JP6589521B2 - Singing standard data correction device, karaoke system, program - Google Patents

Description

  The present invention relates to a technique for correcting singing reference data representing a reference melody for singing karaoke music.

  Conventionally, in a karaoke system, a singing standard data representing a singing melody of a karaoke song is used to score a singing by a singer, or to generate a guide vocal or a piano roll that displays a singing melody. What has a singing support function is known.

  For example, in Patent Document 1, a reference speech synthesis parameter generated from reference data (singing reference data) serving as a singing scoring reference is modified according to a setting input from an external input means, and a modified reference speech synthesis is performed based thereon. Techniques for generating data (modified guide vocals) are disclosed.

Japanese Patent Laying-Open No. 2015-87617

  Generally, singing standard data used in karaoke is artificially created for use in karaoke, and therefore may contain arrangements and data errors by the producer. Therefore, the reference melody of the singing represented by the singing reference data does not necessarily match the singing melody of the original song that is generally recognized by the public.

  If singing standard data that does not match the original song is used for singing, even if the singer sings close to the original song melody (how the singing feels good for the general public), the score is low There are times when it can only be obtained. Further, when a guide vocal or piano roll is generated based on singing reference data that does not match the original music, it may be an obstacle to performing a singing that feels good for the user. As described above, in the conventional singing standard data, there is a problem that the way of singing along the singing standard data is different from the way of singing that seems to be good for general users.

  The present invention has been made to solve the above problems. An object of the present invention is to provide a technique for correcting singing reference data based on evaluation by human subjectivity.

  A singing reference data correction apparatus according to an aspect of the present invention includes a storage unit, an acquisition unit, an extraction unit, a difference generation unit, a correction unit, and an update unit. A memory | storage means memorize | stores the song data by which the singing voice by the singer of karaoke music was matched and recorded, and the song reference data showing the reference | standard melody of the song of karaoke music. The acquisition unit acquires evaluation information representing an evaluation determined by a listener who has listened to the singing voice reproduced based on the singing data stored in the storage unit.

The extraction unit extracts song data in which the evaluation represented by the evaluation information acquired by the acquisition unit satisfies a predetermined high evaluation criterion from the song data stored in the storage unit. The difference generating means compares the singing data extracted by the extracting means with the singing melody data corresponding to the song data of the singing data, and the singing melody represented by the singing data and the reference melody represented by the singing reference data. Difference data representing the difference between and is generated. The correcting means applies the difference data generated by the difference generating means to the singing reference data to correct the singing reference data. The updating means updates the singing reference data before correction stored in the storage means with the singing reference data corrected by the correcting means.

  According to the present invention, the singing reference data can be corrected with respect to the part of the singing by comparing the singing reference data with the singing data highly evaluated based on the subjectivity of the listener of the singing voice. By doing in this way, it is possible to match the singing reference data to the way of singing that feels good in human subjectivity.

  The singing reference data correction device according to the second aspect of the present invention has the following characteristics. That is, the extraction means extracts a plurality of song data for the same piece of music. The averaging means averages singing melody expressed by a plurality of singing data related to the same piece of music extracted by the extracting means. The difference generation means generates the difference data by comparing the song data averaged by the averaging means with the song reference data corresponding to the song of the averaged song data. By doing in this way, singing reference | standard data can be corrected in the aspect which reflected the evaluation based on more people's subjectivity statistically.

The singing standard data correction device according to the third aspect of the present invention has the following characteristics. That is, an acquisition means acquires the evaluation information showing the evaluation determined for every section which divided the music into plurality about song data. The extraction means extracts a plurality of pieces of singing data for the same piece of music in which the evaluation for each section represented by the evaluation information satisfies a predetermined high evaluation criterion. An averaging means averages the song melody of the same area which satisfy | fills high evaluation criteria about the some song data regarding the same music extracted by the extraction means. The difference generation means generates difference data by comparing the singing data for each section averaged by the averaging means with the corresponding section in the singing reference data. In this manner, by extracting a section having a particularly high evaluation from the entire music and correcting the singing reference data, the singing reference data can be corrected in a manner that more accurately reflects the evaluation based on the human subjectivity.

  The singing standard data correction device according to the fourth aspect of the present invention has the following characteristics. That is, the singing technique data stored in the storage means includes singing technique evaluation information that represents an evaluation standard regarding a specific singing technique and a point in time when the singing technique is developed. The detecting means detects a specific singing technique included in the singing voice represented by the singing data extracted by the extracting means and a point in time when the singing technique is expressed. The correcting means corrects the singing technique evaluation information of the singing reference data based on the singing technique detected by the detecting means and the aggregation at the time when the singing technique is expressed. By doing in this way, the evaluation standard regarding the singing technique in singing standard data can be corrected according to the content and timing of the singing technique included in the song highly evaluated by the listener.

  A karaoke system according to an aspect of the present invention includes a singing data generation unit, an upload unit, a reproduction unit, a reception unit, a storage unit, an acquisition unit, an extraction unit, a difference generation unit, a correction unit, An update unit and a control unit are provided. The singing data generating means records the singing voice by the singer of the played karaoke music and generates singing data associated with the music. The upload unit uploads the song data generated by the song data generation unit to the storage unit. The reproduction means reproduces the singing voice based on the song data stored in the storage means. The accepting means accepts an evaluation of the song data reproduced by the reproducing means from the listener.

  The storage means stores singing data uploaded by the uploading means and singing reference data representing a reference melody for singing karaoke music. The acquisition means acquires evaluation information representing the evaluation received by the reception means in association with the song data to be evaluated. The extraction unit extracts song data in which the evaluation represented by the evaluation information acquired by the acquisition unit satisfies a predetermined high evaluation criterion from the song data stored in the storage unit. The difference generation means compares the singing data extracted by the extracting means with the singing melody data corresponding to the song data of the singing data, and the melody represented by the singing data and the reference melody represented by the singing reference data, Difference data representing the difference is generated. The correcting means applies the difference data generated by the difference generating means to the singing reference data to correct the singing reference data. The updating means updates the singing reference data before correction stored in the storage means with the singing reference data corrected by the correcting means. The control means executes a support function for supporting the singing of the karaoke music according to the singing reference data stored in the storage means.

  According to the karaoke system according to the present invention, generation and upload of singing data, acquisition of evaluation of singing data, correction of singing standard data based on comparison between singing data and singing standard data, singing based on the modified singing standard data Various functions such as support can be performed consistently.

  The present invention can also be realized in the form of a program executed on a computer or a storage medium for storing the program.

The block diagram showing the structure of a karaoke system. Explanatory drawing showing the outline | summary of operation | movement of a karaoke system. The flowchart showing the procedure of the process which each part of a karaoke system performs.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In addition, this invention is not limited to the following embodiment, It is possible to implement in various aspects.
[Description of karaoke system configuration]
As illustrated in FIG. 1, the karaoke system 1 is configured such that the karaoke server 2, the karaoke device 3, and the user playback terminal 4 are connected to the Internet 10 and can communicate with each other. In FIG. 1, the karaoke apparatus and the user reproduction terminal 4 are shown one by one for the sake of simplicity. In practice, however, it is assumed that these devices are implemented in a number of connected ways.

  The karaoke server 2 is a server device configured by a computer system, and includes a control unit 21, a storage unit 22, and a communication unit 23. The karaoke server 2 has a function of accumulating song data uploaded from the karaoke apparatus 3 and distributing the accumulated song data via the Internet 10 in response to a request from the user reproduction terminal 4. Further, the karaoke server 2 has a function of holding singing reference data corresponding to karaoke music played in the karaoke apparatus 3 and distributing it to the karaoke apparatus 3. Moreover, the karaoke server 2 has a function of correcting the singing reference data of the corresponding karaoke piece based on the evaluation of the singing data received from the listener of the user reproduction terminal 4 as a characteristic function in the present invention.

The control unit 21 is an information processing apparatus configured mainly with a CPU, RAM, ROM, and the like, and controls the entire apparatus. The control part 21 implement | achieves various functions by performing arithmetic processing according to the program stored in the storage medium. The storage unit 22 is a storage device for storing programs, various databases, and the like. The communication unit 23 is a communication interface for connecting the karaoke server 2 to the Internet 10 and performing data communication with the karaoke apparatus 3 and the user reproduction terminal 4.

  The karaoke server 2 includes in the storage unit 22 a singing database for storing singing data released in the singing recording and distribution service provided by the karaoke system 1. In this singing database, a plurality of singing data uploaded from the karaoke apparatus 3 via the Internet is accumulated. The singing data is composed of digital audio data obtained by sampling the waveform of the singing voice by the singer.

  Further, the karaoke server 2 includes a singing reference database for storing singing reference data corresponding to karaoke pieces played in the karaoke apparatus 3 in the storage unit 22. The singing reference data is composed of information representing a reference singing melody (singing pitch and pronunciation time) that serves as a model for singing. In addition, this singing standard data includes singing technique evaluation information that defines the evaluation standard based on the type of singing technique and the point of onset for a specific singing technique (technique) such as vibrato, shackles, and fist. To do. This singing reference data is distributed to the karaoke device 3 as necessary, and is used as a reference for the singing scoring function, guide vocal creation, and piano roll display in the karaoke device 3.

  The karaoke device 3 is embodied by a business karaoke device installed in a karaoke store, for example. The karaoke apparatus has a function of playing karaoke music requested by a user (singer). Moreover, the karaoke apparatus 3 has a function of uploading to the karaoke server 2 singing data in which singing voices input via a microphone during the performance of karaoke music are recorded. Thereby, singing data of many singers are accumulated in the karaoke server 2. Further, the karaoke apparatus 3 has a function of performing singing support functions such as singing grading, guide vocal creation, and piano roll display using the singing reference data distributed from the karaoke server 2. Since other functions and configurations of the karaoke apparatus 3 are in accordance with known techniques, detailed description thereof is omitted here.

  The user reproduction terminal 4 is an information processing apparatus having a function of acquiring song data from the karaoke server 2 and reproducing and outputting it. The user reproduction terminal 4 is embodied by, for example, a business karaoke apparatus installed in a karaoke store, a high-functional mobile phone terminal such as a smartphone owned by the user, a personal computer, or the like. The user reproduction terminal 4 performs data communication with the karaoke server 2 via the Internet 10. The user reproduction terminal 4 receives a request for song data from the user (listener). And the song data acquired from the karaoke server 2 according to the request is reproduced, and the reproduced song voice is output. Further, the user reproduction terminal 4 has a function of receiving the evaluation determined by the listener for the reproduced song data and notifying the karaoke server 2 of evaluation information representing the evaluation content.

[Explanation of Karaoke system operation]
An outline of a series of operations for correcting the singing reference data stored in the karaoke server 2 will be described with reference to FIG. Hereinafter, the operation of each device of the karaoke system 1 will be described in the order of numbers [1] to [7] shown in FIG.

  [1] The karaoke apparatus 3 associates the singing data in which the singer's singing voice input at the time of the performance of the karaoke music is recorded with the music identification information corresponding to the tune of the played karaoke music. Upload to. The karaoke server 2 stores the song data uploaded from the karaoke apparatus 3 in the song database of the storage unit 22 in association with the music identification information. The song data stored in the karaoke server 2 is disclosed to the user playback terminal 4 through a website on the Internet corresponding to the song recording and distribution service provided by the karaoke system 1.

  [2] When the karaoke server 2 accepts a request for distribution from the user reproduction terminal 4 for the released song data, the karaoke server 2 distributes the song data corresponding to the request to the user reproduction terminal 4 that is the request source. The user reproduction terminal 4 reproduces the singing data acquired from the karaoke server 2 and outputs the singing voice.

  In addition, a “tag”, which is information that functions as a mark for searching and classifying the released song data, is freely given from each user playback terminal 4 to the website where the song data is released. It is assumed that it has a function that can The tag given to the singing data includes various information such as evaluations, impressions, comments, etc. regarding the skill of the singing determined by the listener.

  In addition, this tag can be assigned in association with each performance section obtained by dividing the song data of the song data into a plurality of predetermined sections. Specifically, when a tag that evaluates “good” is input by the listener when a performance section with singing data is being played back on the user playback terminal 4, it is associated with the performance section at the time of input. A “good” tag is given.

  The user reproduction terminal 4 notifies the karaoke server 2 of information representing a tag assigned in association with the performance section of the song data as evaluation information for the song data. This evaluation information includes identification information of the singing data to be evaluated, the contents of the tag given to the singing data, and information representing the performance section to which the tag is given. The karaoke server 2 registers the contents of the tag included in the evaluation information received from the user reproduction terminal 4 in the song database in association with the performance section of the song data corresponding to the evaluation information. In the example of FIG. 2, in the song database of the storage unit 22, tags such as “good”, “interesting”, and “crying” are associated with each performance section for a plurality of song data A to D. Registered.

  [3] The karaoke server 2 highly evaluates singing data stored in the singing database from a singing data group corresponding to a specific karaoke piece (hereinafter, a piece to be corrected) that is a target of correcting the singing reference data. A plurality of song data satisfying the conditions are extracted. Specifically, singing data with a predetermined tag indicating that the singing evaluation is high is extracted as the high evaluation condition. In the example of FIG. 2, singing data with a “good” tag is extracted as singing data satisfying the high evaluation condition from the singing data groups A to D. In addition, about the singing data with which the tag with high evaluation was attached to at least 1 among the several performance sections which comprise one singing data, one whole singing data may be extracted, or the tag with high evaluation was attached. You may extract for each performance section unit.

  [4] The karaoke server 2 performs an averaging process and a singing skill extraction process on the song data group related to the music to be corrected extracted in [3]. Averaging processing averages the pitch of the singing melody in the extracted song data group for each singing point, and averages the length of time each sound constituting the singing melody in the extracted song data group appears It is the process which produces | generates the average data regarding song data with high evaluation by converting. In the present embodiment, from each extracted song data, a portion with a high evaluation tag is extracted for each performance section, and the singing melody in the same performance section with a high evaluation tag is averaged. Shall be

The singing technique extraction process detects a specific singing technique (for example, vibrato, shackle, fist, etc.) from the singing melody represented by each extracted singing data, thereby singing the song used in the highly evaluated singing data. This is a process for obtaining statistical information about the skill. Specifically, the type of singing technique detected in the performance section with a high evaluation tag of the extracted singing data group and the time of its expression are tabulated, and the detection result in which the type of singing technique and the time of expression are common Is identified as an exemplary singing skill. In addition, since the detection method of the singing technique in karaoke is a well-known technique, detailed description is abbreviate | omitted.

  [5] The karaoke server 2 compares the singing melody represented by the average data generated in [4] with the singing melody represented by the existing singing reference data corresponding to the music to be corrected, Difference data representing the difference is generated.

  [6] The karaoke server 2 corrects the singing melody of the existing singing reference data by merging the difference data generated in [5] with the existing singing reference data. Moreover, the karaoke server 2 corrects the evaluation weights according to the exemplary singing technique extracted in [4] for the singing technique evaluation standard represented by the singing technique evaluation information included in the existing singing standard data. Specifically, when the extracted exemplary singing technique is not included in the existing singing technique evaluation information, an evaluation criterion regarding the type and time of expression of the exemplary singing technique is given to the singing technique evaluation information. Moreover, when the singing technique included in the existing singing technique evaluation information does not match the extracted singing technique, the evaluation standard regarding the singing technique is deleted from the singing technique evaluation information. The karaoke server 2 updates the existing singing standard data and singing technique evaluation information stored in the singing standard database to the contents of the modified singing standard data and singing technique evaluation information.

  [7] The karaoke server 2 distributes the updated singing reference data to the karaoke apparatus 3. The karaoke apparatus 3 performs singing support functions such as singing scoring, guide vocal creation, and piano roll display in accordance with the singing reference data distributed from the karaoke server 2.

[Description of processing related to correction of singing standard data]
About the process which concerns on correction of singing standard data, the procedure which each apparatus of the karaoke system 1 performs is demonstrated, referring the flowchart of FIG.

  In S100, the karaoke apparatus 3 plays the karaoke music requested by the user (singer). At this time, the karaoke apparatus 3 records the singing voice input through the microphone during the performance of the karaoke music and generates singing data. Moreover, the karaoke apparatus 3 performs singing support functions, such as a singing score, using the existing singing reference | standard data delivered to each karaoke apparatus 3 from the karaoke server 2. FIG.

  In S102, the karaoke apparatus 3 uploads the singing data recorded in S100 to the karaoke server 2 in association with the music identification information corresponding to the tune of the karaoke music played. The control unit 21 of the karaoke server 2 stores the song data uploaded from the karaoke device 3 in the song database of the storage unit 22 in association with the song identification information.

  In S104, the user reproduction terminal 4 acquires the released song data from the karaoke server 2 and reproduces it. Then, the user reproduction terminal 4 transmits to the karaoke server 2 evaluation information indicating the listener's evaluation on the reproduced song data. The control unit 21 of the karaoke server 2 acquires evaluation information transmitted from the user reproduction terminal 4 that has reproduced the song data. The evaluation information includes identification information of the singing data to be evaluated, the contents of the tag related to the evaluation of the singing given to the singing data, and information representing the performance section to which the tag is given. The control unit 21 of the karaoke server 2 registers the contents of the tag included in the acquired evaluation information in the song database in association with the performance section of the song data corresponding to the evaluation information.

  In S <b> 106, the control unit 21 of the karaoke server 2 extracts “several” singing data satisfying the high evaluation condition from the singing data group corresponding to the music to be corrected among the singing data stored in the singing database. To do. The procedure for extracting the singing data is as described above based on the case of [3] in FIG.

  In S108, the control part 21 of the karaoke server 2 performs a singing technique extraction process for the singing data group extracted in S106, and acquires statistical information on the singing technique of the extracted singing data. Moreover, in S110, the control part 21 of the karaoke server 2 performs an averaging process for the song data group extracted in S106, and generates average data regarding the extracted song data. The contents of the averaging process and the singing technique extraction process are as previously described based on the case of [4] in FIG. In S112, the control unit 21 of the karaoke server 2 compares the average data generated in S110 with the existing singing reference data corresponding to the music to be corrected, and generates difference data representing the difference between the two singing melody. To do.

  In S114, the control unit 21 of the karaoke server 2 corrects the singing melody and singing technique evaluation criteria of the existing singing reference data according to the singing technique statistical information extracted in S108 and the difference data generated in S112. In S116, the control unit 21 of the karaoke server 2 converts the existing singing standard data stored in the singing standard database and the singing skill evaluation information corresponding to the singing standard data into the singing standard data and the singing modified in S114. Update to the skill evaluation information. In S118, the karaoke apparatus 3 acquires new singing standard data corrected from the karaoke server 2, and executes a singing support function such as singing score using the acquired singing standard data.

[effect]
The karaoke system 1 according to the embodiment has the following effects.
By comparing the singing data highly evaluated based on the subjectivity of the listener of the user reproduction terminal 4 and the existing singing reference data, the singing reference data can be corrected with respect to the divergence portion. By doing in this way, the existing singing reference | standard data preserve | saved at the karaoke server 2 can be corrected according to the way of singing which feels good in a human subjectivity.

  In addition, based on the subjectivity of more people by extracting multiple singing data with high evaluation for the same song and modifying existing singing standard data based on average data generated from the extracted singing data group Evaluation can be reflected in singing standard data.

  Moreover, based on the evaluation given to each performance section divided into a plurality of karaoke songs, average data can be generated using song data with high evaluation for each performance section. Thus, by extracting a performance section with particularly high evaluation from the entire music and correcting the singing reference data, the evaluation based on the human subjectivity can be more accurately reflected in the singing reference data. Moreover, the evaluation standard of the singing technique in the existing singing standard data can be corrected based on the statistic information on the type and the time point of expression of the singing technique detected from the highly evaluated singing data.

[Modification]
In the above-described embodiment, a case in which singing data is corrected for each of a plurality of performance sections with respect to the singing data, and the average data is generated for each performance section with a high evaluation tag, and then the singing reference data is corrected Explained. Not only this but the structure which produces | generates average data by dividing | segmenting song data of 1 music whole may be sufficient, without dividing | segmenting song data per performance section unit. In this case, the evaluation with respect to the song data may be given in units of performance sections, or may be given collectively to the song data for one song.

In the above-described embodiment, in S106 of FIG. 3, “single” singing data with a tag having a high evaluation of singing is extracted as singing data satisfying the high evaluation criteria, and the singing melody of these singing data is averaged. We explained an example to make. Separately from this, as a song data satisfying the high evaluation criteria, a configuration may be adopted in which a predetermined number of song data to which more tags with a higher evaluation of the song are attached than other song data is extracted.

  Furthermore, the structure which extracts "one" song data with the most tags with high evaluation of song as song data which satisfy | fills high evaluation criteria may be sufficient. In this case, the process of averaging the extracted song data with other song data (S110 in FIG. 3) is not performed, the extracted song data is compared with the existing song reference data, and the song melodies of both are compared. Difference data representing the difference is generated. And based on the produced | generated difference data, the existing song reference data are corrected.

  The functions of one component in the above embodiment may be distributed as a plurality of components, or the functions of a plurality of components may be integrated into one component. In addition, all the aspects included in the technical idea specified only by the wording described in the claim are embodiment of this invention.

[Correspondence with configuration described in claims]
The correspondence between each configuration of the embodiment and the configuration described in the claims is as follows.
The karaoke server 2 corresponds to a singing standard data correction device. The storage unit 22 of the karaoke server 2 corresponds to storage means. S104 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to an acquisition means. S106 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to an extraction means. S108 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to a detection means. S110 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to an averaging means. S112 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to a difference production | generation means. S114 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to a correction means. S116 of FIG. 3 which the control part 21 of the karaoke server 2 performs corresponds to an update means.

  S100 of FIG. 3 which the karaoke apparatus 3 performs corresponds to a production | generation means. S102 in FIG. 3 executed by the karaoke apparatus 3 corresponds to upload means. S118 in FIG. 3 executed by the karaoke apparatus 3 corresponds to the control means. S104 in FIG. 3 executed by the user reproduction terminal 4 corresponds to reproduction means and reception means.

  DESCRIPTION OF SYMBOLS 1 ... Karaoke system, 2 ... Karaoke server, 21 ... Control part, 22 ... Memory | storage part, 23 ... Communication part, 3 ... Karaoke apparatus, 4 ... User reproduction terminal, 10 ... Internet.

Claims (6)

Storage means for storing singing data in which singing voices by singers of karaoke songs are recorded in association with songs, and singing reference data representing reference melody of singing karaoke songs;
Obtaining means for obtaining evaluation information representing an evaluation determined by a listener who has listened to the singing voice reproduced based on the song data stored in the storage means;
Extraction means for extracting song data in which the evaluation represented by the evaluation information acquired by the acquisition means satisfies a predetermined high evaluation criterion from the song data stored in the storage means;
The singing melody represented by the singing data is compared with the singing melody represented by the singing data and the singing melody represented by the singing data by comparing the singing data extracted by the extracting means and the singing data of the singing data. Difference generation means for generating difference data representing the difference;
Applying the difference data generated by the difference generating means to the singing reference data, correcting means for correcting the singing reference data;
Updating means for updating the singing reference data before correction stored in the storage means with the singing reference data corrected by the correcting means;
A singing standard data correction device comprising: In the singing standard data correction device according to claim 1,
The extraction means extracts a plurality of song data for the same music piece,
An averaging means for averaging singing melodies represented by a plurality of song data relating to the same piece of music extracted by the extracting means;
The difference generation means compares the singing data averaged by the averaging means with the singing reference data corresponding to the song of the averaged singing data, and is expressed by the averaged singing data. Generating difference data representing a difference between the singing melody and the reference melody represented by the singing reference data,
A singing standard data correction device characterized by the above. In the singing standard data correction device according to claim 1,
The acquisition means acquires evaluation information representing an evaluation determined for each section in which the song data is divided into a plurality of pieces of the song data,
The extraction means extracts a plurality of pieces of singing data that satisfy a predetermined high evaluation criterion for each section represented by the evaluation information for the same piece of music,
A plurality of song data for the same music extracted by the pre-Symbol extracting means comprises averaging means for averaging the singing melodies each other the same section that satisfies the high criteria,
The difference generating means compares the singing data for each section averaged by the averaging means with the corresponding section in the singing reference data, and the singing melody represented by the singing data and the singing reference data Generating difference data representing a difference from the reference melody represented by
A singing standard data correction device characterized by the above. In the singing standard data correction device according to any one of claims 1 to 3,
The singing standard data stored in the storage means includes singing technique evaluation information representing an evaluation standard regarding a specific singing technique and a point in time when the singing technique is expressed,
A detection means for detecting a specific singing technique included in the singing voice represented by the singing data extracted by the extracting means and a point in time when the singing technique is expressed;
The correcting means corrects the singing technique evaluation information of the singing reference data based on the singing technique detected by the detecting means and the aggregation at the time when the singing technique is expressed;
A singing standard data correction device characterized by the above. Singing data generating means for generating singing data associated with each song by recording the singing voice by the singer of the karaoke song performed;
Upload means for uploading the song data generated by the song data generating means to the storage means;
Playback means for playing back the singing voice based on the song data stored in the storage means;
Accepting means for accepting an evaluation of the song data reproduced by the reproducing means from a listener;
Storage means for storing singing data uploaded by the uploading means, and singing reference data representing a reference melody for singing the karaoke music;
Obtaining means for obtaining evaluation information representing the evaluation received by the receiving means in association with the song data to be evaluated;
Extraction means for extracting song data in which the evaluation represented by the evaluation information acquired by the acquisition means satisfies a predetermined high evaluation criterion from the song data stored in the storage means;
The singing melody represented by the singing data is compared with the singing melody represented by the singing data and the singing melody represented by the singing data by comparing the singing data extracted by the extracting means and the singing data of the singing data. Difference generation means for generating difference data representing the difference;
Applying the difference data generated by the difference generating means to the singing reference data, correcting means for correcting the singing reference data;
Updating means for updating the singing reference data before correction stored in the storage means with the singing reference data corrected by the correcting means;
In accordance with the singing reference data stored in the storage means, control means for executing a support function for supporting the singing of karaoke music,
A karaoke system comprising: An acquisition procedure for acquiring evaluation information representing an evaluation determined by a listener who has listened to the singing voice reproduced based on the singing data in which the singing voice of the karaoke song singer is recorded in association with the song,
From the storage means for storing the song data, an extraction procedure for extracting song data in which the evaluation represented by the evaluation information acquired in the acquisition procedure satisfies a predetermined high evaluation criterion;
The singing melody represented by the singing data and the singing standard are compared by comparing the singing data extracted in the extraction procedure with the singing standard data representing the singing melody of the karaoke music corresponding to the song data. A difference generation procedure for generating difference data representing a difference from the reference melody represented by the data;
Applying the difference data generated in the difference generation procedure to the singing reference data, correcting the singing reference data,
An update procedure for updating the singing reference data before correction stored in the storage means by the singing reference data corrected in the correction procedure;
A program that causes a computer to execute.

Images