the highest score are stored as the scoring reference data 3 in the storage unit 12 instead of the singing standard data and the scoring data before the evaluation. And the control part 11 displays the scoring result of a singer's song on the display part 13 (step S54). In step S52, if the control unit 11 determines that the singer's singing result is less than the scoring data of the scoring reference data 3, the process of step S53 is skipped and the scoring of step S54 is performed. It progresses to the process which displays a result.

  On the other hand, for example, if the score data in the scoring reference data 3 of the sung song is 0 points or nothing is described, it corresponds to “0 points or no description” in the scoring specific gravity table 124. This means that the scoring result calculated based on the reference data is adopted at a rate of 100%, and the scoring result calculated based on the singing reference data is not adopted at all. That is, the control unit 11 uses a value (70 points) obtained by multiplying the singing result (70 points) of the singing of the singer based on the scoring reference data by 1.00, and the singer's singing result (70 points). ). And the control part 11 implements the process after step S52 similarly to the process mentioned above.

When singing a song based only on the scoring standard data, if the song is sung according to the melody of the music that is determined in advance when composing, the scoring result will be high, but the length of the note will be changed intentionally. When singing using the singing technique, the melody of the music does not match the singing, so the scoring result is low. Therefore, for example, imitating a very good singer like a professional singer does not necessarily give a high scoring result.
On the other hand, when scoring the skill of a singer's singing according to the above-described embodiment, not only the scoring reference data determined in advance at the time of composing, but also the raw singing voice that has the highest scoring result in the music By taking into account the criteria for scoring, scoring can be made that legitimately evaluates singing using a high-level singing skill that is intentionally removed from the melody of the music.

  The scoring specific gravity table 124 is such that the specific gravity of the singing reference data increases as the past highest score increases, and the specific gravity of the singing reference data decreases as the past highest score decreases. For example, a case where the highest score in the past for a certain piece of music is low means that the skill level of a singer who has sung that piece of music in the scoring mode in the past is not so high. In such a case, if scoring is performed with emphasis on the singing of the singer, there is a possibility that the evaluation is not appropriate. For this reason, when the past highest score is low, the specific gravity of the singing reference data is reduced. On the other hand, a case where the highest score in the past for a certain musical piece is high means that the skill level of the singer who has sung the musical piece in the scoring mode in the past is high. In such a case, the singing technique using the singing technique can be properly evaluated by including the singing technique of the singer with a high singing level in the scoring standard.

(3) Modification (3-1) Modification 1
The singing reference data of the scoring reference data 3 is not limited to the singing voice of the singer who uses the karaoke apparatus 1 but may be data recorded in advance by a person who provides the scoring mode service. In addition, the range of scoring data and the corresponding specific gravity (numerical values) described in the scoring specific gravity table 124 shown in FIG. 4 are not only those who provide the scoring mode service, but also users of karaoke devices and singers. May be set / changed depending on the type of scoring mode selected. When the user of the karaoke apparatus sets / changes the specific gravity (numerical value) described in the scoring specific gravity table 124, the user may operate the operation unit 14 and input an arbitrary specific gravity to the karaoke apparatus 1. Also, there are two types of scoring modes: the melodic emphasis mode that increases the specific gravity of the scoring results based on the comparison with the scoring standard data, and the singing skill emphasis mode that increases the specific gravity of the scoring results based on the comparison with the singing standard data. Two kinds of different scoring specific gravity tables 124 may be stored in the storage unit 12, and the control unit 11 may use them according to the scoring mode.
Further, for example, a singing voice of a person (for example, a singer) who has a specialized singing technique is described in the singing reference data, and the singing reference data is 1.00 and the grading reference data is 0.00 in the scoring specific gravity table 124. If the specific gravity is described, the singer can score not only the melody of the music but also the model (simulation) of the singer's singing method.

(3-2) Modification 2
The singing reference data of the grading reference data 3 is that the singing voice at the time of singing is described when the singing grading result in the corresponding music is the highest score. However, the singing reference data may take into account elements other than the scoring results. For example, the structure of the scoring reference data 3 shown in FIG. 3 can be described with 3 singing standard data and 3 scoring data, and the control unit 11 has the singer's scoring result within 3rd place in the corresponding music. In the case of singing, the singing voice at the time of singing is described in the singing standard data, and the scoring result is described in the score data. Then, the singer of the karaoke apparatus 1 views the singing standard data for three items of the scoring reference data 3 of the desired music by operating the operation unit 14, and the singer himself designates suitable singing standard data. . Then, the control part 11 adds a predetermined score to the score data of the singing reference data suitable for the singer. The ranking of the scoring result of the singing standard data of the scoring reference data 3 may vary depending on the score. That is, since the scoring result based on the scoring reference data 3 is not necessarily faithful to the order suitable for the singer, the ranking reflecting the intention of the singer is possible.

(3-3) Modification 3
Another device in place of the karaoke device 1 may store the music data 2 and the scoring reference data 3.
FIG. 6 is an overall configuration diagram of the karaoke system 6.
The karaoke device 61 includes a communication unit 18 in addition to the configuration of the karaoke device 1. The communication unit 18 includes a communication circuit and a communication interface, and is connected to the information management server 63 via a network 62 which is a communication network for transmitting information such as a WAN (Wide Area Network) and a LAN (Local Area Network). Communicate. In addition, the storage unit 631 of the information management server 63 is provided with a music data storage area 121 and a scoring reference data storage area 122 stored in the storage unit 12 of the karaoke apparatus 1, and the music data 2 and scoring reference, respectively. Data 3 is stored. Therefore, the karaoke apparatus 61 is not provided with the music data storage area 121 and the scoring reference data storage area 122 in the storage unit 12, and the song data selected by the singer is sent to the information management server 63 via the network 62. The music number or the music name is notified, the music data 2 and the scoring reference data 3 are requested, and they are acquired and stored. And the control part 11 of a karaoke apparatus performs the process shown in FIG. 5 using the acquired music data 2 and the scoring reference data 3. FIG. Further, the karaoke apparatus 61 does not sequentially acquire the music data 2 and the scoring reference data 3 from the information management server 63 according to a request from the singer, but stores frequently used data according to preset conditions. The time required for sequential acquisition may be shortened by storing (cache) the data.

Further, another device instead of the karaoke device 1 may perform the scoring calculation process.
For example, the scoring program 123 and the scoring specific gravity table 124 stored in the storage unit 12 of the karaoke apparatus 61 in FIG. 6 are stored in the storage unit 631 of the information management server 63 instead of the storage unit 12 of the karaoke apparatus 61. May be. And the control part 11 of the karaoke apparatus 61 transmits the song data of the singer extracted by the voice processing part 16 to the information management server 63 from the communication part 18. The information management server 63 performs scoring calculation processing in the same manner as in step S51 of FIG. 5, and obtains the degree of coincidence between the singers' singing data and scoring reference data, and the singing data and singing reference data, respectively. The degree of coincidence is evaluated in consideration of the contents of the scoring specific gravity table 124, and a scoring result is obtained. Then, the information management server 63 transmits the scoring result to the communication unit 18 of the karaoke apparatus 1. The control unit 11 of the karaoke apparatus 1 displays the received scoring result on the display unit 13.

  According to the above modification, the administrator of the karaoke system 6 can perform unified management of the data provided in the storage unit 631 of the information management server 63, and can store the data in each karaoke device 61. There is no need. In addition, the administrator of the karaoke system 6 can control to distribute predetermined data only to an arbitrary karaoke apparatus 61. In FIG. 6, one information management server 63 and three karaoke devices 61 are illustrated, but the number of karaoke devices 61 and information management servers 63 connected to the network 62 is not limited thereto.

(3-4) Modification 4
In the embodiment described above, singing data and scoring reference data, singing data and scoring reference data are compared with each other, and the pitch and volume, which are characteristic information corresponding to each of the singing data and singing reference data, are compared and the singing is scored according to the difference between the two. . However, the characteristic information is not limited to pitch and volume. That is, the singing skill of a singer may be scored by comparing characteristic information that the singing data, the grading reference data, and the singing reference data have in common.
Moreover, although the singer's song scoring result was displayed on the display unit 13, it is not limited to this. For example, a singer's singing score may be read out as audio, or may be output in the form of data to a personal computer, a portable terminal, or the like.

DESCRIPTION OF SYMBOLS 1 ... Karaoke apparatus, 11 ... Control part, 12 ... Storage part, 121 ... Music data storage area, 122 ... Scoring reference data storage area, 123 ... Scoring program, 124 ... Scoring specific gravity table, 13 ... Display part, 14 ... Operation part , 15 ... voice input unit, 16 ... voice processing unit, 17 ... voice output unit, 18 ... communication unit, 2 ... music data, 3 ... scoring reference data, 6 ... karaoke system, 61 ... karaoke device, 62 ... network, 63 ... information management server, 631 ... storage unit.

Claims (3)

First data representing the melody of the music, second data representing the singing voice that has obtained an evaluation result within a predetermined rank in the singing of the music, and third data representing the evaluation result of the singing by the singing voice Storage means for storing
Obtaining means for obtaining singing data representing the singing voice to be evaluated;
Comparison means for comparing the singing data and the first data, comparing the singing data and the second data obtained the highest evaluation result among the second data ,
Based on the comparison result between the singing data and the first data, and the comparison result between the singing data and the second data obtained the highest evaluation result among the second data , the singing data is An evaluation means for evaluating the singing by the singing voice to represent and outputting the evaluation result ,
When the evaluation result output by the evaluation unit exceeds the third data stored in the storage unit, the storage unit stores the third data and the second data corresponding to the third data. Instead, the evaluation result and song data corresponding to the evaluation result are stored.
A singing evaluation apparatus characterized by that . Before Symbol evaluation means,
The higher the third data stored in the storage means, the higher the specific gravity of the comparison result between the singing data and the second data than the comparison result between the singing data and the first data. The singing evaluation device according to claim 1, wherein Sound emission means for emitting sound according to the second data stored in the storage means;
Operation means for accepting an operation on third data which is an evaluation result of the second data emitted from the sound emission means;
The singing evaluation apparatus according to claim 1, comprising:

Images