JP2002221978A - Vocal data forming device, vocal data forming method and singing tone synthesizer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize natural singing by a virtual singer made to meet accompani ment by uttering phonemes for consonants among the phonemes constituting syllables so as to meet the timing to generate notes. SOLUTION: The vocal data including the timing data to generate the notes by each of the syllables corresponding to lyrics is previously stored. When reproduction processing is started, the timing to generate the vowel 'a', the operation to utter the consonant 's' is started before the timing to generate the note in having the syllable 'sa' corresponding to the note 'doh' uttered and the timing to generate the vowel 'a' is made to meet the timing to generate the note 'doh'. As a result, the natural singing by the virtual singer is made possible without delaying in the accompaniment.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vocal data generating apparatus, a vocal data generating method, and a singing sound synthesizing apparatus suitable for use in, for example, computer music by a personal computer.

[0002]

2. Description of the Related Art Recent computer music is
Due to its development, it has been used in many fields such as commercial karaoke, television commercials, and accompaniment of pops, and its performance level has reached a level close to that of an orchestra or a live band. Computer Music is DTM
(Desktop music). This DTM has a MIDI sound source, and uses this MIDI sound source to imitate a hit song, or to create a personalized song by changing the tone and melody as a creator or arranger, and enjoy playing. It is also possible. Further, the DTM includes a music system that allows a virtual singer (fictional singer) to sing lyrics using vocal data in which syllables of lyrics are assigned to notes of a melody part.

In this DTM system, MIDI data is used, and the MIDI data is composed of a data string of a note-on signal and a note-off signal. The MIDI data generates a note of a certain pitch (note number) by a note-on signal, and stops the note by a note-off signal. The note number is determined by the MIDI standard, for example, note number = “C3”
Is “do” and note number = “D3” is “re”. On the other hand, the DTM generates a syllable in a simulated manner from a singing sound source or the like, and associates the syllable with a note-on signal to produce a voice. Thus, in the DTM system,
It becomes possible to make the virtual singers sing the lyrics based on the vocal data.

[0004]

When a human sings a song, the singer speaks a lyric attached to a note having a certain pitch and length. At this time, instead of generating a voice corresponding to the lyrics at the timing of the note, preparations for singing the lyrics are started unconsciously before the timing of the note. In a syllable that starts with a consonant, the position at which the consonant is emitted is unintentionally shifted forward, and a vowel that is actually heard as a sound is generated at the timing of the note.

However, in the conventional DTM, when a virtual singer utters a syllable of a lyric, vocal data in which a syllable is assigned to a note is used, so that the syllable is uttered based on the note. . For this reason,
If the syllable consists only of vowels, this syllable will be heard slightly after the note, but if it is a syllable that starts with a consonant, it will be heard as a sound when the vowel is uttered. You will hear the syllables.

Here, the delay of the lyrics with respect to the accompaniment will be described by exemplifying the part of "tumorous blooming" at the beginning of the singing of "tulip" shown in FIGS.
FIG. 10 shows MIDI data and syllable utterance corresponding to a note.
ON2, KON3 ... and note-off signal is KOF
F1, KOFF2, KOFF3,... FIG. 11 is a diagram showing the utterance of syllables with respect to the length of a note. FIGS. 10 and 11 show only the case where the syllable is uttered, and omit the stop of the syllable.

[0007] Here, as described above, the person "blossomed".
In order to utter “sa” when the first beat “do” occurs, the phoneme “s” that constitutes “sa” as a syllable must be set in advance from the position where “do” occurs. The vowel “a”, which is a phoneme constituting the syllable “sa”, is also shifted forward to match the position of the note “do”. However, the DTM system uses “note” based on note-on KON1 for generating “do” on the first beat.
The process for producing the syllable is started. For this reason,
The sounding timing of the consonant "s" is delayed by the time .DELTA.t1 from the sounding timing of "do", and the sounding timing of the vowel "a" is further delayed by the time .DELTA.t2 from the occurrence timing of "do" (see FIG. 11). . Here, the time from a consonant of a syllable starting with a consonant to a change to a vowel is set to a fixed time t0 for convenience. Next, the DTM system starts processing for producing a syllable of “I” based on the note-on KON2 for generating “RE” of the second beat. Therefore, the sounding timing of the vowel “i” is delayed by the time Δt1 from the sounding timing of “の” (see FIG. 11). Further, the DTM system uses the note-on KON3 for generating the third beat “mi” based on “ta”.
The process for producing the syllable is started. For this reason,
The sounding timing of the consonant "t" as a phoneme is delayed by a time? T1 from the sounding timing of "mi", and the sounding timing of a vowel "a" as a phoneme is delayed by a time? T2 from the sounding timing of "mi". (Fig. 11
reference).

As described above, in the conventional DTM system,
If you sing the lyrics to the virtual singer along with the accompaniment, the syllables of the lyrics will be uttered later than the melody part note, especially the syllables that begin with consonants will appear noticeably, and the lyrics will not match the accompaniment There was a problem that.

In addition, when the lyrics are matched with the accompaniment, it can be understood that the syllables of the lyrics need only be adjusted in sounding timing. However, if the pronunciation timing is set to a syllable that starts with a consonant, when a syllable consisting of only vowels is reproduced, this syllable will be earlier than the timing of generation of a note. Thus, simply adjusting the tone generation timing uniformly requires
Natural singing by virtual singer could not be obtained.

[0010] The present invention has been made in view of the above problems, and has a vocal data generating apparatus, a vocal data generating method, and a vocal data generating method for generating vocal data capable of realizing a natural singing by a virtual singer according to accompaniment. It is an object of the present invention to provide a singing sound synthesizer.

[0011]

According to one aspect of the present invention, there is provided a vocal data generating apparatus for generating vocal data corresponding to a melody and lyrics. Lyric allocating means for allocating to a note corresponding to the syllable, and a pronunciation corresponding to each syllable in order to match the pronunciation timing of a vowel phoneme to the pronunciation timing corresponding to the note among the phonemes constituting each syllable. Sounding timing data generating means for generating timing data; data output means for generating phoneme string data of the syllables, the sounding timing data and pitch data corresponding to the syllables as vocal data, and outputting the vocal data; It is characterized by having.

According to a second aspect of the present invention, there is provided a vocal data generating apparatus for generating vocal data corresponding to a melody and lyrics, wherein the lyrics allocating means assigns a syllable of the lyrics to a note corresponding to the melody; Sounding timing data generating means for generating sounding timing data corresponding to each syllable in order to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note, out of the phonemes constituting each syllable, divided into phonemes; Data output means for generating pitch data corresponding to the phoneme string data, the tone generation timing data and the syllables as vocal data, and outputting the vocal data included in the system exclusive message. I have.

[0013] The invention according to claim 3 is the invention according to claim 1 or 2.
In the vocal data generation device described above, the data output means outputs the vocal data separately for each chorus.

The invention described in claim 4 is the first or second invention.
In the vocal data generation device described above, the data output means outputs the vocal data divided for each phrase.

The invention according to claim 5 is the first or second invention.
In the vocal data generation device described above, the data output means outputs the vocal data separately for each breathing section.

The invention according to claim 6 is the first or second invention.
In the vocal data generation device described above, the data output means outputs the vocal data for each bar.

The invention according to claim 7 is a vocal data generating method for generating vocal data corresponding to a melody and lyrics, wherein a lyric allocating step of assigning a syllable of the lyrics to a note corresponding to the melody, and A sounding timing data generating step of generating sounding timing data corresponding to each syllable in order to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note, out of the phonemes constituting each voice by dividing into phonemes; And generating pitch data corresponding to the syllable, the phoneme sequence data, and the syllable timing data, and outputting the vocal data.

The invention according to claim 8 is a vocal data generating method for generating vocal data corresponding to a melody and lyrics, wherein a lyric allocating step of allocating a syllable of the lyrics to a note corresponding to the melody, and A sounding timing data generating step of generating sounding timing data corresponding to each syllable in order to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note, out of the phonemes constituting each voice by dividing into phonemes; A data output step of generating pitch data corresponding to the phoneme string data, the sounding timing data and the syllables as vocal data, and outputting the vocal data included in a system exclusive message. I have.

The invention according to claim 9 is a singing sound synthesizing apparatus for singing lyrics to a virtual singer in accordance with an accompaniment, wherein lyrics are assigned to notes of accompaniment data and melody data, and the sound timing of the notes is set at the sounding timing of the notes. Data output means for outputting vocal data including sounding timing data corresponding to the vowel sounding timing for each syllable, syllable phoneme string data and pitch data corresponding to the syllable, and data for transmitting these data during reproduction Control means, an accompaniment sound source that generates the accompaniment sound by receiving the accompaniment data transmitted by the data control means, and receives the vocal data transmitted by the data control means, and generates a note based on the vocal data. And a singing sound source for producing syllables at corresponding intervals.

According to a tenth aspect of the present invention, there is provided a singing sound synthesizing apparatus for singing lyrics to a virtual singer along with an accompaniment, wherein lyrics are assigned to musical notes of accompaniment data and melody data, and the sounding timing of the musical notes is set. Data output means for outputting vocal data including vocal data including vocal data including vowel sound timing corresponding to vowels for each syllable, phoneme string data of syllables and pitch data corresponding to the syllables in a system exclusive message, A data control means for transmitting these data, an accompaniment sound source for generating an accompaniment tone upon receiving the accompaniment data transmitted by the data control means, and vocal data in the system exclusive message transmitted by the data control means To the note based on the vocal data. It is characterized by comprising a singing sound source to sound the syllables in pitch to respond, the.

[0021]

Embodiments of the present invention will be described below with reference to the drawings.

A. FIG. 1 is a diagram showing an example of a system configuration of a singing sound synthesizer (hereinafter, referred to as DTM) according to the present invention. This DT
The M system generally includes a personal computer 1 (hereinafter, referred to as a personal computer 1), a tone generator 31, a speaker 50, and a MIDI keyboard 60. FIG. 2 is a diagram showing the configuration of the personal computer 1 and the tone generator 31, and further shows the connection state. The personal computer 1 may be a sequencer, and the tone generator 31
Includes a voice pronunciation device.

A-1. 1. Configuration of the Personal Computer 1 The personal computer 1 includes a personal computer body 10 having a CPU 13, a RAM 11, a ROM 12, and a hard disk 14 (see FIG. 2), a monitor 27 that receives image signals from the personal computer body 10 and performs various displays, and a personal computer body. Keyboard 28 for inputting various information such as lyrics and command information to 10
And a mouse 29.

As shown in FIG. 2, a detection circuit 16, a detection circuit 17, and a display circuit 1 are connected to a bus 15 of the personal computer body 10.
8, the interface 19, the hard disk 14,
The RAM 11, the ROM 12, the CPU 13, and the like are connected.

The detection circuit 16 detects an input of a key (numerical key, character key, or the like) on the keyboard 28 and generates a key signal. The detection circuit 17 detects a moving operation and a switch operation of the mouse 29 and generates a mouse signal. The worker
Using the keyboard 28 or the mouse 29, accompaniment data is generated and vocal data is edited.

The display circuit 18 is connected to a monitor 27.
An edit screen or the like can be displayed on the monitor 27, and the operator generates the accompaniment data and edits the vocal data while referring to the edit screen on the monitor 27.

The ROM 12 stores various parameters such as accompaniment data and vocal data and a control program.
The RAM 11 stores flags, buffers, and the like.
It is also possible to store the control program and the like supplied from the hard disk 14. The CPU 13 is a RAM 11
Alternatively, according to a control program stored in the ROM 12, calculation or control for editing vocal data or the like is performed. The control program stored in the ROM 12 or the like performs an accompaniment data generation process, a musical instrument sound generation process, a vocal data generation process, a music reproduction process, a display process, and the like, which will be described later. A timer 20 is connected to the CPU 13, and the CPU 13 performs an interrupt process at predetermined time intervals according to the time information of the timer 20.

The interface 19 comprises a MIDI interface. The interface 19 is connected to the interface 38 of the tone generator 31 via a MIDI cable. Thereby, the personal computer 1 transmits the accompaniment data and the vocal data to the tone generator 31 via the interface 8.

A-2. Configuration of Tone Generator 31 The tone generator 31 has a MIDI sound source 32 and a singing sound source 33. The bus 34 has a MIDI sound source 3
2. Effect circuit 37, singing sound source 33, detection circuit 35, display circuit 36, interface 38, RAM 39, ROM
40, a CPU 41 and the like are connected.

The detection circuit 35 is provided with an operation element 42 or MID
An operation of the I keyboard 60 is detected, and an operation signal is generated. The operation element 42 is a performance operation element such as a switch. The display circuit 36 displays the operation state of the tone generator 31 on the display panel 43.

The ROM 40 stores formant data for synthesizing voice, various other data, and a control program. The RAM 39 stores accompaniment data and vocal data transmitted from the personal computer 1.
Instead of storing the control program in the ROM 40,
The information may be stored in the RAM 39. CPU 41 is a ROM
According to the control program stored in 40, the music is reproduced by performing calculation or control.

The CPU 41 obtains time information from the connected timer 44, and reproduces accompaniment data or vocal data according to the time information. The CPU 41 generates a tone parameter and an effect parameter based on the note-on signal and the like, and supplies them to the MIDI tone generator 32 and the effect circuit 37, respectively. The MIDI tone generator 32 generates a tone signal in accordance with a supplied tone parameter.
The effect circuit 37, depending on the effect parameter supplied,
The tone signal generated by the MIDI tone generator 32 is given an effect such as delay or reverb, and is supplied to the DA conversion circuit 45. The DA conversion circuit 45 converts the supplied digital musical tone signal into an analog musical signal, and generates an accompaniment from the speaker 50. The MIDI sound source 32 has a waveform memory system, an FM system, a physical model system, a harmonic synthesis system,
Any system such as an analog synthesizer system of VCO + VCF + VCA may be used.

The interface 38 comprises a MIDI interface. For example, the tone generator 31
When the interface 38 of the personal computer 1 and the interface 19 of the personal computer 1 are connected by a MIDI cable, MIDI communication is performed between the tone generator 31 and the personal computer 1.

The CPU 41 is provided with the interface 3
The vocal data is received from the personal computer 1 through the PC 8 and stored in the RAM 39. The vocal data includes note data, phoneme string data, and sounding timing data, and will be described later in detail.

Further, the CPU 41 reads the vocal data stored in the RAM 39 and supplies the vocal data to the singing sound source 33 based on the formant data stored in the ROM 40. The formant data includes, for example, formant center frequency (center frequency of a mountain forming a formant) data required to form a formant corresponding to each phoneme, and formant band width (bandwidth of a mountain forming a formant)
Data, formant level (peak level of a mountain forming a formant) data, and the like. Singing sound source 33
Generates an audio signal according to the supplied formant data. The audio signal corresponds to a singing voice having a predetermined pitch. The singing sound source 33 may be a formant synthesis method (singing sound source) or another method. The formant synthesis method is described in, for example, JP-A-3-200299.

The effect circuit 37 applies an effect such as a delay to the audio signal generated by the singing sound source 33 in accordance with the supplied effect parameter, and supplies the result to the DA conversion circuit 45. The DA conversion circuit 45 converts the supplied digital audio signal into an analog audio signal and causes the speaker 50 to produce a singing voice.

The singing sound source 33 and the MIDI sound source 3
2 is not limited to a configuration using dedicated hardware, but may be configured using a DSP + microprogram, or may be configured using a CPU + software program. Further, a plurality of sounding channels may be formed by using one singing sound source or sound source circuit in a time-division manner, or one singing sound source for one sounding channel using a plurality of singing sound sources or sound source circuits. Alternatively, a plurality of tone generation channels may be constituted by a tone generator circuit.

B. Control Processing Next, DTM control processing will be described. This control program causes the CPU 13 or the CPU 41 to perform accompaniment data generation processing, musical instrument sound generation processing, vocal data generation processing, music reproduction processing, display processing, and the like.

Here, the outline of each process described above will be described. Accompaniment data generation processing The user enters a musical score in the musical notation input window displayed on the monitor 27 with the keyboard 28 (mouse 29) or MI
The musical note is written using the DI keyboard 60, and the musical note is converted to MIDI format accompaniment data on the hard disk 14.
To memorize. At this time, the accompaniment data is stored individually for each part (for example, musical instrument). Musical instrument sound generation processing The user selects a musical instrument (for example, a drum, a guitar, an electric tone, etc.) for each part data written by the music creation processing, and further arranges and effects processing. 14 is stored. Vocal Data Generation Processing Vocal data corresponding to a melody and lyrics is generated, and the vocal data is stored in the hard disk 14. Music Playback Processing Using the MIDI sound source 32 and the singing sound source 33 of the tone generator 31, a musical tone based on vocal data and accompaniment data is generated from the speaker 50. Display Processing Various screens are displayed on the monitor 27.

Among these processes, the accompaniment data generation process, the musical instrument sound generation process, and the display process are conventionally performed by DTM.
Although the detailed description is omitted, the vocal data generation processing and the music reproduction processing will be described later in detail.

B-1. Vocal Data Generation Process First, the vocal data generation process described above is described with reference to FIG.
A description will be given based on the flowchart of FIG. First, CPU1
No. 3 performs a lyrics assignment process based on a control program for generating vocal data (step S1).

Here, the lyrics assignment process is a process of assigning the syllables of the lyrics to the notes of the melody data.
For example, processing such as assigning the syllable data number and the note data number together from the beginning is performed.

The vocal data table (see FIG. 4) generated by the lyrics assigning process is stored in the hard disk 14. FIG. 4 shows, for example, a table of a portion of the lyrics “flower bloomed” of “tulip”.
This data table includes note (note) data, generation timing data, lyrics, and phoneme string data. The note data is a note number (indicating a pitch). The phoneme string data includes data relating to each phoneme and data expressing exhalation (that is, breathing when a person pronounces a voice). The phoneme string data corresponds to each syllable of “lyrics” (a kana character in the present embodiment).

For example, the syllables constituting the lyrics “sa” on the first line are replaced with phoneme string data “s + a”,
Means that the word "i" is pronounced at the pitch (note) of the second line, and the lyrics "i" on the second line are replaced with phoneme string data "i".
It means that it is pronounced at the pitch of D3.

The sounding timing data indicates the sounding timing of the first phoneme. As shown in FIG. 9, for the syllable assigned to the note, the sounding timing when the vowel of the phoneme string data is sounded, This shows the timing to match the sounding timing when a note is sounded. In the case of a syllable represented only by a vowel, the sounding timing of the syllable almost coincides with the sounding timing of the note, and in the case of a syllable starting with a consonant, the sounding timing of the syllable is earlier than the sounding timing of the note.

Returning to FIG. 3, the CPU 13 returns to FIG.
Then, a system exclusive message including the vocal data is created (step S3). Generally, this system exclusive message is information used for operating a function unique to a MIDI sound source, and is a transmission / reception message that is set uniquely by a manufacturer for a MIDI instrument or the like manufactured by the manufacturer. This message is sent to the sound source at the very beginning of the music data, so that the tone color and effects are set using the unique functions of the MIDI sound source. By using this system exclusive message,
Advanced musical tone expression is realized by MIDI sound source.

The system exclusive message according to the present embodiment is configured as shown in FIG. That is, at the beginning, an exclusive status “F0” indicating a system exclusive message, the next position is a manufacturer ID, and the third position is an identification number for transmitting / receiving an exclusive message (device I).
D), next is a phoneme string data message, next is a sounding timing message, and a syllable pitch message. The phoneme string data message, sounding timing message, and pitch message each indicate one syllable. Then, these three data are repeated for each syllable to form vocal data for one song, and the end of exclusive "F7" is assigned to the last position.

Further, the CPU 13 stores the system exclusive message generated in step S3 on the hard disk 14, and ends this processing.

B-2. Music Playback Processing Next, music playback processing will be described based on the flowchart of FIG. First, the CPU 13 of the personal computer 1 receives a signal from the keyboard 28 or the mouse 29 and sends the accompaniment data stored on the hard disk 14 and the vocal data expressed as a system exclusive message to the tone generator 31 via the interfaces 19 and 37. Send. Tone generator 31
CPU 41 starts music playback processing upon receiving the accompaniment data and the system exclusive message (step S11).

The CPU 41 stores the received accompaniment data and the system exclusive message in the RAM 39 based on the control program, and generates vocal data from the stored system exclusive message (step S12). At this time, note data, tone generation timing data, lyrics and phoneme string data are read from each message written in each part of the system exclusive message for each syllable, and a vocal data table as shown in FIG. 4 is created.

Next, the CPU 41 starts an accompaniment reproduction process based on the accompaniment data (step S13). In this accompaniment reproduction process, reproduction is performed with a musical instrument for each part. Since this processing is the same as that of the conventional technique, its details are omitted. Further, the CPU 41 starts the playback processing of the lyrics by the virtual singer from the vocal data table (step S14). Then, let the virtual singers sing the lyrics along with the performance.

Here, the flow of the lyrics reproduction process will be described with reference to the flowchart of FIG. That is, the CPU 41
Starts the timer 44 (step S21), monitors whether or not the count value of the timer has reached the sound generation timing t in the data table, and sounds the syllables in order each time the sound generation timing is reached (step S22). . Then, the CPU 41 monitors whether or not all the vocal data is generated (step S23), and repeats this process until all the vocal data is generated.

Regarding the sounding timing of notes and syllables,
A specific example will be described with reference to FIGS. This specific example shows the part of "tulips bloomed" at the beginning of the singing of "tulips". As described above, the sounding timing data is data obtained by combining the sounding timing when a vowel is uttered and the sounding timing when a note is generated in the syllable assigned to the note. Therefore, the CPU 41 generates the syllable “sa” corresponding to the note-on KON1 for generating the first beat “do” by the time t0 before the note generation timing, that is, the timer 44. After t1 from the start, the sounding operation of the consonant "s" which is a phoneme constituting the syllable "sa" has begun. Thus, the sounding timing of the vowel "a", which is a phoneme constituting the syllable "sa", substantially matches the sounding timing of "do".

Further, the CPU 41 generates the syllable “i” corresponding to the note-on KON2 for generating the second beat “re” by outputting the vowel “i” after t2 from the start of the timer 44. He has begun sounding. Thus, the sounding timing of the vowel "i" substantially coincides with the sounding timing of "RE".

Further, the CPU 41 sets the third beat of "mi"
In order to sound a syllable "ta" corresponding to the note-on KON3 for sounding a note, a time t0 before the sounding timing of the note, that is, t from the start of the timer 44,
Three seconds later, a sounding operation of a consonant "t" which is a phoneme constituting the syllable "ta" has begun. Thus, the sounding timing of the vowel "a", which is a phoneme constituting the syllable "ta", substantially coincides with the sounding timing of "mi".

As described above, in the DTM system according to the present embodiment, the syllables of the lyrics can be reproduced without delaying the notes of the melody, and the virtual singers can sing the lyrics. This makes it possible to make the lyrics sing along with the accompaniment.

C. Effects of the Embodiment In the present embodiment, the sounding timing data set so that the sounding timing of the melody note and the sounding timing of the vowel of the syllable corresponding to the lyrics are transmitted to the tone generator 31 in advance when the music is reproduced. By doing so, it is possible to sing in real time by the virtual singer. As a result, in the DTM system, it is possible to realize a timing of starting singing close to natural singing.

D. Modifications (1) In the above-described embodiment, the case where vocal data for one song is included in the system exclusive message has been exemplified. However, the present invention is not limited to this. It is good enough to transmit the tone generator 31 in a certain unit in advance of the note generation timing.

(2) The performance data may be data distributed from the Internet or the like. In this case, existing lyrics or original lyrics are assigned to the melody part of the performance data, and the vocal information is used. It is also possible to make the set virtual singers sing.

(3) The note number or the like may be stored in the message of the system exclusive message. The pitch at the time of reproducing the lyrics uses the pitch included in the note-on signal even if the pitch is written in the system exclusive message. You may do so.

(4) In the above embodiment, the singing device is DT
Although the case of application to M has been described, the present invention is not limited to this, and may be used for an electronic musical instrument or voice response device capable of outputting a singing sound, or an amusement device such as a game machine or karaoke.

[0062]

As described above, according to the present invention,
Of the phonemes that make up the syllable, the phoneme that faces the consonant is uttered in time with the note generation timing, so that a natural singing can be realized by the virtual singer that matches the accompaniment.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a configuration of a DTM system according to an embodiment of the present invention.

FIG. 2 is a block diagram showing an electrical configuration of the DTM system according to the embodiment.

FIG. 3 is a flowchart showing vocal data generation processing according to the embodiment.

FIG. 4 is a diagram showing a table of vocal data used in the embodiment.

FIG. 5 is a diagram showing a configuration of a system exclusive message used in the embodiment.

FIG. 6 is a flowchart showing a music reproduction process according to the embodiment;

FIG. 7 is a flowchart showing lyrics reproduction processing according to the embodiment;

FIG. 8 is a diagram showing MIDI data and syllable utterances corresponding to notes according to a specific example.

FIG. 9 is a diagram illustrating utterance of syllables with respect to a note length according to a specific example.

FIG. 10 is a diagram showing MIDI data and syllable utterances corresponding to notes according to the related art.

FIG. 11 is a diagram illustrating utterance of syllables with respect to a note length according to the related art.

[Explanation of symbols]

1 ... PC, 28 ... Keyboard, 29 ...
Mouse, 31 ... tone generator, 32 ... M
IDI sound source, 33 ... singing sound source, 50 ... speaker, 60 ... MIDI keyboard

Claims (10)

[Claims] 1. A vocal data generating apparatus for generating vocal data corresponding to a melody and lyrics, a lyrics assigning means for assigning a syllable of the lyrics to a note corresponding to the melody, dividing the syllable into phonemes, Of the phonemes that make up
Sounding timing data generating means for generating sounding timing data corresponding to each syllable so as to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note; and the phoneme string data of the syllable, the sounding timing data and the syllable. A vocal data generating device, comprising: data output means for generating corresponding pitch data as vocal data and outputting the vocal data. 2. A vocal data generating apparatus for generating vocal data corresponding to a melody and lyrics, a lyrics assigning means for assigning a syllable of the lyrics to a note corresponding to the melody, dividing the syllable into phonemes, and Of the phonemes that make up
Sounding timing data generating means for generating sounding timing data corresponding to each syllable so as to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note; and the phoneme string data of the syllable, the sounding timing data and the syllable. A vocal data generation device comprising: data output means for generating corresponding pitch data as vocal data, and outputting the vocal data included in a system exclusive message. 3. The vocal data generation device according to claim 1, wherein said data output means outputs said vocal data separately for each chorus. 4. The vocal data generation device according to claim 1, wherein said data output means outputs said vocal data divided for each phrase. 5. The vocal data generation device according to claim 1, wherein said data output means outputs said vocal data separately for each breathing section. 6. The vocal data generating device according to claim 1, wherein said data output means outputs said vocal data for each bar. 7. A vocal data generating method for generating vocal data corresponding to a melody and lyrics, a lyrics assigning step of assigning a syllable of the lyrics to a note corresponding to the melody, dividing the syllable into phonemes, and Of the phonemes that make up
A sounding timing data generating step of generating sounding timing data corresponding to each syllable in order to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note; and the phoneme string data of the syllable, the sounding timing data and the syllable. A vocal data generating method, comprising: generating corresponding vocal data as vocal data; and outputting the vocal data. 8. A vocal data generating method for generating vocal data corresponding to a melody and lyrics, a lyrics assigning step of assigning a syllable of the lyrics to a note corresponding to the melody, dividing the syllable into phonemes, and Of the phonemes that make up
A sounding timing data generating step of generating sounding timing data corresponding to each syllable in order to match the sounding timing of the vowel phoneme to the sounding timing corresponding to the note; and the phoneme string data of the syllable, the sounding timing data and the syllable. A vocal data generating method, comprising: generating corresponding pitch data as vocal data; and outputting the vocal data included in a system exclusive message. 9. A singing sound synthesizer for singing lyrics to a virtual singer in accordance with an accompaniment, wherein lyrics are assigned to musical notes of accompaniment data and melody data, and the vowel of each syllable is pronounced at the timing of the musical note. Data output means for outputting vocal data including pronunciation timing data, syllable phoneme string data, and pitch data corresponding to the syllable, data control means for transmitting these data during reproduction; and the data An accompaniment sound source for generating an accompaniment sound by receiving the accompaniment data transmitted by the control means; and receiving the vocal data transmitted by the data control means, and forming a syllable at a pitch corresponding to the note based on the vocal data. A singing sound synthesizer comprising: a singing sound source to be pronounced. 10. A singing sound synthesizer for singing lyrics to a virtual singer in accordance with an accompaniment, wherein lyrics are assigned to musical notes of accompaniment data and melody data, and the vowel of each syllable is pronounced at the musical note timing. Data output means for outputting the vocal data including the sounding timing data, syllable phoneme string data and pitch data corresponding to the syllable in a system exclusive message, and transmitting these data during reproduction Data control means, an accompaniment sound source for generating an accompaniment sound upon receiving the accompaniment data transmitted by the data control means, and receiving vocal data in the system exclusive message transmitted by the data control means;
A singing sound source for producing a syllable at a pitch corresponding to the note based on the vocal data.