JP2010237380A - Singing support system, support content estimation device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To specify whether it is suitable to support what kind of contents by which support medium for a singer. <P>SOLUTION: Time difference sequence between voice output timing and ideal timing is measured (delay measurement) for each support medium, by supporting measurement by each of a plurality of support media for an object singer. It is estimated that singing support by a predetermined support amount performed by a support medium in which a distribution degree of the time difference sequence is less than a threshold, or the smallest, is a support content suitable for the object singer. When singing support is requested in singing of a musical piece from a specified singer, a support content corresponding to singing frequency of the musical piece by the request is extracted, and singing support by this support content, that is, singing support with the support content in which the singer can stably sing with a fixed time difference, is performed. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention is a singing support system that performs singing support so that the timing at which a singer sings each singing sound can be matched with the timing at which each singing sound is uttered when the song is properly sung. About.

  In recent years, as a technique for supporting singing as described above, a method of adjusting the display timing and display mode of characters by using a telop (see Patent Documents 1 to 3), a volume of a guide melody (see Patent Document 4), and the like. Has been proposed.

JP-A-2005-173256 Japanese Patent Laid-Open No. 10-039882 JP2007-292922 JP-A-2005-148627

  However, the above-mentioned method can specify what kind of support is suitable for a single support medium used for support, such as telop or guide melody. It cannot be specified what kind of support is appropriate for the medium.

  The present invention has been made to solve such a problem, and its purpose is to identify what kind of support medium is suitable for a singer by using which support medium. Is to provide technology for.

  In order to solve the above-mentioned problem, the first invention is that when a singer sings each singing sound when singing (hereinafter referred to as “uttering timing”), each singing sound is uttered when the song is appropriately sung. One or more support media used at the time of support, and one or more support amounts indicating the degree of support by the support media, as support content to be matched with timing (hereinafter referred to as “ideal timing”) It is a singing support system that can support singing with multiple types of support content combining any of the above, and an information collection unit that collects information about singing by a specific singer, and estimation of support content suitable for the singer A content estimation unit, and a singing support unit that performs singing support based on the support content estimated by the content estimation unit.

  Among these, each of the one or more support media is configured so that the information collecting unit sings a song (hereinafter referred to as “target song”) by a specific singer (hereinafter referred to as “target song”). Measurement support means for performing singing support for measurement (hereinafter referred to as “measurement support”) that is performed by changing the amount of support by the support medium repeatedly for a predetermined measurement period for each support medium; In each measurement period in which measurement support is performed by the measurement support means, each utterance timing by the target singer is compared with a corresponding ideal timing among the ideal timings in the target music, and the utterance Acquire a time difference series consisting of each of the time differences in a partial period when the time difference between the timing and the ideal timing converges within a predetermined range. Measurement information storage for storing measurement information including a time acquisition sequence acquired by the sequence acquisition unit, the time difference sequence acquired by the sequence acquisition unit, and the support content of the measurement support performed at the time of acquisition in association with the target singer Means.

  In addition, the content estimation unit groups the measurement information of the same target singer stored in the information collection unit for each support medium that defines the support content, and the dispersion of the time difference series in the measurement information of each group A variance calculating unit for calculating the degree, and a group of the measurement information in which the degree of dispersion calculated by the variance calculating unit is less than a predetermined threshold or the smallest, corresponding to the group Support estimation means for estimating the support by the support medium as support contents suitable for the target singer who has been calculated by the variance calculation means, and the support contents estimated by the support estimation means are associated with the target singer. Support content storing means for storing the information.

  Then, the singing support unit is a support corresponding to the singer according to the request among the support contents stored by the content estimation unit when a specific singer requests a singing support when singing a song. Content extraction means for extracting content, and singing support means for performing singing support based on the support content extracted by the content extraction means.

  With this configuration, the time difference series between the utterance timing and the ideal timing is measured for each support medium by performing measurement support for each of the support media for the target singer, and the degree of dispersion of the time difference series is determined. The singing support by the support medium that is less than the threshold or the smallest can be estimated as the support content suitable for the target singer.

  Singing support whose degree of variance of the time difference series is less than the threshold or the smallest means that the time difference has converged within a narrow range and has been supported in a stable manner with a certain time difference. Yes. In other words, it can be said that the support content that can be stably sung is estimated here.

  When there is a request for singing support from a specific singer when singing a song, the support content corresponding to the singer is extracted, and the singing support by this support content, that is, the singer is stable at a certain time difference. Singing support can be carried out with the support content that can be sung in an effort.

  In the present invention, the support medium is “to change the character display color” in terms of the support content for adjusting the timing for changing the character display color in the telop. In this case, the change timing is “the amount of support”. Is. The support content for adjusting the character display amount (number of characters) in the telop is “changing the character display amount”. In this case, the display amount is the “support amount”. The support content for changing the character display mode (size, etc.) in the telop is “changing the character display mode”. In this case, the display mode is “support amount”. Further, the support content for adjusting the volume of the guide melody in accordance with the ideal timing is “changing the volume of the guide melody”. In this case, the volume is the “support amount”. The support content for adjusting the performance sound (for example, bass sound) to notify the ideal timing is “adjusting the performance sound”. In this case, the degree of adjustment of the performance sound (the volume of the sound) Is the “amount of support”.

  In addition, the measurement support in the above invention is repeatedly performed for each support medium for each measurement period. For example, a part of the period during which the same music is sung may be performed as a “measurement period”. Moreover, you may implement the whole period when a music is sung as a "measurement period."

  In addition, in order to repeat measurement support, for example, if a part of the period during which the same music is sung is a “measurement period”, the target singer is singing, and The measurement support may be performed by a support medium different from that which has been performed so far, provided that the time corresponding to the measurement period remains.

  In the above configuration, the time difference obtained by comparing the utterance timing and the ideal timing may be specified as follows. First, when the singer sings a song, each singing sound is a singing section, and the transition pattern of the pitch in the singing section is a section where each singing sound is uttered when the same song is sung appropriately. Contrast with the transition pattern of the pitch in the guide section corresponding to the corresponding singing section in the guide section. And the position on the time axis in the singing section is corrected (along the time axis) until the transition pattern of the pitch in the singing section is the closest to the transition pattern of the pitch in the guide section (the degree of approximation is maximized). The amount of correction when shifting back and forth is specified as the time difference described above.

  Any method may be adopted as a method for correcting the position on the time axis so that the transition pattern is approximated in this way, for example, as described in JP-A-2005-107330. It is conceivable to adopt such a method.

  In addition, the condition for acquiring the time difference series composed of the time differences thus identified is that the time difference has converged within a predetermined range. For example, each time difference specified during measurement support is only within a certain range. A state in which it does not fluctuate or the time difference calculated during measurement support continuously becomes a small value within a certain range.

  Moreover, in the said structure, when the request | requirement of the song of a music was received from the singer, the singer assigned this identification information was specified by acquiring the identification information assigned to the singer related to the request. In addition, the measurement information and the support content should be stored with this singer as the target singer. In addition, when receiving a request for singing support when singing a song, by obtaining the identification information of the singer, the singer to whom this identification information is assigned is specified, and this singer is supported. It is recommended to extract support contents.

  In the above configuration, when a specific singer requests singing support when singing a song, the support contents corresponding to the singer include not only a specific support medium but also support by the support medium. It is desirable that the support amount indicating the degree can be extracted.

As a configuration for this purpose, the first configuration is desirably a second configuration (claim 2) shown below.
In the information collection unit having the above configuration, the measurement information storage unit includes, as the support content of the measurement support performed when acquiring the time difference series, a support medium for measurement support performed when acquiring the time difference series, and The support amount of the measurement support performed during a partial period corresponding to the time difference series is stored in association with the target singer.

  In the content estimation unit, the support estimation unit is configured to measure the group of measurement information in which the degree of dispersion calculated by the dispersion calculation unit is less than a predetermined threshold value or the smallest. The support based on the support medium corresponding to the group and the support amount in the measurement information of the group is estimated as the support content suitable for the target singer.

  With this configuration, singing support with a predetermined amount of support carried out with a support medium in which the degree of variance of the time difference series is less than the threshold value or the smallest is suitable for the target singer. It can be estimated as content. And when there is a request for singing support when a specific singer sings a piece of music, the support content corresponding to that singer is extracted, and this support content sings with a specific support amount by a specific support medium Support can be implemented.

  In this configuration, the support amount stored in association with the coping singer may be the support amount at the time when the time difference has converged, among the support amounts of the measurement support performed during a partial period, or a part thereof The average amount of support for the entire period may be used.

  By the way, in general, the more the number of times a singer sings the same piece of music, the more the singing of that piece of music will improve. It is also possible that singing support by content is more suitable for singing at the ideal timing. As a specific example, for example, as a result of memorizing lyrics and singing timing by repeatedly singing the same music, it is suitable for singing at the same time to use text display by telop as a support medium until then. However, there are cases where it is easier to synchronize the timing with the volume of the guide melody as a support medium.

  Therefore, when estimating the support content suitable for the target singer, it is desirable to estimate the singing frequency of the same music. As a configuration for this purpose, for example, it is conceivable to make each of the above configurations as a third configuration (claim 3) shown below.

  In the information collecting unit configured as described above, the measurement information storage unit includes measurement information including the time difference sequence acquired by the sequence acquisition unit, the support content of the measurement support performed at the time of acquisition, and the target music piece. Is stored in association with the target singer.

  Moreover, in the said content estimation part, the said dispersion | distribution calculation means consists of the measurement information of the same object singer preserve | saved in the said information collection part, and when the said object music is a music sung by the object singer. Each of the large groups of measurement information according to the singing frequency (or the number of singings) is further divided into small groups of measurement information in which the support medium that defines the support content is the same. The degree of dispersion is calculated, and the support estimation means has a degree of dispersion calculated by the dispersion calculation means less than a predetermined threshold for each large group of measurement information according to the singing frequency, or The small group of the measurement information that is the smallest is identified, and the support medium corresponding to the identified small group is supported by the large group to which the small group belongs. Estimating the content of the singing frequency corresponding to the loop as the support content suitable when the target singer sings, the support content storage means, the support content for each singing frequency estimated by the support estimation means Save it in association with the target singer.

  In addition, in the singing support unit, when there is a request for singing support from a specific singer when singing a song, as the measurement information stored in the information collecting unit in association with the singer, the music according to the request The frequency specifying means for specifying the singing frequency of the music related to the request is provided depending on how many pieces of measurement information that are the same target music as the above are included. And when the content extraction means is a request for singing support when singing a song from a specific singer, among the support contents stored by the content estimation unit, the singer according to the request, and the The support content corresponding to the singing frequency specified by the frequency specifying means is extracted.

  With this configuration, for each singing frequency, the singing support by the supporting medium in which the degree of variance of the time difference series is less than the threshold or the smallest is estimated as the supporting content suitable for the target singer. can do.

  And when there is a request for singing support from a specific singer when singing a song, the support content corresponding to the singer and corresponding to the singing frequency of the song related to the request is extracted, and this support Singing support according to the contents, that is, the supporting contents that the singer can sing stably with a certain time difference according to the singing frequency can be implemented.

  Further, in each of the above configurations, a specific method for calculating the degree of dispersion is not particularly limited. For example, a method like the fourth configuration (claim 4) shown below can be considered.

  In the content estimation unit configured as described above, the variance calculating unit calculates the degree of variance according to Equation 1 below, where dti is the time difference sequence in the measurement information of the target group.

If it is this structure, based on the dispersion | distribution degree computed by the said Formula 1, the support content suitable for the object singer can be estimated.

In order to solve the above-described problem, a support content estimation apparatus (claim 5) including all the means in any one of the first to fourth configurations may be provided.
If it is this structure, the effect | action and effect similar to the said any one of the said 1st to 4th structure can be acquired.

  In order to solve the above-mentioned problem, a program (Claim 6) for causing a computer system to execute various processing procedures for causing all the means in any one of the first to fourth configurations to function. Good.

If it is a computer system that executes this program, it can constitute a part of the singing support system according to any of the first to fourth aspects.
The above-mentioned program is composed of an ordered sequence of instructions suitable for processing by a computer system, and uses a singing support system, a singing content estimation device, and the like via various recording media and communication lines. It is provided to the user who performs.

Block diagram showing the overall configuration of the karaoke system Flow chart showing information collection process Diagram showing the operation of the karaoke system Flow chart showing singing support processing Flow chart showing support content identification processing Diagram showing how information is registered in the measurement table The figure which shows a mode that each information of measurement information is registered into a support content table Flow chart showing support content notification processing

Embodiments of the present invention will be described below with reference to the drawings.
(1) Overall Configuration The karaoke system 1 is configured such that a server 2 composed of a well-known computer system and each of one or more karaoke apparatuses 3 are communicably connected via a network 100. In addition, this karaoke system 1 functions as a singing support system in the present invention.

  The server 2 includes a control unit 21 that controls the entire server, a storage unit 23 that stores various information, a communication unit 25 that controls communication via the network 100, and a user interface (U / I) unit 27 including a keyboard and a display. , A media drive 29 for inputting / outputting information via a recording medium.

  The karaoke device 3 includes a control unit 31 that controls the entire device, a storage unit 33 that stores music data and video data indicating the accompaniment content and lyrics of the performance music, a communication unit 35 that controls communication via the network 100, and the like. A display unit 41 for displaying video, an operation unit 43 including a plurality of keys and switches, a voice input / output unit 49 for controlling voice input from the microphone 45 and voice output from the speaker 47, and the like are provided. .

  In this karaoke system 1, the timing at which a singer utters each singing sound when singing (hereinafter referred to as “voice timing”), and the timing at which each singing sound is uttered when the song is properly sung (hereinafter “ideal timing”). As a support content to be adapted to the above), any one of a plurality of support media used for support and one or more support amounts indicating the degree of support by the support media are respectively combined. The singing support is implemented by a plurality of types of support contents.

The “support medium” here refers to support contents for adjusting the timing for changing the character display color in the telop, which is “to change the character display color”. In this case, the change timing is “ Amount of support ". The support content for adjusting the character display amount (number of characters) in the telop is “changing the character display amount”. In this case, the display amount is the “support amount”. The support content for changing the character display mode (size, etc.) in the telop is “changing the character display mode”. In this case, the display mode is “support amount”. Further, the support content for adjusting the volume of the guide melody in accordance with the ideal timing is “changing the volume of the guide melody”. In this case, the volume is the “support amount”. The support content for adjusting the performance sound (for example, bass sound) to notify the ideal timing is “adjusting the performance sound”. In this case, the degree of adjustment of the performance sound (the volume of the sound) ”) Is the“ support amount ”.
(2) Processing by Karaoke Device 3 Hereinafter, various processing procedures started when the control unit 31 of the karaoke device 3 executes a program stored in the storage unit 33 will be described.
(2-1) Information Collection Processing First, when the singing of a song is performed by a singer in a state where the collection mode for collecting information in the karaoke apparatus 3 is set to “ON”, the information is repeatedly executed. A processing procedure of the information collection processing will be described with reference to FIG. Here, “the song is being sung by the singer” is determined based on the fact that the music is being played and the period during which the song is to be performed has not ended.

  When this information collection process is started, first, a singer who performs singing (hereinafter referred to as “target singer”) is specified (s110). In this embodiment, since it is comprised so that the user of the karaoke apparatus 3 can select the music which should be sung after performing the login operation with respect to the karaoke apparatus 3, here, the music in performance (henceforth "target music"). The user who selected (the user with the user ID accompanying the login operation) is identified as the target singer.

  Next, measurement support using the i-th support medium is selected from a plurality of types of measurement support as singing support for measurement (hereinafter referred to as “measurement support”) performed in the subsequent processing (s120). In this embodiment, since a variable i for setting a value i indicating the order of support media used for measurement support is prepared for each user ID, here, the user specified in the above s110 Measurement support by the i-th support medium is selected based on the value of the variable i corresponding to the ID. The variable i is generated with the initial value “1” being set when the user ID is registered. The variable i is sequentially incremented to the upper limit value in the process described later, and then reset to the initial value “1”. is there.

  Next, measurement support using the i-th support medium selected in s120 is started (s130). Here, singing support for measurement is started, in which singing support is performed while sequentially changing the support amount in the i-th support medium to each of Xi [1] to Xi [n] (n: any number). (See “Measurement Support” in FIG. 3A).

  Next, the time difference between the utterance timing and the ideal timing is sequentially calculated by comparing each utterance timing of the singing sound by the target singer with the ideal timing of the corresponding singing sound among the ideal timings in the target music. Processing (time difference series calculation processing) is started (s140). With this process, a time difference sequence including a time difference between the utterance timing and the ideal timing is generated as the measurement support progresses (see “delay measurement” in FIG. 3A).

  In this time difference series calculation process, first, when a singer sings each song sound when singing a song, the song segment is set as a song segment, and the pitch transition pattern in the song segment is sung when the same song is properly sung. Each of these is compared with the transition pattern of the pitch in the guide section corresponding to the corresponding singing section among the guide sections which are sections to be uttered. And the position on the time axis in the singing section is corrected (along the time axis) until the transition pattern of the pitch in the singing section is the closest to the transition pattern of the pitch in the guide section (the degree of approximation is maximized). The correction amount in the case of shifting back and forth is calculated as the time difference described above. At this time, any method may be employed as a method of correcting the position on the time axis so that the transition pattern is approximated. For example, a method described in JP-A-2005-107330 is used. Can be considered.

  Next, it is checked whether or not an end condition FNi for ending the i-th measurement support is satisfied (s150). Here, the time difference calculated in the process of changing the support amount in measurement support has converged to a predetermined range, or the i-th measurement support has ended (for example, the time required for changing to all support amounts) It is determined that the end condition FNi is satisfied.

  The above-mentioned “when the time difference converges within a predetermined range” means that, for example, each time difference calculated during measurement support varies only within a certain range, or the time difference calculated during measurement support continues. In other words, it may be a small value within a certain range.

  If it is determined in s150 that the termination condition is not satisfied (s150: NO), it is checked whether or not the singing of the target song by the target singer has ended (s160). Here, it is determined that the singing of the target music by the target singer has ended when the performance of the target music has ended.

If it is determined in s160 that the singing by the target singer has not ended (s160: NO), the process returns to s150.
Moreover, when it determines with completion | finish conditions being satisfy | filled by said s150 (s150: YES), or when it is determined that the singing by an object singer is complete | finished by said s160 (s160: YES), the above The measurement support started in s130 is ended (s170).

  Next, among the time differences calculated in the measurement support carried out so far, a time difference series dti [k] (k is an ideal timing) consisting of a predetermined number of time retroactive from the time point when the measurement support is ended in s170. , K = 1, 2,...) Is acquired (s180). In this s180 (especially, when “YES” is determined in s160), the time difference series dti is not calculated as a time difference series dti consisting of a predetermined number of time differences retroactive from the time point when the measurement support is ended. Among the time differences, a time difference sequence in a section where the average of a predetermined number of time differences is smallest may be acquired.

  In s180 in the case where “YES” is determined in s160, the time difference calculated in the time difference series calculation process is not a time difference series dti consisting of a predetermined number of time differences retroactive from the time point when the measurement support is ended. Among them, it is possible to acquire a time difference series in a section where the average of a predetermined number of time differences is the smallest.

  Next, the time difference series dti [k] acquired in s180, the support content (support medium i and support amount Xi [j]) at the time when measurement support is ended in s170, and the target music (assigned to Measurement information consisting of the song number TR_N) and the target singer (identification information SI assigned to) is generated and transmitted to the server 2 (s190) (see FIG. 3A).

  Here, the support amount X [j] that defines the support content is calculated by a feedback function FBi represented by the following Equation 1.

In the server 2 that has received the measurement information, the support content suitable for the target singer is specified and stored in the support content specifying process described later.

  Next, it is checked whether or not the measurement support by all the support media is completed (s200). Here, when the value of the variable i is the upper limit value, it is determined that the measurement support by all the support media is completed.

  If it is determined in s200 that measurement support by all support media has not been completed (s190: NO), the variable i described above is incremented (i + 1 → i) (s210), and then this information collection process Ends.

  On the other hand, when it is determined that the measurement support by all the support media has been completed (s200: YES), after the variable i described above is initialized (1 → i) (s220), this information collection process Ends.

Thus, after the information collection process is completed, the information collection process is executed again based on the value of the variable i changed in s210 or s220.
(2-2) Singing Support Processing Subsequently, in the state where the support mode for performing singing support in the karaoke apparatus 3 is set to “ON”, the singing support process is performed at the timing when the singing of the music by the singer is started. The singing support process procedure will be described with reference to FIG.

  When this singing support process is activated, first, a singer who starts singing (hereinafter referred to as “target singer”) is specified (s310). Here, as in the information collection process described above, the user (user ID associated with the login operation) who has selected the music to be played next (hereinafter referred to as “target music”) is specified as the target singer.

  Next, a request command to notify the support content suitable for the target singer identified in s310 is transmitted to the server 2 (s320). Here, the request command is transmitted to the server 2 together with the user ID of the target singer identified in s310 and the song number of the target song (see “request command” in FIG. 3B). From the server 2 that has received this request command, a request indicating support contents (support medium iopt and support amount Xopt_i) suitable for the target singer identified by this user ID based on the user ID transmitted together with the request command. A response is returned (see “request response” in FIG. 3B).

  Thus, after the request command is transmitted in the above s320, when the request response from the server 2 is received (s330), the support content indicated by this request response is determined as the support content when the target singer sings the target music. (S340).

Next, along with the performance of the target music, singing support based on the support content determined in s340 is started (s350) (see FIG. 3B).
Then, after the performance of the target music is ended (s360: NO), the performance of the target music is completed (s360: YES), and the singing support started by s350 is completed (s370). ) Thereafter, the singing support process ends.
(3) Processing by Server 2 Hereinafter, various processing procedures started when the control unit 21 of the server 2 executes a program stored in the storage unit 23 will be described.
(3-1) Support Content Specifying Process First, the processing procedure of the support content specifying process started every time the server 2 receives measurement information from the karaoke apparatus 3 will be described with reference to FIG. This measurement information is information that the karaoke apparatus 3 transmits in s190 of the information collection process, and includes a time difference series dti [k], support contents (support medium i and support amount Xi [j]), and target music (music) Number TR_N) and information consisting of a singer (identification information SI) (see FIG. 3A).

  When this support content specifying process is activated, first, the measurement information received prior to the activation is registered in a data table (measurement table) for registering measurement information (s410). Here, the target music (music number TR_N), the time difference series dti [k] in the measurement information, and the support content (support medium i and support amount Xi [j]) correspond to the singer (identification information SI). Information is registered in the measurement table (see FIG. 6).

  Next, one or more pieces of measurement information corresponding to the singer registered in s410 (hereinafter referred to as “target singer”) are extracted from the measurement information registered in the measurement table at this time (s420). ).

  Next, each of the measurement information extracted in s420 is divided into large groups of measurement information according to the singing frequency (or the number of singing times) of the target music, and each large group further supports the support contents. The medium i is divided into small groups of measurement information that are the same (s430).

  Here, first, it is assumed that the target singer has sung the target music included in the measurement information, and the singing frequency is assigned to each measurement information when the number of measurement information that is the same target music is the number of times of singing. (See FIG. 7 (a)). The singing frequency is, for example, singing frequency 0 to singing frequency 0 to singing frequency D, singing frequency 3 to 5 times, singing frequency C, singing frequency 6 to 9 times, singing frequency B, singing frequency 10 times or more. It is expressed as singing frequency A or the like. Then, after each measurement information to which the singing frequency is assigned is divided into large groups corresponding to the singing frequencies, each of the measurement information belonging to the large group is a small group corresponding to each support medium i that defines the support content. (See FIG. 7B).

  Next, the degree of variance of the time difference series dti [k] is calculated for each measurement information of the small group grouped in s430, and this is averaged for each support medium i that defines the support content in the same small group. (S440). Here, the degree of dispersion vdti of each measurement information belonging to the small group is calculated by the following equation 2 based on the time difference sequence dti [k] of the measurement information, and these are averaged for each support medium i in the same small group. Accordingly, the degree of dispersion for each support medium i is obtained for each singing frequency (see FIG. 7C).

Note that “m” in Equation 2 is an index corresponding to the ideal timing, similarly to k.

  Next, based on the degree of dispersion vdti for each support medium i obtained for each singing frequency in s440, a support medium iopt that defines support content suitable for the target singer is estimated for each singing frequency ( s450). Here, a support medium iopt having a degree of variance vdti that is minimum or less than a predetermined threshold is specified for each singing frequency, and when the supporting medium iopt sings the target music of the corresponding singing frequency, the target singer It is presumed that this is a support medium that defines the support content suitable for.

  As described above, the singing support by the support medium i having the smallest dispersion degree vdti or less than the threshold value converges in a narrow range of the time difference of the time difference series dti [k], and is stable at a constant time difference. It means that it was support contents that could sing. Therefore, it can be said that the support medium iopt with the degree of dispersion vdti being the minimum or less than the predetermined threshold is a support medium that defines the support contents that the singer can sing stably at a certain time difference.

  Next, for each support medium iopt for each singing frequency specified in s450, the support amount Xopt_i [j] for performing the singing support by the support medium iopt is specified (s460). Here, the support amount Xi [j] that defines the support content referred to when specifying the support medium iopt in the above s450, that is, the support amount of the support medium that defines the support content suitable for the target singer is the support medium. It is specified as a support amount Xopt_i [j] when performing singing support by iopt. When there are a plurality of support amounts Xi [j] that define the referred support contents, one of the support amounts or the average support amount is set as the support amount Xopt_i [j]. It should be specified.

  This will be described with reference to FIG. 7. When the support medium 4 is specified as the support medium iopt for the singing frequency A (see FIG. 7C), the reference is made when the support medium 4 is specified. The support amount X4 [9] that defines the supported content (support medium 4, support amount X4 [9]) is specified as the support amount Xopt_i [j].

  And the support content consisting of the support medium iopt and the support amount Xopt_i [j] for each singing frequency specified by the above s450 and s460 is associated with the target singer in the data table for support content registration (support content table). It is registered (s470).

  Since this support content table is prepared for each user ID, the support content table corresponding to the user ID in the measurement information received prior to the start of the support content specifying process corresponds to each singing frequency. Is registered as the support content to be performed (see FIG. 7D). In addition, when the support content corresponding to the singing frequency is already registered in the corresponding support content table, this support content is overwritten.

After finishing s470 in this way, this support content specific process is complete | finished.
(3-2) Support Content Notification Process Next, the processing procedure of the support content notification process that is activated in response to a request command to notify the support content from the karaoke apparatus 3 will be described with reference to FIG. This request command is transmitted by the karaoke apparatus 3 at s320 of the singing support process, and is transmitted together with the user ID of the singer and the music number of the target music (FIG. 3 (b) “Request command”. reference).

  When this support content notification process is activated, first, the frequency of singing the song with the same song number (singing frequency) is specified by the singer of the user ID received prior to the activation (s510). .

  Here, first, among the measurement information registered in the measurement table corresponding to the user ID at this time, each measurement information corresponding to the singer of the user ID received prior to the start of the support content notification process is extracted. Is done. Subsequently, it is checked how many target music pieces having the same music number as the received music number are included in the target music pieces of the measurement information thus extracted. And the singing frequency according to the number of the object music is allocated similarly to the support content specific process mentioned above, and this is specified as the singing frequency in the music of the music number which this received.

  Next, the support content suitable for singing the music with the singing frequency specified in s510 is extracted (s520). Here, first, among the support contents (support medium iopt and support amount Xopt_i) registered in the support contents table corresponding to the user ID at this time, the support contents corresponding to the singing frequency specified in s510 above. Is extracted.

Then, after the request response indicating the support content extracted in s520 is returned to the karaoke apparatus 3 that is the transmission source of the request command (s530), the support content notification processing is ended.
This request response is received by the karaoke apparatus 3 at s330 of the singing support process (see “request command” in FIG. 3B).
(4) Action and Effect In the karaoke system 1 configured as described above, the time difference series between the utterance timing and the ideal timing is measured for each support medium by performing measurement support for each of the target singers using a plurality of support media. (Single information collection process in FIG. 2), subject to singing support with a predetermined support amount implemented with a support medium in which the degree of variance of the time difference series is less than the threshold or the smallest It can be estimated as the support content suitable for the singer (s450, s460 in FIG. 5).

  When there is a request for singing support from a specific singer when singing a piece of music, the support content corresponding to the singing frequency of the music related to the request is extracted (s510, s520 in FIG. 8), and this support content. Singing support can be performed with the support content that allows the singer to sing stably at a certain time difference (s530 in FIG. 4 and s340 and s350 in FIG. 4).

At this time, the support content suitable for the target singer can be estimated based on the degree of dispersion calculated by Equation 2 above.
Thus, when estimating the support content suitable for the target singer, the estimation is performed in consideration of the singing frequency of the same music.

  This is because, as the number of times the same song is sung increases, the singing of that song will improve, so it depends on other support contents than the support support provided so far This is because it is assumed that the singing support is more suitable for singing at the ideal timing.

  As a specific example, for example, as a result of memorizing lyrics and singing timing by repeatedly singing the same music, it is suitable for singing at the same time to use text display by telop as a support medium until then. However, there are cases where it is easier to synchronize the timing with the volume of the guide melody as a support medium.

That is, as in the above embodiment, if the support content is estimated after considering the singing frequency of the same music, the support content according to the progress of the singer is set as the support content suitable for the singer. Can be estimated.
(5) Modifications Embodiments of the present invention have been described above, but the present invention is not limited to the above-described embodiments, and can take various forms as long as they belong to the technical scope of the present invention. Needless to say.

  For example, in the above embodiment, the karaoke apparatus 3 and the server 2 are configured to operate in cooperation with each other, but all the processes are performed only on the karaoke apparatus 3 side. Moreover, you may comprise so that each process implemented by the karaoke apparatus 3 may be implemented by another apparatus.

  Moreover, in the said embodiment, the information collection process by the karaoke apparatus 3 is comprised so that the performance of a music is performed and it may be repeatedly performed in the condition where the period which should sing in the music in performance has not ended. ing. That is, a part of the period during which the same music is sung is a measurement period by measurement support. However, the information collection process may be repeated each time a song is sung by a singer so that the entire period in which the song is sung becomes a measurement period by measurement support.

Moreover, in the said embodiment, it is comprised so that the assistance content specific process by the server 2 may be started whenever the measurement information from the karaoke apparatus 3 is received. However, this support content specifying process may be configured to be started for a certain period or when a predetermined event occurs regardless of reception of the measurement information. In this case, starting from s420 without performing s410, the measurement information corresponding to any one of the singers is extracted in s420, and after s470, after s420, about another singer (unprocessed singer) It should be done.
(6) Correspondence with the present invention In the embodiment described above, the control unit 31 of the karaoke apparatus 3 that executes the information collecting process is the information collecting unit in the present invention, and controls the server 2 that executes the support content specifying process. The unit 21 is a content estimation unit in the present invention, and the control unit 31 of the karaoke apparatus that executes the singing support process and the control unit 21 of the server 2 that executes the support content notification process are the singing support unit in the present invention.

  2 corresponds to the measurement support means in the present invention, s180 in FIG. 2 corresponds to the sequence acquisition means in the present invention, and s190 in FIG. 5 and s410 in FIG. 5 correspond to the measurement information storage means in the present invention. S440 corresponds to the variance calculation means in the present invention, s450 and s460 correspond to the support estimation means in the present invention, s470 corresponds to the support content storage means in the present invention, and FIG. S520 and s530 correspond to content extracting means in the present invention, s340 and s350 in FIG. 4 correspond to singing support means in the present invention, and s510 in FIG. 8 corresponds to frequency specifying means in the present invention.

  The server 2 that executes the support content specifying process corresponds to the support content estimation apparatus according to the present invention.

  DESCRIPTION OF SYMBOLS 1 ... Karaoke system, 2 ... Server, 21 ... Control part, 23 ... Storage part, 25 ... Communication part, 27 ... User interface part, 29 ... Media drive, 3 ... Karaoke apparatus, 31 ... Control part, 33 ... Storage part, 35 ... Communication unit, 41 ... Display unit, 43 ... Operation unit, 45 ... Microphone, 47 ... Speaker, 49 ... Audio input / output unit, 100 ... Network.

Claims (6)

The timing at which the singer utters each singing sound (hereinafter referred to as “speech timing”) when singing is matched with the timing (hereinafter referred to as “ideal timing”) when each singing sound is sung when the song is properly sung As a support content for doing so, a plurality of types of support in which any one or more support media used for support and one or more support amounts indicating the degree of support by the support media are combined. A singing support system capable of supporting singing by content,
An information collection unit that collects information related to singing by a specific singer, a content estimation unit that estimates support content suitable for the singer, a singing support unit that performs singing support based on the support content estimated by the content estimation unit, and Consists of
The information collecting unit
In a state in which a song (hereinafter referred to as “target song”) by a specific singer (hereinafter referred to as “target song”) is being sung, it is singing support by each of the one or more support media. Measurement support means for repeatedly performing singing support for measurement (hereinafter referred to as “measurement support”) performed while changing the support amount, for each of the support media, for a predetermined measurement period;
For each measurement period in which measurement support is performed by the measurement support means, each utterance timing by the target singer is compared with a corresponding ideal timing among the ideal timings in the target music, and the utterance timing And a sequence acquisition means for acquiring a time difference sequence consisting of each of the time differences in a partial period in which the time difference between the ideal timing and the ideal timing has converged within a predetermined range;
Measurement information storage means for storing measurement information consisting of the time difference series acquired by the series acquisition means and the support contents of the measurement support performed at the time of acquisition in association with the target singer. And
The content estimation unit
Dispersion calculation means for grouping the measurement information of the same target singer stored in the information collection unit for each support medium that defines the support content, and calculating a degree of dispersion for a time difference series in the measurement information of each group; ,
For the group of measurement information whose degree of variance calculated by the variance calculation means is less than a predetermined threshold value or the smallest, the support by the support medium corresponding to the group is used for the variance calculation. Support estimation means for estimating the support content suitable for the target singer who is the target of calculation by the means;
Support content storing means for storing the support content estimated by the support estimation means in association with the target singer, and
The singing support section
Content extraction means for extracting the support content corresponding to the singer according to the request out of the support content stored by the content estimation unit when there is a request for singing support when singing a song from a specific singer ,
A singing support means for carrying out singing support based on the support content extracted by the content extraction means. In the information collecting unit,
The measurement information storage means includes, as the support content of the measurement support performed when acquiring the time difference series, a support medium for measurement support performed when acquiring the time difference series, and a partial period corresponding to the time difference series Is stored in association with the target singer,
In the content estimation unit,
The support estimation means includes a support medium corresponding to the group of measurement information in which the degree of dispersion calculated by the dispersion calculation means is less than a predetermined threshold value or the smallest. And the assistance by the assistance amount in the measurement information of this group is estimated as the assistance content suitable for the said target song person. The singing assistance system of Claim 1 characterized by the above-mentioned. In the information collecting unit,
The measurement information storage means associates the measurement information including the time difference series acquired by the series acquisition means, the support content of the measurement support performed at the time of acquisition, and the target music with the target singer. Save
In the content estimation unit,
The variance calculation means includes measurement information of the same target singer stored in the information collecting unit, and the singing frequency (or the number of singing times) when the target music is a music sung by the target singer. Each of the corresponding large groups of measurement information is further divided into small groups of measurement information in which the support medium that defines the support content is the same, and the degree of variance of the time difference series is calculated for each of the small groups,
For each large group of measurement information according to the singing frequency, the support estimation unit has a degree of variance calculated by the variance calculation unit that is less than a predetermined threshold or the smallest Identifies a small group of measurement information, supports by a support medium corresponding to the identified small group, and supports suitable for the target singer singing a song having a singing frequency corresponding to the large group to which the small group belongs Estimating as content,
The support content storage means stores the support content for each singing frequency estimated by the support estimation means in association with the target singer,
In the singing support section,
When there is a request for singing support when a specific singer sings a song, the measurement information stored in the information collecting unit in association with the singer becomes the same target song as the song according to the request. Frequency identification means for identifying the frequency of singing the music according to the request, depending on how many pieces of measurement information are included.
The content extracting means includes a request for singing support for a song sung by a specific singer, among the support content stored by the content estimation unit, and the frequency specifying The support content corresponding to the singing frequency specified by the means is extracted. The singing support system according to claim 1 or 2. In the content estimation unit,
4. The variance calculation unit according to claim 1, wherein the variance calculation unit calculates a variance degree according to the following equation 1 when dti is a time difference sequence in the measurement information of a target group. The singing support system described in 1.
(K and m are indexes corresponding to ideal timing) All the means with which the content estimation part in any one of Claim 1 to 4 is provided. The assistance content estimation apparatus characterized by the above-mentioned.   The program for making a computer system perform the various process procedures for functioning as all the means with which the content estimation part in any one of Claim 1 to 4 is provided.