JP2002149179A - Device and method for generating melody and recording medium recorded with melody generating program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To generate a natural melody to match the lyrics. SOLUTION: The various conditions such as the key and the first note, etc., of the melody to be generated are inputted by the function A of a 'condition input'. The lyrics are inputted by a word processor function, etc., by the function B of a 'lyrics input'. The lyrics are divided into words by the function C of 'division by word'. The divided words are also divided into syllables by the function D of 'division by syllable'. Rhythm is generated by syllable as unit in accordance with the inputted lyrics by the function E of 'rhythm generation'. Intonation information which indicates the height of the note by syllable when the word is uttered is stored in a word intonation table. Intonation information corresponding to the divided word is extracted by referring to the word intonation table by the function F of 'word intonation table reference'. The high and low change of a musical interval is obtained at each syllable in accordance with the generated rhythm through the use of intonation information at each word by the function G of 'interval high/low progress generation'. The generation of the melody is completed by the function H of 'melody completion'.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a melody generating device, a melody generating method, and a recording medium on which a melody generating program is recorded.

[0002]

2. Description of the Related Art Conventionally, in the field of computer music and the like, there is a technique of an automatic tune for automatically generating a melody as disclosed in, for example, JP-A-10-49154. This automatic song processing consists of the composition of the passage, the number of phrases in the passage, the number of bars in the phrase, the melody data in the bar (pitch at the beginning and end of the bar), etc., corresponding to the time signature and genre of the song. Song template is used. Then, a melody is generated in accordance with the music template selected by the user and a predetermined music rule.

[0003]

However, the conventional automatic music technique requires a certain degree of musical knowledge, such as the need to consider the melody configuration when selecting a music template. It was not very suitable for composing. In addition, it is difficult to compose a melody that matches the lyrics unless the user is good at songwriting, such as when the user wants to add a melody to the lyrics after writing the lyrics. However, the conventional device does not consider the relationship with the lyrics as described above, and it is still insufficient to create a melody that matches the lyrics.

[0004] It is an object of the present invention to be able to generate a natural melody that matches lyrics with a simple operation.

[0005]

[Means for Solving the Problems] Here, a song that is easy to sing,
A song that is familiar to the ear has some relationship between the intonation of the words in the lyrics and the pitch up or down of the melody or the rhythm of the melody. For example, considering the lyrics "Let's go over the hill", if you read these lyrics as words, the intonation shown in FIG. In FIG. 8, the vertical position of a circle attached to the hiragana indicates intonation (to raise or lower the pitch (tone) of the sound: intonation).

When this is read with inflections as shown in FIG. 8B, it gives the impression that a person who is not fluent in Japanese reads it. It may be pronounced in some parts of the world, but it may be inconsistent with the inflection of the standard language, and in some cases it may be confusing or have a different meaning.

[0007] The same applies when this is applied to the lyrics of a song.
If you add a melody that goes against the intonation of the original word,
The meaning of the words is lost, and it gives a very unnatural impression. In some cases, such effects are intentionally aimed at, but in general, it is more natural to match the up-and-down progression of the melody with the original intonation of the word. Therefore, the present invention focuses on the above points, and is configured as follows so as to generate a melody according to the intonation of the words constituting the lyrics.

[0008] A melody generating device according to claim 1 of the present invention comprises:
Lyric input means for inputting lyrics, word decomposing means for decomposing the lyrics into words when the lyrics are input or after the lyrics are input, and inflection of the words decomposed by the word decomposing means. Pitch information generating means for generating corresponding pitch information.

According to the melody generating apparatus of the first aspect configured as described above, pitch information corresponding to the inflection of the words constituting the lyrics can be obtained, so that a natural melody matching the lyrics can be created. Can be. In addition, it becomes easy to give a variation to the melody by analyzing the words.

[0010] The melody generating device according to claim 2 of the present invention comprises:
Inflection information storage means for storing intonation information indicating each inflection of a plurality of words, lyrics input means for inputting lyrics, and when the lyrics are input by the lyrics input means or after the lyrics are input, the lyrics are read. Word decomposition means for decomposing into words, pitch information generation means for reading intonation information corresponding to the words decomposed by the word decomposition means from the intonation information storage means, and generating pitch information according to the intonation information , Is provided.

According to the melody generating apparatus of the second aspect configured as described above, in addition to the operation and effect of the first aspect, the user does not need to be conscious of the inflection of the word, so that the melody is created. Effective support can be provided.

According to a third aspect of the present invention, there is provided a melody generating apparatus.
Inflection information storage means for storing intonation information indicating each inflection of a plurality of words, lyrics input means for inputting lyrics, and when the lyrics are input by the lyrics input means or after the lyrics are input, the lyrics are read. Word decomposing means for decomposing into words, syllable decomposing means for decomposing words decomposed by the word decomposing means into syllables, reading intonation information corresponding to the words decomposed by the word decomposing means from the intonation information storage means And pitch information generating means for generating pitch information corresponding to the intonation information in accordance with the syllable decomposed by the syllable decomposing means.

According to the melody generating device of the third aspect configured as described above, in addition to the functions and effects of the first and second aspects, it is easy to generate a fine melody in syllable units.

[0014] A melody generating device according to a fourth aspect of the present invention comprises:
Inflection information storage means storing inflection information indicating each inflection of a plurality of words, and word decomposition means for decomposing the lyrics into words when the lyrics are input by the lyrics input means or after the lyrics are input, A syllable decomposing means for decomposing the word decomposed by the word decomposing means into syllables, a rhythm generating means for generating rhythm information based on the syllable decomposed by the syllable decomposing means, and a word decomposed by the word decomposing means Reading out the intonation information corresponding to the intonation information storage means, and generating pitch information according to the intonation information, corresponding to the rhythm information generated by the rhythm generation means, pitch information generation means,
It is characterized by having.

According to the melody generating apparatus of the fourth aspect configured as described above, in addition to the functions and effects of the first and second aspects, the rhythm can be easily created and the melody corresponding to the rhythm can be generated. .

A melody generating device according to a fifth aspect of the present invention provides:
An inflection information storage means storing inflection information indicating each inflection of a plurality of words, wherein the plurality of words are independent words meaningful only by the words, and dependent words used successively to other independent words. The independence word is managed by an absolute value, the pitch of the sound is managed by an absolute value, and the pitch of the dependent word is managed by a relative value from a previous independent word to store the intonation information. Inflection information storage means, lyrics input means for inputting lyrics, and sound for reading inflection information corresponding to the words of the input lyrics from the inflection information storage means and generating pitch information in accordance with the intonation information High-information generating means.

According to the melody generating device of the fifth aspect configured as described above, in addition to the functions and effects of the first and second aspects, the pitch of the sound of the dependent word following the independent word is changed. Processing for appropriately setting the pitch according to the pitch of the sound is facilitated.

Each of the melody generating methods according to claims 6 to 9 of the present invention respectively corresponds to the melody generating apparatus according to claims 1 to 4, and has the same operation and effect as the corresponding claims. Can be Claims 10 and 1 of the present invention.
The recording medium of claim 1 and claim 12 is claim 1 (or claim 6), claim 4 (or claim 9) and claim 5
The melody generation programs corresponding to the respective claims are recorded. By executing these melody generation programs, the same operational effects as those of the corresponding claims can be obtained.

[0019]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram of an embodiment in which the melody generating device of the present invention is constituted by a personal computer and software. The personal computer includes a CPU 1, a ROM 2, a RAM 3, a display 4, a keyboard 5, a mouse 6, an external storage device 7, a sound source 8, a sound system 9, a communication interface 10, and a MIDI interface 11. Various interfaces of the display 4, the keyboard 5, the mouse 6, and the external storage device 7 are not shown. The sound source 8 is implemented by a sound card or the like, and includes a D / A converter and the like.

The CPU 1 includes an OS installed in, for example, a hard disk drive (HDD) of the external storage device 7.
(Operating system) performs various processes using the working area of the RAM 3. For example, a process of outputting data to the display 4 and a process of inputting data in accordance with an operation of the keyboard 5 and the mouse 6 are performed. In addition, a melody corresponding to the inputted lyrics is generated by executing the melody generation program. When a melody is reproduced, a timer interrupt process of a built-in timer is performed, data of various notes in the melody data is reproduced, and data corresponding to each performance is output to a sound source. And the sound source 8 is a CPU
A digital tone signal corresponding to the data input from 1 is generated and converted into an analog audio signal, and a tone is generated by a sound system 9 such as an amplifier or a speaker.

The external storage device 7 includes a hard disk device (HDD), a floppy disk device (FDD), and a CD.
A ROM device, a magneto-optical disk (MO) device, a digital versatile disk (DVD) device, etc., for example, a melody generation program is supplied from the external storage device 7 to the RAM 3. The external storage device 7 is also used to store a table including intonation information indicating inflection of a word (a “word intonation table” described later). Further, the external storage device 7 is used to store the input lyrics and the created melody data corresponding thereto. Note that MI
The DI interface 11 transmits and receives various data to and from another MIDI device. For example, the melody data can be output to the MIDI device.

Further, it is connected to a communication network 20 such as a LAN (local area network), the Internet or a telephone line via the communication interface 10 and receives distribution of various data such as a melody generation program and a word inflection table from the server computer 30. You can also do so. In this embodiment, the melody generation program and the word intonation table are stored in a hard disk device (HDD) of the external storage device 7, and the CPU 1 develops the melody generation program of the hard disk device in the RAM 3, and the RAM 3 To execute the melody generation process.

FIG. 1 is a functional block diagram of a main part of the melody generating apparatus according to the embodiment. Each function is realized by executing a melody generating program. Function A of “condition input” is tonality of the melody to be generated, pitch of the first sound,
Alternatively, it manages input processing of various conditions such as beats (beats per bar) for rhythm generation. Note that the function A of the “condition input” may be appropriately executed at a necessary point in time, as indicated by a dashed arrow in FIG.

Next, a function B of "lyric input" manages input processing of lyrics input from the keyboard 5 by, for example, a word processor function, and stores the input lyrics data in the RAM 3.
In a predetermined area. Alternatively, already created lyrics data may be read from the external storage device 7 or distributed from the server computer 30 via the communication interface 10. For example, as an example,
Enter "Overcome, let's go." The function C of "decompose for each word" decomposes the input lyrics into words and stores the decomposed words together with data such as the parts of speech in a predetermined area of the RAM in association with the original lyrics. As the function C of "decompose for each word", a kanji conversion function such as a word processor may be used to automatically decompose. In the above input example, "oka", "wo", "koe", "iko", "~ u", "~"
Yo ”. Words to which "~" is attached are dependent words described later.

Note that, as shown by the dashed arrow in FIG.
The function B of “lyric input” and the function C of “decompose into words” may be executed in parallel. For example, one word constituting the lyrics is input by the function B of “lyric input”, and every operation of specifying a word delimiter by the function C of “decompose into words” is accepted for each input of the word, and the lyrics data is received. In addition, a method of inputting data of a word delimiter may be used.

The function D of "decompose into syllables" is to decompose each of the decomposed words into syllable units, and associate data such as the number of decomposed syllables (the number of syllables) with the original word.
M is stored in a predetermined area. In this example, "Oka" →
“O ・ ka”, “to” → “o”, “koe” → “ko ・
e ”,“ iko ”→“ i · ko ”,“ ~ u ”→“ u ”,
Decompose as "~ yo" → "yo". In the process of decomposing into syllables, a syllable delimiter may be input manually while selecting a word, or may be automatically decomposed. Further, syllable information may be stored in a word intonation table as described later, and the syllable may be decomposed by referring to this word intonation table. Further, information such as an accent dictionary may be used.

The function E of "generate rhythm" assigns a rhythm in units of disassembled syllables in accordance with the input lyrics, and generates a rhythm as shown in FIG. 7, for example. As the rhythm assignment process, for example, a note length within a predetermined range may be randomly generated using a random number,
The rhythm may be generated with reference to a rhythm variation table storing various kinds of rhythms created in advance.

The function F of "refer to the word intonation table" refers to the word intonation table and extracts intonation information corresponding to the decomposed word. This intonation information is information indicating the pitch of the syllable when the word is pronounced (increase or decrease in tone), as described later. The function G of “generate up-and-down progression of pitch” uses the intonation information for each word to determine the up-down variation of the pitch for each syllable corresponding to the generated rhythm. Then, note information is generated based on the rhythm and the pitch by the function H of “complete melody”, and the generation of the melody is completed.

Next, the intonation information will be described. In this embodiment, words that make sense only by the word are classified as “independent words”, and words that make sense by using them after other independent words are classified as “dependent words”. Put another way, independent words can be the beginning of a clause, and dependent words cannot be the beginning of a clause. Classification is performed in this way, and for independent words, one corresponding to the pitch of the sound
The absolute values of the five steps of ~ 5 are associated with each syllable as intonation information, and for the dependent word, the five steps of -2 to +2 corresponding to the amount of pitch change from the last syllable of the previous word The relative value is associated as intonation information.

FIG. 3 is a diagram showing an example of the word intonation table. FIG. 3 shows a specific example of a portion which is referred to when the words "overcome, go ahead" are input. The word (noun) of “oka” stores data of “o” and “ka” corresponding to the syllables “o” and “ka”, and the intonation information of each syllable is stored as “3” and “4”. Have been. Also, the words (verbs) of "i-ku"
"I" and "k" corresponding to syllables "i" and "ku"
u ”is stored, and the intonation information of each syllable is stored as“ 3 ”and“ 4 ”. In addition,
The word (verb) of “ru” stores data of “ko”, “e” and “ru” corresponding to the syllables “ko”, “e” and “ru”, and the inflection information of each syllable is “3”. , "4" and "4". In addition, the data (o) corresponding to the syllable “wo” is stored in the word (particle) “to wo”, and the inflection information of the syllable is stored as “± 0”. In addition, the data (u) corresponding to the syllable “u” is stored in the word (particle) of “to u”, and the inflection information of the syllable is stored as “−1”. further,"~
The word (particle) of “yo” includes “y” corresponding to the syllable “yo”.
The data “o” is stored, and the syllable intonation information is stored as “−1”.

When the word intonation table is created, the following may be performed. Independent words are further categorized as those with or without inflection. Unutilized items may be different from linguistic part of speech classification, but are all classified as nouns. On the other hand, if there is a conjugation, the five-stage conjugation verb, the la line transformation conjugation verb, the adjective verb,
Classify as adjectives. Note that, in FIG. 3, it is shown as a "verb" for simplicity. In addition, for words that have been utilized, the utilized endings are clearly registered in the word intonation table. For example, “kor” clarifies the inflection ending of “koru”. In practice, the inflected ending may be combined with the dependent word, so rather than registering as
It is desirable to register "koe / ru" so that the consonant and vowel of the inflected ending "ru" can be separated.

In some cases, the pronunciation of the same word has a different meaning, but in such a case, each is registered. In addition, Chinese characters may be added to the contents of the table to clarify the meaning of the word. Dependent words are classified as those used in combination with independent words, such as particles and auxiliary verbs, but all of them are shown as “particles” in FIG. 3 for simplicity. However, in practice, some of them involve inflections, so it is better to classify them as "particles" and "auxiliary verbs" (classified according to inflection).

FIG. 4 is a diagram conceptually showing an example of a reference result of the word intonation table. Lyrics obtained by chaining the intonation information of each word of the input words "Overcome, let's go" It represents the whole intonation. Here, in the following description, a numerical value assigned to each syllable indicating the inflection of the lyrics, for example, the syllable "yo" (syllable of the word "-yo") in FIG.
Is referred to as “lyric inflection information” in distinction from “inflection information” in the word intonation table. It should be noted that the lyrics intonation information "1-5" and the five horizontal lines shown in the figure do not directly correspond to the staff of the musical score.

As shown in the figure, "o.ka", "ko.
The lyrics inflection information of the syllables of the independent words “e” and “i · ko” are the inflection information of the word inflection table, respectively, but the lyrics inflection information of the syllables of the dependent words “o”, “u” and “yo” , The amount of change in the syllable inflection information of the immediately preceding syllable “±
0 "," -1 "," -1 "(inflection information), and the seventh syllable" ko "is the word" i.
This corresponds to the ending change of “ku” of “ku”, and the inflection information of such changed ending also has the same value as the ending before the change registered in the word intonation table (“ku” in this example). (“4”).

Next, a description will be given of an embodiment in which a pitch change is determined by the function G of "generate pitch up / down" based on such lyrics intonation information. If the pitch is simply determined in accordance with the five-step numerical value of the lyrics inflection information, only a fixed pitch change with a narrow pitch change can be obtained, which is boring. Therefore, in this embodiment, the pitch change uses only the information of the direction of the vertical change corresponding to the change of the numerical value of the lyrics intonation information. For example, in the example of "Let's Overcome Oka" in FIG. 4, between the first syllables of "o-ka"
Since the lyric intonation information changes from "3" to "4" by "+1", the pitch is increased between these syllables. In the next “ka · o”, the change of the lyrics intonation information is “± 0”. However, if the pitch is not changed as it is, a completely flat melody is obtained, so that a certain width change is allowed. In order to realize such an allowance, for example, a random number is used as follows.

First, when the change in the lyrics intonation information is positive, the random number value is Rnd, the absolute value of the changes in the lyrics intonation information is ΔP, the set value is RndMax, and the pitch change ΔKey.
Is determined by Δkey = ΔP × (Rnd + 1) × RndMax. When the change in the lyrics inflection information is minus, the random number value is Rnd, the absolute value of the changes in the lyrics inflection information is ΔN, and the pitch change ΔKey is obtained by Δkey = ΔN × (Rnd−1) × RndMax. When the change in the lyrics intonation information is ± 0, the random number value is set to Rnd, and the pitch change ΔKey is obtained by ΔKey = Rnd × RndMax.

Here, the change range of the random number Rnd is −1 ≦ R
By setting nd ≦ + 1 and defining the range of change by the value of the set value RndMax, the degree of variation in pitch change can be controlled. For example, when RndMax is set to “2”, ΔP = ΔN = 1 in this example,
If the change in the lyrics inflection information is positive, 0 ≦ Δkey ≦
4 if the change in the lyrics intonation is negative-
4 ≦ Δkey ≦ 0, and when the change of the lyrics intonation information is ± 0, −2 ≦ Δkey ≦ 2. Note that a value obtained by rounding off the decimal part of Δkey is set as an actual pitch change. In addition,
Any number of values of RndMax may be used, but it is preferable that the value of RndMax is set to a value that does not greatly change the pitch (for example, within 4).

FIG. 5 shows an example of a pitch change obtained by the above processing.
FIG. 6 shows an example of a melody generated by adding a pitch change to the generated rhythm. The pitch of the first syllable is set by the user in advance,
In this example, the sound is set to “do”. The pitch of the first syllable may be set by the automatic music composition device. Also, in this example, the unit of the pitch change is one whole note,
The tonality is input in advance, and the pitch is corrected so as to be a chromatic sound corresponding to the key. For example, in this example, the set tonality is C major. Therefore, as shown in FIG. 5, the pitch change between syllables “i-ko” is “+3”,
The sound for all three sounds from “do” to the top is “♯FA”,
The sound of the syllable "ko" has been modified to "Fa" according to the scale of C major.

When the pitch of the melody is determined as described above, the generated melody note and the input lyrics are displayed on the display 4, and note information is generated based on the rhythm and pitch. The note information corresponds to performance data such as automatic performance, and is composed of note length data (or gate time), note data (or key code), timing from the beginning of the music (or duration data between notes), and the like. Enable melody playback.

In the above embodiment, five levels of absolute values are given as the inflection information of the independent word. Further, for example, “0”, “1”, and “1” may be used as inflection information for each syllable “ko”, “e”, and “ru” of the word “koeru”. May be expressed by a relative value that relatively indicates the height from the syllable. Further, for example, each syllable "ko" of the word "koeru",
“0”, “1” as the intonation information of “e” and “ru”
For example, the inflection may be expressed by a relative value that relatively indicates the level from the “preceding syllable”, such as “0”. In addition, regardless of whether an absolute value or a relative value is used as the intonation information, the entire intonation of the word may be shifted up or down with respect to the syllable at the end of the immediately preceding word (the shift amount is selected by the user, for example). Good.

In the above embodiment, the rhythm is assigned to each of the decomposed syllable units. However, it is preferable that the compatibility between the lyrics and the tune is obtained in consideration of the time signature, the break between words, and the like. For example, as a method of considering the time signature, one method is that one word does not straddle a bar line due to the time signature. On the other hand, depending on the length of the word, the length of the bar, and the specification of the rhythm, it may be difficult or even unnatural to prevent the word from straddling the bar line. Therefore, it may be effective to prevent a word from straddling a bar line and preventing only the first or last syllable of the word from entering another bar. In addition, in the case of a song with a break, the phrase itself of the song may straddle a bar line. It becomes more natural to use virtual bar information shifted by the amount of afteract.

On the other hand, the note length of the rhythm may be simply generated by random numbers, but it is also possible to refer to a rhythm table or the like so as to reflect the inherent rhythmic characteristics of the word. Good. Or there is a simpler approach. For example, syllables that become unnatural when stretched or syllables that become unnatural when shortened are related to pronunciation.
For example, it is unnatural to extend the syllable of "n". Therefore, it is also an effective method to prepare and refer to a table such as whether the syllable is unnatural when extended or unnatural when contracted.

The various processing methods described above are suitable so that the user can select them as desired. However, 2
If the above methods conflict, the user is allowed to make an alternative choice.

In the embodiment, the pitch is determined in accordance with the tonality. However, the chord progression is given, and the pitch is given so that the sound of the chord progression on the available note scale is adopted. Is also good. Alternatively, a melody generation parameter (for example, dynamics or a tone range) may be given, and the pitch may be given based on the parameter.

After the melody is generated, the user may listen to the melody, and if he / she does not like the melody, the process for determining the pitch change may be executed again to generate the melody again (change the random value).

Further, the melody generation method of the present invention may be applied to application software for composition support or automatic music composition, and portions of the music that use the melody generation method of the present invention may be provided. Also, in this case,
The user may be able to specify where to use the melody generation method of the present invention.

In the processing for determining the pitch, for example, the Rn
A plurality of pitches meeting predetermined conditions, such as the condition of the preceding expression when a plurality of dMax are set and the upper and lower conditions for correcting the pitch to an available note scale, are generated as candidates, and the plurality of candidates are generated. (Display or pronunciation)
Either one may be made to be selected by the user.

The rhythm may be manually input without automatically generating it. For example, various rhythm generation methods can be applied, such as a user inputting note information or tapping a predetermined key to input a rhythm by tapping. Alternatively, lyrics may be input to an existing rhythm, and a pitch may be given to the lyrics. In the case of automatically generating a rhythm, any method may be used.

In a system comprising a server and a terminal,
Lyrics or lyrics and rhythm are input from the terminal, transmitted to the server, and the server generates the rhythm and melody or generates the melody by the device or method of the present invention, and distributes the melody to the terminal. Is also good. On this occasion,
You may charge. The terminal may be a computer or a mobile phone.

The lyrics to be input are not limited to Japanese but may be foreign.

[0051]

According to the first aspect of the present invention, there is provided a melody generating apparatus.
According to the melody generation method of the present invention or the execution of the melody generation program recorded on the recording medium of claim 9, pitch information corresponding to the inflection of the words constituting the lyrics can be obtained.
You can create a natural melody that matches the lyrics. In addition, it becomes easy to give a variation to the melody by analyzing the words. Also, by replacing the intonation information with another one, the atmosphere of the melody can be changed.

According to the melody generating apparatus of the second aspect or the melody generating method of the sixth aspect, in addition to the effect of the first aspect, the user does not need to be conscious of the inflection of the word, so that the melody is generated. Effective support can be provided.

According to the melody generation device of the third aspect or the melody generation method of the seventh aspect, in addition to the effects of the first and second aspects, it is easy to generate a fine melody in syllable units.

According to the melody generating device of claim 4, the melody generating method of claim 8 or the execution of the melody generating program recorded on the recording medium of claim 10.
In addition to the effects of (2) and (3), the rhythm can be easily created and a melody corresponding to the rhythm can be generated.

According to the melody generation device of the fifth aspect or the execution of the melody generation program recorded on the recording medium of the twelfth aspect, in addition to the effects of the first and second aspects, in addition to the effects of the independent words, the sound of the dependent word following the independent word is obtained. This facilitates the process of setting the pitch to be appropriate depending on the pitch of the sound of the preceding independent word.

[Brief description of the drawings]

FIG. 1 is a main functional block diagram of a melody generating device according to an embodiment of the present invention.

FIG. 2 is a block diagram of an embodiment in which the melody generating device of the present invention is configured by a personal computer and software.

FIG. 3 is a diagram illustrating an example of a word intonation table according to the embodiment.

FIG. 4 is a diagram conceptually illustrating an example of a reference result of the word intonation table in the embodiment.

FIG. 5 is a diagram illustrating an example of a pitch change obtained by processing according to the embodiment.

FIG. 6 is a diagram illustrating an example of a melody generated by adding a pitch change to a rhythm in the embodiment.

FIG. 7 is a diagram illustrating an example of a rhythm in the embodiment.

FIG. 8 is a diagram illustrating an example of inflection of lyrics;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... CPU, 2 ... ROM, 3 ... RAM, 7 ... External storage device

Claims (12)

[Claims] 1. Lyrics input means for inputting lyrics, word decomposition means for decomposing the lyrics into words when the lyrics are input by the lyrics input means or after the lyrics are input; And a pitch information generating means for generating pitch information corresponding to the inflection of the word. 2. Inflection information storage means for storing intonation information indicating each inflection of a plurality of words; lyrics input means for inputting lyrics; and when lyrics are input by the lyrics input means or when lyrics are input. Word decomposing means for decomposing the lyrics into words after reading the intonation information corresponding to the words decomposed by the word decomposing means from the intonation information storage means, and generating a pitch information according to the intonation information A melody generating device comprising: a high information generating unit. 3. Inflection information storage means for storing intonation information indicating each inflection of a plurality of words; lyrics input means for inputting lyrics; and when lyrics are input by the lyrics input means or when lyrics are input. Word decomposing means for decomposing the lyrics into words, and syllable decomposing means for decomposing the words decomposed by the word decomposing means into syllables; and inflection information corresponding to the words decomposed by the word decomposing means. A melody generating device comprising: a pitch information generating unit that reads out from an information storage unit and generates pitch information corresponding to the intonation information in accordance with the syllable decomposed by the syllable decomposing unit. . 4. Intonation information storage means for storing intonation information indicating each inflection of a plurality of words, and when the lyrics are input by the lyrics input means or after the lyrics are input, the lyrics are decomposed into words. Word decomposing means; syllable decomposing means for decomposing words decomposed by the word decomposing means into syllables; rhythm generating means for generating rhythm information based on the syllable decomposed by the syllable decomposing means; Reading the intonation information corresponding to the word decomposed in the inflection information storage means, and generating pitch information corresponding to the intonation information in correspondence with the rhythm information generated by the rhythm generation means And a melody generating device. 5. An inflection information storing means for storing intonation information indicating each inflection of a plurality of words, wherein the plurality of words are stored in an independent word having a meaning only by the word and another independent word. It is classified into at least two types of dependent words to be used successively, the pitch of the independent word is managed by an absolute value, and the pitch of the dependent word is managed by a relative value from the previous independent word. Inflection information storage means for storing intonation information, lyrics input means for inputting lyrics, inflection information corresponding to the words of the input lyrics is read from the intonation information storage means, and a pitch corresponding to the inflection information is read. A melody generating device, comprising: a pitch information generating means for generating information. 6. A method of inputting lyrics and decomposing the lyrics into words when inputting the lyrics or after inputting the lyrics, and generating pitch information corresponding to the intonation of the decomposed words. Melody generation method. 7. An inflection information storage unit that stores intonation information indicating each inflection of a plurality of words, and inputs lyrics and converts the lyrics into words when the lyrics are input or after the lyrics are input. A melody generating method, comprising: decomposing, reading inflection information corresponding to the decomposed word from the intonation information storage means, and generating pitch information corresponding to the intonation information. 8. An inflection information storage unit that stores intonation information indicating each inflection of a plurality of words, and inputs lyrics and converts the lyrics into words when the lyrics are input or after the lyrics are input. Decomposing, decomposing the decomposed word into syllables, reading intonation information corresponding to the decomposed word from the intonation information storage means, and generating pitch information according to the intonation information,
A melody generating method for generating a melody corresponding to the decomposed syllable. 9. An inflection information storage unit that stores intonation information indicating each inflection of a plurality of words, and inputs lyrics and converts the lyrics into words when the lyrics are input or after the lyrics are input. Decomposing, decomposing the decomposed word into syllables, generating rhythm information based on the decomposed syllable, reading inflection information corresponding to the decomposed word from the intonation information storage means, and responding to the intonation information. Pitch information,
A melody generating method, wherein the melody is generated in accordance with the generated rhythm information. 10. A process of inputting lyrics, a process of decomposing the lyrics into words when or after inputting the lyrics, and a process of generating pitch information corresponding to the intonation of the decomposed words. And a recording medium on which a melody generation program for executing by a computer is recorded. 11. A process of inputting lyrics, a process of decomposing the lyrics into words when inputting the lyrics or after inputting the lyrics, a process of decomposing the decomposed words into syllables, Generating rhythm information based on the inflection information, reading inflection information corresponding to the decomposed word from inflection information storage means storing inflection information indicating each inflection of a plurality of words, and responding to the inflection information. A recording medium recording a melody generation program for executing, by a computer, a process of generating pitch information in accordance with the generated rhythm information. 12. An inflection information storage unit that stores lyrics input processing, and inflection information corresponding to the words of the input lyrics, and inflection information indicating each inflection of a plurality of words. The plurality of words are classified into at least two types, an independent word meaningful only by the word, and a dependent word used after the other independent word, and the pitch of the independent word is managed by an absolute value. For the dependent word, a process of managing the pitch of the sound with a relative value from the previous independent word, reading the information from the intonation information storage means storing the intonation information, and generating pitch information according to the intonation information, Recording medium for storing a melody generation program for executing the program on a computer.