JP4447540B2 - Appreciation system for recording karaoke songs - Google Patents

Description

  The present invention relates to a karaoke recording service that creates and provides a song recording work that includes song accompaniment music and a user's singing voice, and particularly relates to a technique for appropriately managing and providing a song recording work.

In karaoke, a service that can create a CD-ROM or MD in which a user's singing voice is recorded together with accompaniment music is well known and has been well received. In connection with such a karaoke recording system, for example, a number of techniques such as Japanese Patent No. 3515069 and Japanese Patent No. 3522215 have been disclosed. According to this service, singing remains as a record, not just singing like karaoke. For this reason, the user naturally enters the heat, and a professional song recording work is created by putting a lot of energy into a professional style.
Japanese Patent No. 3515069 Japanese Patent No. 3522215

  It is a natural feeling that not only I hear, but also many people want to hear the work I made with a great deal of energy. There is also a singing test service that allows singing professionals to listen to and record their singing recordings. Such services are also synergistic, and new song recordings are being created one after another.

  Therefore, the present inventor has considered creating a database of works in order to use them effectively because the works are born but wasteful. When creating a database of works, it is assumed that many people will ask, and if a database is constructed so that it can be searched from various angles according to the mood of various people, the possibility of providing new services is also possible The present invention was created because it was spread and convenient.

The system according to the present invention is specified by the following items (1) to (11).
(1) A music index database, a song recording database, a user interface, and a karaoke song recording work appreciation system provided with control means. (2) The music index database is for each song played on a karaoke device. The song ID and the song attributes including the song title, artist name, and genre are described in association with each other. (4) The table of contents includes the song ID, singer attributes including the singer name, and singing for each karaoke song recording work. (5) The user interface is the control of the control means. Accordingly, the information is displayed to the user and the user's operation input is received. (6) The control unit causes the user to perform music search based on the music index database via the user interface. (7) Control unit When the user specifies one piece of music by searching for a piece of music, the first process to the fourth process are performed. (8) The first process is based on the song ID of the specified song. (9) The second process is the number of works extracted in the first process, and the singing, and extracts the singer attributes, voice quality evaluation description, and skill evaluation description for a plurality of works associated with the music ID. User attribute, voice quality evaluation description, skill evaluation description, attribute value choices for each of multiple attribute items are displayed on the user interface, and user selection for attribute value choices (10) In the third process, the user selection information received in the second process and the singer attributes, voice quality evaluation description, and skill evaluation description for multiple works extracted in the first process are matched using a matching algorithm. Collating and selecting a karaoke song recording work that is highly compatible with the user selection information (11) The fourth process is a user interface for each work ID of the karaoke song recording work selected in the third process. When the user input indicating that the work being displayed is instructed is received, the corresponding karaoke song recording work is extracted from the song recording database based on the work ID and played.

  According to the present invention, it is possible to select and listen to or use a work that matches the sensibility of the occasion from among many people's singing works. The way of selecting works greatly expands and you may encounter unexpected works, so you can enjoy the search itself. You can effectively use the singing songs that increase one after another. In addition, the number of people who use karaoke is expected to increase as the number of people who make new singing songs will increase.

=== Data structure of the song recording work ===
The song recording work according to the present embodiment exists as data recorded in a well-known karaoke device or a hard disk device of a server operated by a karaoke company. An analog signal obtained by mixing karaoke accompaniment music reproduced by a karaoke apparatus and a singer's singing voice collected from a microphone is digitally converted by a PCM method or the like. It can be played back using a music playback device with a built-in D / A conversion circuit and can be output as sound from a speaker or earphone. In the present invention, this digital data is stored and managed in a song song recording work database described below.

  A combination of recording data of singing voice without karaoke accompaniment music and accompaniment music in the form of karaoke data may be used as a song recording work. When playing a song recording with a karaoke device or the like, karaoke accompaniment music is played in synchronization with the singing voice. Or it is good also as a structure which does not include accompaniment music in a song recording work on the assumption that the accompaniment music of a karaoke data format is combined when reproducing a song recording work.

=== Structure of the Choral Recording Database ===
The singing song recording work database of the present embodiment is constructed by storing the work in a server device operated by a karaoke company and using the work ID as key information. The server device is a general-purpose computer having a well-known configuration. A computer main body including a CPU, a RAM, and a ROM, a hard disk, a communication control unit for data communication with an external device via a network, and a user interface device that accepts user input and presents output to the user Controls peripheral devices such as. On the hard disk, works appropriately ordered through the communication control unit are stored, and a table of contents of works using the work ID as key information is created and stored, and managed as a song recording work database.

  FIG. 1 shows an example of the data structure of the song recording database of this embodiment. The table of contents of this database is configured as a set of records in which attribute values of various attribute items are described for each work. In addition to work ID information such as work ID, work storage location, creation date, etc., the music item identification information such as music ID, music title, music tone information, singer ID, singer's gender and age, etc. Identification information, descriptions that evaluate the voice quality of the song, descriptions that evaluate the skill of the song, and the like are listed. When a database is searched by specifying an attribute value for one or more of any of the attribute items, the corresponding work can be extracted with reference to this table of contents.

  The work ID is an identification code assigned as appropriate when the work data is generated. It may be determined by combining a song ID for identifying a sung karaoke song, a singer ID for identifying a singer, a singing date (creation date), and the like.

  As is well known, the song ID is composed of a 6-digit numerical value that can identify a karaoke song. When the song ID is designated, various attribute information of the song attached to the karaoke device is stored, as is well known in the electronic table of contents technique disclosed in, for example, Japanese Patent No. 3634070 and Japanese Patent No. 3643071. By using the well-known music index database, it is possible to specify attributes related to music such as music name, artist name, and music genre. Of course, the same attribute items as those of the music index database may be stored in this work database.

  The singer ID is information that can identify a singer, such as a member number at a karaoke shop, a home phone number, or a combination of a karaoke shop name and name.

  In this embodiment, music tone information is included as one piece of music identification information. For example, attribute values such as fantastic, realistic, idyllic, lyrical, rhythmic, and dynamic are described. As in the above-described music index database, attribute values are appropriately described by a karaoke company or the like.

  As other music identification information, an attribute item describing characteristics of the music such as a theme expressed by each music and a keyword of each music may be included.

=== Voice quality evaluation description ===
The table of contents of the work database of this embodiment includes a description that evaluates the voice quality of the singer as the attribute item. Here, for each of a plurality of items specified in opposite expressions, such as high or low, masculine or feminine, faint or clear, thick or thin, or not tensioned, Indicates the degree of bias within the item. For example, for masculine or feminine items, the masculine and feminine categories are classified into 5 levels from 1 to 5, with a very masculine voice quality of 1, a very feminine voice quality of 5, The intermediate voice quality that cannot be said to be feminine is 3, so that it describes in which stage the voice quality to be evaluated is located.

  As another description method, for example, the degree of similarity with the voice quality of a famous artist may be described. When determining the degree of similarity, it is also possible to determine by comparing a plurality of items as described above.

  In addition, the evaluation description from various angles is known for voice quality. For example, the expression of soprano, meso-soprano, alto, tenor, baritone, bass is well known in vocal music, but soprano coloratura (voice quality suitable for strong expression), soprano spread (cheerful young) There are also known expression methods that combine sound range and voice quality, such as voice quality suitable for the role of a daughter) and Tenor Ririco (voice quality that has both lyrical and powerful). In research on audio montage that embodies the listening impression of voice quality, reliable and stable voice quality expression is being studied. Although further explanation is omitted in this specification, a database capable of more detailed search can be constructed by including abundant voice quality evaluation descriptions from various aspects.

=== Technical evaluation description ===
As described above, the table of contents of the singing song recording work database of the present embodiment includes a description that evaluates singing skills such as singing evaluation results as attribute items. This is a description of the evaluation results when the karaoke device evaluates the singing content in parallel with the recording process at the time of singing, or the personal computer or music playback device plays the work and evaluates the content after singing It is.

  In the present embodiment, evaluation and description are based on evaluation items of singing skills such as pitch accuracy, vibrato characteristics, and utterance timing characteristics. As a technique evaluation technique, for example, a singing scoring technique as disclosed in Japanese Patent No. 3290918, Japanese Patent No. 3452792 and the like is applied. The details of the skill evaluation technique are not mentioned here.

  The accuracy of the pitch is calculated, for example, by comparing the pitch of the guide vocal included in the karaoke data with the pitch of the voice of the singer for each appropriate section in the performance time series. For example, the score is 100 points when it completely matches. The average value of the values calculated for all singing sections is classified into about five levels according to the score. The pitch accuracy is described as 1 when the number is 80 points or more or fairly accurate, and conversely, when the number is 20 points or less or 5 when it is completely inaccurate.

  The vibrato characteristic describes the depth and speed of vibrato. For example, the vibrato depth is determined by analyzing the amplitude of the pitch fluctuation for each sound of the work data. For example, the depth of the vibrato is determined as 5 levels, for example, 5 for very deep and 1 for very shallow. It is described in principle. In addition, the pitch speed is determined by analyzing the frequency of pitch fluctuation. The speed of the pitch is determined according to the classification of about 5 levels, such as 5 when very fast and 1 when very slow. Is described.

  The utterance timing characteristic describes the accumulation of utterance and the degree of running. For example, each sound of the work data is compared with each sound of the guide vocal, and if the delay in timing at which the first sound of each phrase is pronounced is large, it is determined that the reservoir is large, and if the delay is small, the reservoir is determined to be small. Conversely, if it is ahead rather than delayed, it is determined that it is running (and its degree). These levels are classified into about five levels, for example, 5 when the pool is very large, 3 when there is no pool or running, and 1 when the running is very large.

  In addition, regarding the skill evaluation description, an attribute item such as a fist characteristic describing a degree of how to apply the fist may be included.

=== Search method for a database of vocal recordings Part 1 ===
The system is configured so that the database of singing song recordings can be searched and used when using karaoke with a well-known karaoke apparatus. What is necessary is just to comprise a server apparatus which stored the song recording work database in a karaoke apparatus, or the server apparatus which stored the song recording work database so that data communication is possible. A database of singing song recordings is stored in a portable computer capable of data communication with a karaoke device, and this portable computer is used as a user interface device to accept a user's designated input or present a result to the user to search for a work. You may comprise so that a karaoke apparatus may be notified and a work may be reproduced.

  And as one of the services which can be utilized with a karaoke apparatus, the appreciation menu of a song recording work is prepared on the user interface apparatus of a karaoke apparatus, for example as illustrated in FIG. When the user designates the appreciation menu for the recorded song song through the user interface of the karaoke apparatus, the attribute value of each attribute item can be designated and input on the user interface. The screen transition of the portable computer as a user interface apparatus of a karaoke apparatus is illustrated in FIGS. Here, the top screen of the appreciation menu is configured so that “search by selecting music” and “search by feature” can be selected and input (see FIG. 3).

  When the user selects “Search by searching”, select the song you want to listen to by specifying the song name or by directly specifying the song ID by referring to the table of contents etc. according to the instructions on the user interface. (See FIG. 4). At this time, it is also possible to configure to select music by data communication with the music index database described above. When one song is selected, one song ID is designated.

  When the music ID is designated, the portable computer refers to the table of contents of the song recording database and extracts a record of the work related to the music ID. The number of corresponding works (the number of extracted records) is displayed on the user interface. If the number of works is less than or equal to a predetermined number, the works may be displayed in a list using the singing date or the name of the singer.

  The user can further specify and narrow down the features of the work he wants to listen to from the works singing the specified music. In other words, it is a feature of the present invention that the works can be narrowed down by specifying the voice quality and technical features of the singer. When the number of works is narrowed down by designating music, the screen is displayed so that the characteristics of the singer's voice quality and singing technique can be designated and input as illustrated in FIGS.

  While appropriately performing the operation of repeating the pages of the screen, the user designates the feature by selecting from among the items displayed appropriately or inputting numerical values. For example, when searching for a work with a somewhat feminine voice, 4 is designated and input for “masculine or feminine”. A plurality of attribute value designation inputs for the same item, such as 4 and 5, may be received and designated as 4 or more. When a user finishes specifying features and inputs a search, for example, by pointing to the “Search” button, the mobile computer extracts the corresponding records from the records previously extracted from the table of contents and sings the singing dates of those works. A list of works is displayed using a singer or a singer (see FIG. 7). If the number of works is greater than or equal to a predetermined number, it may be displayed over a plurality of pages so that the user can appropriately turn the page back and forth.

  When only one feature is specified, it is only necessary to specify and input an attribute value for only one corresponding attribute item as described above, and to specify a search without specifying other items. It is also possible to specify and input features for a plurality of attribute items. For example, in addition to the above-mentioned “masculine or feminine”, “2” indicating that the “high or low” is slightly high is designated and input, and then the “search” button is designated. It is possible to specify and input for all displayed items. The portable computer extracts records from which the “masculine or feminine” is 4 or 5 and the “high or low” is 2 from the records previously extracted from the table of contents, and displays the corresponding works in a list. .

  It is also possible to search again by returning to the screen for designating and inputting features as appropriate. If there are many items specified and the corresponding record becomes zero as a result of the search, the specified features may be reduced and the search may be performed again.

In addition, when inputting features, it is possible to select and specify "Check by voice" and "Check by singing", and if the user selects "Check by voice", it will be included in the table of contents of the song recording database Change to a screen (page) where you can input voice quality items among the attribute items, and select "Check by singing" to change to a screen where you can enter technical items among the attribute items included in the table of contents of the song recording database If it is configured so that it is easy to find the designated input location of the feature to be designated, it is convenient. Also in this case, it is configured so that all items relating to voice quality / techniques can be designated by appropriately performing an operation of turning the page forward or backward.
In addition, the singer's gender and age may be designated and input.

  When the user selects and designates one work out of works narrowed down and displayed as described above, the karaoke apparatus is notified of the identification information and storage location of the work. The karaoke machine orders the work and plays it back. For the work selected by the user, the table of contents information may be presented as appropriate, such as the attribute value of the item specified as the search condition.

  In the database search process, the search condition and the table of contents of the database are collated, the degree of coincidence and similarity are determined and digitized by a predetermined matching algorithm, and records that fall within a predetermined range are extracted. In the above, in order to simplify the description, a record that matches the search condition is extracted. When determining the similarity, a well-known data mining method is used to set the similarity between attribute values within the same attribute item in advance, or a correlation rule between multiple attribute items (high frequency For the combinations that occur, the degree of support and the certainty are defined in advance. According to such a similar search, for example, a combination of a plurality of items even if the attribute value of each item does not match the search condition so that there is a high correlation between sales of shampoo and rinse. Can be determined to be similar and extracted by comparing them in pairs.

=== How to search the database of singing song recordings part 2 ===
When the user designates “Search by feature” on the top screen of the appreciation menu (see FIG. 3), the tune of the piece he wants to listen to, the voice quality of the singer, and the features of the singer, without specifying the song being sung Can be specified to narrow down the works (see FIG. 8). In addition to the tune, it may be possible to specify the characteristics of the tune such as the theme and keywords of the tune. The search method after selecting a song when "Search by selecting songs" is specified is to specify the characteristics of the singer's voice quality and skill, except that instead of selecting a song, select and enter the tune, etc. The same applies to input.

  For example, when a user wants to cheer himself up by listening to a piece of dynamic music, he selects and designates “dynamic” for the music. Except that the tune tone can be selected and input, the specification input of the voice quality and skill of the singer is the same as in the case of the search method 1 described above. For example, if you want to appreciate a dynamic work with rich female voice according to the mood at that time, the singer's gender “female”, the tone “dynamic”, and the voice quality attribute item “Is there tension?” A plurality of features are designated and input, such as 1 indicating that there is tension and 5 indicating that the vibrato characteristic is very large. As described above, the user appropriately designates an attribute value for an attribute item corresponding to one or more features, and instructs the “search” button to input and search.

  The portable computer searches the singing song recording work database as in the case of the search method 1 described above, extracts the corresponding work records, and displays those works in a list. When the number of works is not less than a predetermined number and cannot be displayed on one page, it is displayed over a plurality of pages (see FIG. 9). When a user selects and designates one of the works that are narrowed down and displayed while performing page operations as appropriate, the items specified in the singer's name, song title, and other search conditions for the selected work Table of contents information, such as attribute values, is presented as appropriate. Then, the identification information and storage location of the work are notified to the karaoke apparatus. The karaoke machine orders the work and plays it back.

=== Application example of the singing song recording database ===
As described above, when the user wants to appreciate the work in accordance with the mood at that time, the desired song can be narrowed down by searching the database of the song recording work. When a song is selected from the narrowed songs on the user interface and specified, the work can be taken out and listened to.

  If you're looking for someone you want to duet, you can narrow down the singers and their works that match your voice. For example, it is possible to search for a work by specifying voice quality and skill, and to find a target singer from among the singer of the narrowed work. You can specify the tune and voice quality without selecting a song, narrow down the works of the same genre as the song you want to duet, find the works, and find those singers.

  The use of the song recording database is not limited to karaoke. The work database should be accessible via a public network such as the Internet. In this case, the work database may be disclosed for a fee. For example, even when trying to find a singer who wants to be scouted as a singer, the target singer can be found in the same manner as described above.

  You can also search for works and singers you want to use for various advertisements and commercial BGM. It is possible to specify the voice quality and tune according to the image of the advertisement, etc., narrow down the works and search for the desired work, and specify the singer.

It is the figure which illustrated the data structure of the song song recording work database which concerns on a present Example. It is the figure which illustrated the top menu displayed on the user interface apparatus screen of the song song recording work database which concerns on a present Example. It is the example of a screen at the time of the work appreciation menu selection displayed on the user interface device. It is an example of a screen at the time of selecting a “select and search for music” menu displayed on the user interface device. It is an example of a screen of feature input displayed on the user interface device. It is an example of a feature input screen (continued) displayed on the user interface device. It is the figure which illustrated the search result at the time of the "select music and search" menu displayed on the user interface device. It is the example of a screen of the feature input at the time of selecting the “Search by feature” menu displayed on the user interface device. It is the figure which illustrated the search result at the time of selecting the “Search by feature” menu displayed on the user interface device.

Explanation of symbols

1 Attribute item 2 Attribute value

Claims (1)

A karaoke song recording work appreciation system equipped with a music index database, a song recording work database, a user interface, and a control means,
The song index database describes each song played on the karaoke device in association with the song ID and the song attributes including the song title, artist name, and genre,
The singing song recording database stores karaoke accompaniment music performed on a karaoke device and karaoke song recordings recorded by singing along with the song ID in association with the work ID, and has a table of contents.
The table of contents includes a song ID, a singer attribute including a singer's name, a voice quality evaluation description of the singing, and a technique evaluation description of the singing for each karaoke song recording work,
The user interface displays information to the user according to the control of the control means, accepts the user's operation input,
The control means causes the user to perform music search based on the music index database via the user interface,
The control means performs the first process to the fourth process when the user specifies one music by music search,
The first process searches the table of contents of the song recording database based on the song ID of the identified song, and extracts the singer attributes, voice quality evaluation description, and skill evaluation description for a plurality of works associated with the song ID. And
The second process displays on the user interface the number of works extracted in the first process and attribute value options for each of a plurality of attribute items related to the singer attributes, voice quality evaluation description, and skill evaluation description. Accepts user selection input for,
In the third process, the user selection information received in the second process and the singer attributes, voice quality evaluation descriptions, and skill evaluation descriptions for the plurality of works extracted in the first process are collated using a matching algorithm, and the user selection information Select a karaoke song recording that is highly compatible with
The fourth process displays the singer name on the user interface for each work ID of the karaoke song recording work selected in the third process, and receives a user input for instructing the reproduction of the displayed work. A system for appreciating karaoke song recordings that extracts and reproduces the corresponding karaoke song recordings from the singing song recording database.

Images