JP2016009153A - Karaoke device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a karaoke device which gives a pattern voice data using a voice of a singer himself/herself that is recorded by one time singing so that the singer can sing in his/her voice range.SOLUTION: A karaoke device comprises: determination means 30 which detects singing pitch of singing voice data for each singing section of a music, compares the singing pitch with the reference pitch to calculate a pitch difference, and verifies that the pitch difference is within an allowable range to determine whether the singing pitch for each singing section is required to be corrected or not; correction means 31 which creates the standard pattern voice data in which the singing pitch required to be corrected is corrected based on the reference data; identification means 32 which identifies, when the pitch of the singing voice data corresponding to the highest reference pitch is required to be corrected, the singing upper limit pitch where a singer can sing; and transposition means 33 which transposes all singing sections of the standard pattern voice data into lower pitch by a semitone unit transposition quantity on the basis of the pitch difference between the highest reference pitch and the singing upper limit pitch.

Description

  The present invention relates to a karaoke apparatus including means for creating and providing exemplary singing voice data for a singer's singing.

  In recent years, various services can be provided with the enhancement of the functionality of karaoke apparatuses. For example, in order to improve the singing ability of a singer, a service that enables creation and listening of exemplary singing voice data for the singing of the singer is one of them. In providing such an exemplary singing, it is desired to be an exemplary singing in a range where the singer can sing.

  Conventionally, karaoke apparatuses are also known that can score a user's singing voice data in comparison with model voice data and store singing voice data. A technique is known that can store singing voice data, evaluate singing power by comparing singing voice data and model voice data, and compare the singing voice and model voice for sections that could not be sung correctly. (Patent Document 1).

  In this case, the model voice data is often sung by a singer with excellent singing ability, such as a professional singer, and is not a professional singer but a model voice sung by the user's own voice for comparison. Is desirable for improving singing ability. Therefore, the technique which can correct | amend singing voice data by a user based on the pitch of the data of model singing is also disclosed (patent document 2), and the technique of patent document 2 is added to the technique of patent document 1. By using it, it can be compared with the model voice sung by the user's own voice.

  On the other hand, there is a user's own vocal range for singing, and a karaoke device provided with a function for measuring and announcing the vocal range of the singer is also provided (Patent Document 3), depending on the measured vocal range. The key can be set.

JP 2008-233736 A Japanese Patent Laid-Open No. 11-038987 JP 2004-326133 A

  However, when correcting the pitch of the singing voice using the technique of the above-mentioned Patent Document 2, it is not only necessary that the singing pitch is accurate, but it is necessary to be able to use it as an example when practicing singing. As a result of correcting the pitch, if the user's voice range (especially the singing upper limit note) is exceeded, the user cannot sing that part, which is inappropriate as an exemplary singing.

  In addition, it is conceivable that the vocal range of the singer is measured using a technique such as Patent Document 3 and singing after setting a key based on the measurement result. Since the key change is repeated during singing, it is unsuitable for singing for generating model voice data from the singing voice data, and there is a problem that it is necessary to sing again with the optimal key.

  Then, this invention is made | formed in view of the said subject, and provides the karaoke apparatus which can create the example audio | voice data by a singer's own voice so that it can sing in the vocal range of a singer by the singer's one time singing. Objective.

  In order to solve the above-mentioned problem, in the invention of claim 1, for each song section, for each song section, a reference pitch indicating a pitch to be sung with respect to a pitch at least sung by a singer in accordance with the song section; A karaoke apparatus having reference data as an analysis standard to be analyzed in relation to the above, wherein recording means for recording the singing voice of a predetermined song by a singer as singing voice data, and the recorded song for each song section of the song Detecting the singing pitch of the audio data, calculating the pitch difference compared with the corresponding reference pitch of the reference data, whether the singing pitch of each singing section needs to be corrected or not required depending on whether the pitch difference is within an allowable range or not Determining means for determining the singing pitch of the singing section to be corrected by the determining means. The singing pitch of the singing voice data corresponding to the highest reference pitch among all the singing sections of the music is required based on the determination result, based on the determination result and the correcting means for correcting the reference model voice data corrected based on the corresponding reference pitch. In the case of correction, a specifying means for specifying the singing pitch of the singing voice data corresponding to the reference pitch as a singing upper limit pitch that can be sung by the singer, and the highest reference pitch and the specifying among all singing sections of the song A transposing means that transposes the entire singing section of the reference model voice data to a lower one to make the transposition model voice data with a transposition amount in semitones based on the pitch difference with the singing upper limit pitch specified by the means. And

  In the invention of claim 2, as an analysis standard for analyzing each piece of music for each singing section in relation to a reference pitch indicating a pitch to be sung with respect to at least the pitch sung by the singer in accordance with the singing section. Karaoke apparatus comprising the reference data of the recording device, wherein the singing voice of the predetermined music by the singer is recorded as singing voice data, and the singing pitch of the recorded singing voice data is detected for each singing section of the music. Determining means for calculating a pitch difference in comparison with a corresponding reference pitch of the reference data, and determining whether the singing pitch of each singing section needs to be corrected or not to be corrected depending on whether the pitch difference is within an allowable range; and The singing pitch required to be corrected in the corresponding singing section by the judging means is the reference pitch corresponding to the reference data. When the singing pitch of the singing voice data corresponding to the highest reference pitch among all the singing sections of the music is based on the determination result, the correcting means for creating the reference model voice data corrected based on The specifying means for specifying the highest singing pitch as a singing upper limit pitch that can be sung by the singer, among the singing pitches that do not need to be corrected according to the determination result, and the highest reference pitch and the specifying means among all singing sections of the song A transposition means that transposes the entire singing section of the reference exemplary voice data to a lower one to make the transposition exemplary voice data with a transposition amount in semitones based on the pitch difference from the specified singing upper limit pitch. .

In the inventions of claims 3 and 4, "the transposing means further transposes all singing sections of the recorded singing voice data by the transposition amount and transposes the performance data of the music by the transposition amount". age,
“Contains selection means for selectively listening to the transposed singing voice data and performance data, or the transposition model voice data and the transposed performance data”.

  According to the first and second aspects of the invention, the singing pitch of the singing voice data is detected for each singing section of the music, and the pitch difference is calculated by comparing with the reference pitch, and whether or not the pitch difference is within the allowable range. It is determined whether the singing pitch of each singing section needs to be corrected or not required, and the reference model voice data in which the singing pitch to be corrected is corrected based on the reference data is created, and the entire singing of the song is performed based on the determination result. When the singing pitch of the singing voice data corresponding to the highest reference pitch in the section is a correction required, the singing pitch of the singing voice data corresponding to the reference pitch, or the singing pitch that does not need to be corrected based on the determination result The highest singing pitch is specified as the singing upper limit pitch that can be sung by the singer, the reference pitch and the singing upper limit pitch specified above. By transposing the entire singing section of the reference model voice data to the lower side to make the transposition model voice data with the transposition amount in semitones based on the pitch difference of the highest pitch of the reference model voice data after correction Singing by the singer once, so that the singing voice data of the singer's own voice can be sung in the singer's voice range. Is something that can be done.

  In the third and fourth aspects of the invention, the transposing means transposes the entire singing voice data in the singing voice data by the determined transposition amount, and the performance reproducing means also transposes the performance data of the music by the transposition amount, Listening comparison with a model singing that can be sung by the user by selectively listening to the transposed singing voice data and performance data, or the transposing model voice data and the transposed performance data. Can contribute to the convenience of the user.

It is a block block diagram of the karaoke apparatus which concerns on this invention. It is a process flowchart of the determination means of FIG. 1, and explanatory drawing of a determination result. It is a process flowchart of the specific means of FIG. It is a process flowchart of the transposing means of FIG. It is processing explanatory drawing of the transposing means of FIG. It is another process flowchart of the identification means of FIG. It is process explanatory drawing of the transposition means based on the process of FIG.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram of a karaoke apparatus according to the present invention. In FIG. 1, each karaoke terminal 11 has a display unit 13, a mixing amplifier 14, a microphone 15, a speaker 16, and a remote input / output device 17 that are externally connected to a karaoke main body 12 as a main device by wire or wirelessly. Connected.

  The display unit 13 displays a normal music selection display, a background image during karaoke performance, and the like. For example, a liquid crystal display (LCD), a plasma display (PDP), and other various displays can be employed. The mixing amplifier 14 mixes the audio signal from the microphone 15 with the music performance signal sent from the karaoke main body 12, amplifies it, and outputs it from the speaker 16.

  The remote input / output device 17 transmits / receives data to / from the karaoke main body 12 using a wired system or a wireless system (IR system, Bluetooth (registered trademark) mechanism piconet connection system, etc.) by a terminal transmission / reception unit (not shown). For this purpose, at least a terminal display unit 17A and a music selection piece registration unit 17B are provided as appropriate.

  The terminal display unit 17A can be used for input / output by laminating a liquid crystal display (LCD) and a touch sensor, and data such as selection of music can be input by the touch sensor corresponding to the displayed icon or the like. It has a GUI user interface function.

  The music selection music registration means 17B is a program including a table for searching for music and selecting music, and the music selection is registered in a reservation queue (41) in the karaoke main body 12 to be described later. In addition, at the time of music selection, display is selected that allows the user to select the creation and playback of exemplary audio data for his / her song after singing, or the playback of audio data transposed to a singable pitch. It is provided as a means, and when it is selected, it is attached to the selected music and registered in the reservation queue.

  The karaoke main body 12 includes a bus 20, a central control unit 21, a ROM 22, a RAM 23, a video display control means 24, a music performance control unit 25, a sound source (synthesizer) 26, a transmission / reception unit 27, an A / D conversion unit 28, and a storage unit 29. , Determination means 30, correction means 31, identification means 32, transposition means 33, and performance reproduction means 34.

  In the RAM 23, a storage area for a reservation queue 41, singing voice data 42, performance data 43, a determination result 44, and reference model voice data 45 is formed. The storage unit 29 also stores a music database (music DB) 51, a video database (video DB) 52, and a reference database (reference DB) 53. In addition, about each said structure, even if it is an element part which is not directly related to the summary of this invention, in order to show that most can be applied also in the conventional karaoke system, the whole component is demonstrated.

  The central control unit 21 is a physical CPU that performs overall processing control of the system, and performs algorithm processing based on a program stored in the ROM 22. The RAM 23 has a storage area for the reservation queue 41, singing voice data 42, performance data 43, determination result 44, and reference model voice data 45, and a work area for developing and executing the various programs. It is a concept that includes a case where it is configured by, for example, a semiconductor memory and is virtually built on a hard disk.

  The video display control means 24 is a program or an electronic circuit that outputs the background video extracted from the video DB 52 and the lyrics data of the music extracted from the music DB 51 to the display unit 13 during performance. The music performance control unit 25 includes, for example, a sequence program, and drives a sound source (synthesizer) 26 according to note data as performance data extracted from the music DB 51 with a music ID. The output of the sound source 26 is used as a performance signal. It is output to the mixing amplifier 14. The note data (performance data) is stored in the performance data storage area 43 of the RAM 23.

  The transmission / reception unit 27 is used to exchange data with the remote input / output device 17 using a wired system or a wireless system (such as an IR system or a Bluetooth (registered trademark) mechanism piconet connection system). Electronic circuit and program. The A / D conversion unit 28 is a program for digitally converting an audio signal that has passed through the mixing amplifier 14 from the microphone 15, and the digitally converted audio signal is recorded in the song audio data storage area 42 of the RAM 23 as song audio data. The The A / D converter 28 and the singing voice data storage area 42 constitute recording means.

  The music DB 51 stored in the storage unit 29 stores performance data (note data) and lyrics data for each music. Specifically, it has a music table in which a music ID, a music title, and an artist ID (artist name) are associated, and for each music, karaoke musical note data (for example, MIDI ( This is a database in which the music ID is stored as a file name for music data (file) composed of (registered trademark) format note data) and the like.

  The video DB 52 stored in the storage unit 29 is a database in which music IDs are stored as file names for background video data corresponding to each music. Reference DB53 memorize | stored in the said memory | storage part 29 is used as evaluation criteria for evaluating and analyzing the song by the singer matched with the karaoke music memorize | stored in the said music DB51 and the sufficiently small song area of the said karaoke music. It is a database that stores reference data used in association with each other. The reference data includes a reference pitch indicating a pitch to be sung according to at least a note number and velocity (size) for each music piece.

  In addition, by providing the reference DB 53, the karaoke apparatus 11 may be provided with a scoring means for singing using this reference data. The scoring means analyzes and sings the singing voice for each singing section of the music with respect to the song singing, and the audio signal from the microphone 15 passes through the mixing amplifier 14 and is digitally converted by the A / D converter 28. This is a program that performs scoring processing based on reference data for scoring the singing included in the music data for the recorded audio signal, and specifically uses, for example, the technique described in Japanese Patent No. 4229915 Can do.

  The determination unit 30 detects the singing pitch of the recorded singing voice data for each singing song section sufficiently small as described above, and extracts corresponding reference data from the reference DB 53 to cope with the sung music. This is a program that calculates the pitch difference in comparison with the singing pitch of the notes in the singing section, and determines whether the singing pitch in each singing section needs to be corrected or not required depending on whether the pitch difference is within an allowable range (details This will be described with reference to FIG. The determination result in this case is stored in the determination result storage area 44 of the RAM 23 (described in FIG. 5).

  The correction means 31 is a program for creating reference model voice data in which the singing pitch required to be corrected by the determination means 30 based on the reference data extracted by the determination means 30 is created. The sound data is stored in the reference model sound data storage area 45 of the RAM 23.

  As one method, the specifying means 32, when the singing pitch of the singing voice data corresponding to the highest reference pitch among all the singing sections of the song is based on the determination result, needs to be changed to the reference pitch. It is a program which specifies the singing pitch of corresponding singing voice data as a singing upper limit pitch which a singer can sing. (This will be explained with reference to FIGS. 4 and 5.)

  Further, as another method, when the singing pitch of the singing voice data corresponding to the highest reference pitch among all singing sections of the music is a correction required, the highest singing of the singing pitches that do not need to be corrected according to the determination result It is assumed that the pitch is specified as a singing upper limit pitch that can be sung by the singer (described with reference to FIGS. 6 and 7).

  The transposition means 33 reads the determination result by the determination means 30 and the reference model voice data created by the correction means 31 from the determination result storage area 44 and the reference model voice data storage means 45 of the RAM 23 for the sung song, and the determination result. The lower one of all the singing sections of the reference model voice data with the transposition amount in semitones based on the pitch difference between the highest reference pitch and the singing upper limit pitch specified by the specifying means 32 among all the singing sections of the singing song obtained more. Is a program for creating transposition model voice data by transposing to (described with reference to FIGS. 4 and 5). Note that the transposition amount transposed to the lower side includes a transposition amount “zero” (this point will also be described with reference to FIGS. 4 and 5).

  The transposing means 33 further transposes the entire singing section of the corresponding music of the read singing voice data by the transposition amount, and transposes the performance data of the music by the transposition amount to the performance reproducing means 34. A program for controlling playback is provided.

  The performance reproducing means 34 is a program for transposing and reproducing the musical performance data by a predetermined transposition amount (output to the music performance control unit 25), and the transposition amount is controlled by the transposing means 33.

  Therefore, FIG. 2 shows a processing flowchart of the determination unit of FIG. 1 and an explanatory diagram of the determination result. In FIG. 2A, after singing a predetermined music, singing voice data sung by the music and recorded in the singing voice data storage area 42 of the RAM 23 is acquired, and corresponding reference data is acquired from the reference DB 53 (step). 1 (S1)), the singing pitch of the singing voice data sung for each singing section is detected (S2).

  Subsequently, the pitch difference is calculated by comparing each singing pitch detected for each singing section with the pitch (reference pitch) of the corresponding singing section of the reference data (S3). Here, the allowable range of the pitch difference is set to be, for example, less than ± 10 cent, and it is determined whether or not the calculated pitch difference is within the allowable range (S4). If it exceeds the allowable range, it is determined that correction is required ("NG") (S5-1), and if it is within the allowable range, it is determined that correction is not necessary ("OK") (S5-2). And these determinations are processed in all the singing sections (S6).

  In addition, in the determination result by the determination means 30, the song can be sung at the correct pitch to be sung once anywhere in the song to be sung (the difference from the reference pitch is within an allowable range and the pitch determination result is “OK”). ) Is considered to be singable, that is, below the upper limit singing note. When the same note is determined a plurality of times, the difference is the smallest difference among the differences.

  The determination result by the determination means 30 is shown as an example in FIG. FIG. 2B shows pitch differences and determination results of, for example, five notes in order from the highest note “G5” among the notes to be sung. In the MIDI device, the “do” in the center of the piano keyboard (88 keys) is “C4”, the highest note “do” is “C8”, and the lowest note “la” is “A0”. It is normal to represent. Further, although not shown, the determination result is accompanied by values of the reference pitch and the singing pitch that are to be determined.

  FIG. 3 shows a process flowchart of the specifying means of FIG. In FIG. 3, the specifying unit 32 reads the determination result by the determination unit 30 of the song from the RAM 23, and first, the song of the singing voice data corresponding to the highest pitch (reference pitch) among all the song sections of the song The pitch determination result is referred to (S11). If the determination result does not require correction (S12), the reference exemplary voice data created by the correction means 31 is assumed not to be transposed, and the process ends.

  On the other hand, in the case of correction required (S12), the pitch of the singing voice data corresponding to the highest reference pitch among all the singing sections of the singing song is specified as the singing upper limit pitch that can be sung by the singer (S13). That is, from the determination result shown in FIG. 2 (B), the pitch difference between the highest reference pitch (note number “G5” pitch) of all the song sections of the song and the song pitch of the corresponding song voice data is obtained. Since it is “−30 cent” and the pitch determination is “NG” and correction is necessary, the singing pitch of the singing voice data (the pitch obtained by subtracting 30 cent from the pitch of the note number “G5”) is the singing of the singer. It is specified as a possible singing upper limit pitch.

  In addition, the above-mentioned singing upper limit pitch can be sung without problems with "F # 5" or less that is one semitone (100 cents) lower than "G5". Even if the pitch determination of note numbers “F5” and “E5” is “NG”, these determinations and the singing upper limit pitch are irrelevant.

  FIG. 4 shows a process flowchart of the transposing means of FIG. 1, and FIG. 5 shows a process explanatory diagram of the transposing means of FIG. In FIG. 4, the transposing means 33 reads the determination result by the determination means 30 from the determination result storage area 44 of the RAM 23 (S21). From the read determination result, the transposition amount is determined based on the singing upper limit pitch specified by the specifying means 32, that is, the pitch difference of the singing voice data of the singing voice data corresponding to the highest reference pitch in all singing sections (S22). That is, as shown in FIG. 5, “−30 cent” is obtained as the pitch difference between the reference pitch of the highest note number “G5” in the entire singing section and the singing pitch of the corresponding singing voice data, and the pitch is based on this. The transposition amount including the difference is determined as “100 cent” of one semitone.

  Then, the reference model voice data corrected by the correction means with the determined transposition amount is transposed to the lower one in the entire singing section to create the transposition model voice data (S23). That is, as shown in FIG. 5, the transposition model is transposed (transposed so that the highest note number is “F # 5”) by one semitone (100 cent) in which all the singing sections of the reference model audio data are determined. Create audio data. In other words, by transposing down one semitone (100 cent), the highest note to be sung becomes “F # 5”, so even if the pitch of the highest note that was “NG” is corrected, The pitch does not exceed the singing upper limit pitch of the singer, and is useful as transposition model voice data that can be sung by the singer.

  Note that if the pitch difference between the note number “G5” and the corresponding singing pitch is “−110 cent (NG)”, for example, even if transposition is performed so that the highest note number is “F # 5”. The upper “−10 cent” is the limit, and it is impossible to sing at the correct pitch, and the transposition is down by two semitones (200 cent) so that the highest note number becomes “F5”.

  By the way, the transposition amount transposed to the lower side in this case includes the transposition amount “zero”. For example, if the pitch difference between the highest note singing pitch and the reference pitch is positive and “NG (requires correction)”, the singing pitch will be adjusted downward, so that the singing pitch will remain below the upper limit of the singing and transposition is unnecessary ( Transposition amount “zero”). Therefore, in effect, transposing the entire singing section of the reference model voice data to the lower one with a transposition amount in semitones based on the pitch difference in which the pitch difference between the highest reference pitch and the singing upper limit pitch is negative. Become.

  Next, FIG. 6 shows another processing flowchart of the specifying means of FIG. 1, and FIG. 7 shows a process explanatory diagram of the transposing means based on the processing of FIG. In FIG. 6, the specifying unit 32 reads the determination result by the determination unit 30 of the song from the RAM 23, and first, the singing of the singing voice data corresponding to the highest pitch (reference pitch) among all the song sections of the song. The pitch determination result is referred to (S31). If the determination result does not require correction (S32), the reference exemplary voice data created by the correction means 31 is assumed not to be transposed, and the process ends.

  On the other hand, in the case of correction required (S32), the highest singing pitch among the singing pitches that do not need to be corrected based on the song determination result is specified as the singing upper limit pitch that can be sung by the singer (S33). That is, from the determination result shown in FIG. 7 (same as FIG. 2B), the highest singing pitch among the singing pitches that do not need to be corrected is specified as the singing upper limit pitch, and the note number “D5” is the object of transposition amount determination. It becomes.

  In addition, although the above-mentioned singing upper limit pitch is specified, although “G5” can be sung at “−30 cent”, “F5” and “E5” have never been sung at the correct pitch, that is, the reference If the difference from the pitch is outside the allowable range and the pitch determination result is “NG”, it is difficult to sing at the correct pitch with notes of “D # 5” or higher when singing the melody of the music at least. It is based on the idea that “D5” is the upper limit of notes that can be sung at the correct pitch.

  Therefore, as shown in FIG. 7, the transposing means 33 calculates a pitch difference between the pitch value of the singing upper limit pitch specified by the specifying means and the pitch value of the highest reference pitch in all singing sections from the determination result. Then, the transposition amount “500 cent” of five semitones to be lowered from the note “G5” to “D5” as a semitone unit corresponding to the pitch difference is determined.

  Then, with the determined transposition amount, the reference model voice data corrected by the correction means is transposed to the lower one in the entire singing section to create the transposition model voice data. That is, as shown in FIG. 7, by transposing down “5 semitones (500 cents)”, the highest note to be sung becomes “D5”, and the corrected pitches of the notes that were “NG” are both singers. This is useful as transposition model voice data that can be sung by a singer without exceeding the upper singing upper limit pitch (note “D5” pitch).

  By the way, in the music selection music registration means 17B of the remote input / output device 17, at the time of music selection, generation and reproduction of exemplary voice data for the user's own song after singing, or voice data obtained by transposing the user's song to a singable pitch. A display capable of inputting to select reproduction is provided as selection means. For example, when reproduction of transposition model audio data and transposed performance data is selected, the transposition means 33 outputs the created transposition model audio data to the music performance control unit 25 and determines the above-mentioned determination for the performance reproduction means 34. The performance reproduction means 34 reads out the performance data of the music from the performance data storage area 43 of the RAM 23 and transposes the transposition amount instructed by the transposition means 33 with the transposition amount. Then, it is reproduced by being output to the music performance control unit 25.

  Also, when playback of the transposed singing voice data and performance data is selected, the transposing means 33 reads the singing voice data from the singing voice data storage area 42, transposes it by the transposition amount, and outputs it to the music performance controller 25. At the same time, the performance reproducing means 34 is controlled to transpose the performance data of the music piece with the determined transposition amount, and the performance reproduction means 34 reads the performance data of the music piece from the performance data storage area 43 of the RAM 23. It is reproduced by transposing the transposition amount instructed by the transposing means 33 and outputting it to the music performance control unit 25.

  In other words, it is possible for the user to compare his / her own song before and after the correction and the singable model song, and at that time, the user's voice and singable voice range are played back, which contributes to the convenience of the user. It is something that can be done.

  As described above, since the pitch of the highest note of the modified reference model voice data is assumed to be transposition model voice data that does not exceed the upper limit pitch of the singer, the singer's own singing, It is possible to provide model voice data by voice so that it can be sung in the vocal range of the singer.

  The karaoke apparatus of the present invention can be used in the industry of manufacturing and using an apparatus including means for generating and providing exemplary singing voice data for a singer's song.

DESCRIPTION OF SYMBOLS 11 Karaoke apparatus 12 Karaoke main body 17 Remote input / output apparatus 25 Music performance control part 28 A / D conversion part 30 Judgment means 31 Correction means 32 Specification means 33 Transposition means 34 Performance reproduction means 44 Judgment result storage area 45 Reference | standard model voice data storage area 53 Reference Database

Claims (4)

A karaoke apparatus provided with reference data as an analysis standard for analyzing a song by a singer for each song section in relation to a reference pitch indicating a pitch to be sung at least with respect to a pitch sung in accordance with the song section. Because
Recording means for recording the singing voice of the predetermined music by the singer as singing voice data;
For each singing section of the music, the singing pitch of the recorded singing voice data is detected, and a pitch difference is calculated by comparing with the corresponding reference pitch of the reference data. A determination means for determining whether the singing pitch of the singing section requires correction or correction, and
Correction means for creating reference model voice data in which the singing pitch required for correction of the corresponding singing section by the determination means is corrected based on the corresponding reference pitch of the reference data;
Based on the determination result, when the singing pitch of the singing voice data corresponding to the highest reference pitch among all the singing sections of the tune is to be modified, sing the singing pitch of the singing voice data corresponding to the reference pitch. Specific means for specifying the upper limit pitch that can be sung by the person,
Transpose the entire singing section of the reference model voice data to the lower one with a transposition amount in semitones based on the pitch difference between the highest reference pitch and the singing upper limit pitch specified by the specifying means among all singing sections of the song Transposing means for transposing model voice data,
A karaoke apparatus comprising: A karaoke apparatus provided with reference data as an analysis standard for analyzing a song by a singer for each song section in relation to a reference pitch indicating a pitch to be sung at least with respect to a pitch sung in accordance with the song section. Because
Recording means for recording the singing voice of the predetermined music by the singer as singing voice data;
For each singing section of the music, the singing pitch of the recorded singing voice data is detected, and a pitch difference is calculated by comparing with the corresponding reference pitch of the reference data. A determination means for determining whether the singing pitch of the singing section requires correction or correction, and
Correction means for creating reference model voice data in which the singing pitch required for correction of the corresponding singing section by the determination means is corrected based on the corresponding reference pitch of the reference data;
Based on the determination result, when the singing pitch of the singing voice data corresponding to the highest reference pitch among all singing sections of the music is a correction required, the highest singing of the singing pitches that need not be corrected based on the determination result A specifying means for specifying the pitch as a singing upper limit pitch that the singer can sing,
Transpose the entire singing section of the reference model voice data to the lower one with a transposition amount in semitones based on the pitch difference between the highest reference pitch and the singing upper limit pitch specified by the specifying means among all singing sections of the song Transposing means for transposing model voice data,
A karaoke apparatus comprising: The karaoke apparatus according to claim 1 or 2,
A performance reproduction means for transposing and reproducing the performance data of the music by the transposition amount determined by the transposition means,
The transposing unit further transposes the entire singing section of the recorded singing voice data by the transposition amount, and controls the performance reproducing unit to transpose and reproduce the performance data of the music by the transposing amount. Karaoke device characterized by.   4. The karaoke apparatus according to claim 3, further comprising selection means for selectively listening to the transposed singing voice data and performance data, or the transposition model voice data and the transposed performance data. Karaoke device.

Images