JP4139878B2 - Music data generation system, music data generation device, and music data generation method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a system for generating chorded music, and more particularly to a system for generating music in consideration of an input melody approach note.
[0002]
[Prior art]
Currently, software that is a singer-songwriter, for example, is commercially available to support user composition activities. This software is intended for pop-type music, and accepts melodic input via a personal computer or the like, and analyzes its constituent sounds to generate chords corresponding to melodies.
[0003]
[Problems to be solved by the invention]
However, this singer-songwriter software does not have a function for identifying approach notes such as progress sounds, stuttering sounds, embroidery sounds, mooring sounds, preceding sounds, and unusual sounds that often appear in music. For this reason, there may be cases where chording for a given melody is not suitable. Normally, the chording for a given melody is handled differently depending on which of the additional melodies is not the main one of the melody. Therefore, it is not possible to generate a chord perfectly adapted to the melody simply by generating a chord from a given string.
On the other hand, the following patent document 1 etc. exist as literature which handled the approach note.
[0004]
[Patent Document 1]
Japanese Patent Publication No.7-111636
[0005]
What is described in this publication considers whether the sound in the melody can be interpreted as a chord component or approach note when determining whether a chord raised as a chord setting sound can be set. is there.
[0006]
However, in Japanese Patent Publication No. 7-111636, when a chord candidate group that can be set is obtained, firstly, a sound that can be interpreted as an approach note is excluded from consideration when setting as a chord candidate group. Therefore, chord candidates for all constituent sounds must be obtained, and a huge number of chord candidate groups must be generated. Furthermore, the interpretation of approach notes generally has multiple interpretation issues. For example, even if the sound is the same, the interpretation of the approach note may be different when the preceding and following chord settings are taken into account. In this regard, in the method disclosed in Japanese Patent Publication No. 7-11636, the approach note is determined without considering the preceding and following chord settings, and therefore there is a possibility that the approach note may be determined incorrectly. For this reason, approach notes cannot be used sufficiently or correctly, and correct chords cannot be added to the melody.
[0007]
Accordingly, the present invention has been made paying attention to the above-described problems, and an object thereof is to provide a system that can perform correct chording with a small amount of calculation for a given melody. is there.
[0008]
[Means for Solving the Problems]
That is, in order to solve the above-mentioned problem, the present invention receives an input of a melody and determines an approach note in the common chord region from the melody. Then, a chord is generated based on the component sound excluding the determined approach note, and integrated with the melody so that it can be output as music data for viewing.To do. When determining an approach note in this way, the approach note may not be determined only within an area where chords are common (for example, one bar), so a temporary approach note that is assumed from a given melody. And a chord candidate generated based on the temporary approach note is generated. Then, it is determined whether or not the chord progression including the temporary chord candidate is suitable for the chord progression stored in the storage unit in advance. Correct the interpretation of the approach note for the common chord region to which is assigned.
[0009]
By configuring in this way, the chord is generated based on the component sound excluding the approach note, so that the generated volume can be reduced, and the chord corresponding to the melody is generated. It becomes possible to generate music data suitable for.In addition, since approach notes that cannot be determined only within the common chord area are determined based on the preceding and following common chord areas, the approach notes can be determined more reliably and chords suitable for the melody can be generated. become able to.
[0012]
Furthermore, when generating a chord in this way, a plurality of chords are generated in the common chord region based on the constituent sound excluding the approach note, and the degree of consistency with the melody, that is, the approach note from the melody is generated. A chord is extracted in consideration of the degree of overlap between the excluded component sound and the chord component sound.
[0013]
If comprised in this way, even if it is a case where a some chord can be produced | generated within a common chord area | region, it will become possible to produce | generate the natural music data which can adapt more to a melody.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 1, the music data generation system 1 according to the present embodiment includes a plurality of portable information terminals 2 constituted by mobile phones and the like, and music data connected to the portable information terminal 2 via a network 4. For example, when generating an incoming melody or a melody mail, the portable information terminal 2 accepts input of a melody, and the music data generation device 3 is an approach note from the melody. Musical data with chords is generated from the remaining constituent sounds by determining elapsed sound, stuttering sound, embroidery sound, mooring sound, preceding sound, and anomalous sound. Hereinafter, a specific configuration of the present embodiment will be described in detail.
[0015]
As shown in the block diagram on the left side of FIG. 2, the portable information terminal 2 transmits and receives data between the input receiving means 20 that receives input such as melody and the music data generating device 3 via the network 4 such as the Internet. Transmitting means 21 and receiving means 22, storage means 23 for storing various information including music data transmitted from the music data generating device 3, and output means 24 for outputting music data. Do it. In addition, the portable information terminal 2 has other functions for achieving the original functions of the portable information terminal 2, for example, a necessary configuration for functioning as a normal telephone.
[0016]
The input receiving means 20 includes operation keys provided in the portable information terminal 2 and applications for receiving inputs such as melody and music genre. The input accepting means 20 accepts the melody as individual pitch information, note value information, and the like, and the music genre is accepted by selecting a pre-registered music genre such as pop, jazz, or classic. . In addition, when the finally generated music data is transmitted to another person as a melody mail, the input receiving means 20 receives destination information such as a mail address of the other person.
[0017]
The transmission unit 21 transmits predetermined data to the music data generation device 3 via the network 4. For example, the melody, the music genre input via the input reception unit 20, and the mail of the portable information terminal 2 of itself Transmission destination information such as an address, model information of the portable information terminal 2 of itself is transmitted. The receiving means 22 receives music data transmitted from the music data generating device 3.
[0018]
The storage means 23 stores information necessary for operating the portable information terminal 2, and as shown in FIG. 3, an execution program area 230, a music data storage area 231, a terminal information storage area 232, a work area 233 etc. are provided to store information corresponding to these. Among these, the execution program area 230 stores an execution program for operating the portable information terminal 2, and specifically stores a program corresponding to the flowchart shown in FIG. In the music data storage area 231, music data received from the music data generating device 3 is stored together with identification information corresponding to the music data. Further, the terminal information storage area 232 stores identification information such as a mail address of the own terminal, model information of the own terminal, and the like.
[0019]
The output unit 24 includes a display output unit such as a display and a voice output unit such as a speaker. The output unit 24 displays and outputs the music data transmitted from the music data generation device 3, and the music data is output as audio. The output means outputs the music data so that it can be viewed. This display output and audio output incorporate a program that outputs them in synchronism with each other, for example, to change the display form in accordance with the progress of melody or chords. When used as an incoming melody, the music data is output upon detection of the incoming call.
[0020]
On the other hand, the music data generation device 3 automatically generates music data with chords based on the melody transmitted from the portable information terminal 2, and has the music theory knowledge, accompaniment pattern, file format as shown in FIG. The receiving means 31, the melody / key judging means 32, the approach note judging means 33, the chord generating means 34 as shown on the right side of FIG. , Accompaniment data generation means 35, integration means 36, media data generation means 37, transmission means 38, and the like. The information stored in the storage unit 30 will be described later.
[0021]
The receiving means 31 receives data transmitted from the portable information terminal 2 via the network 4, and transmits the melody, music genre, destination information, model information, and music data as melody mail. The destination information is received, and these data are temporarily stored in the storage means 30.
[0022]
The melody / key determination means 32 determines the melody and key of the input melody, and specifically performs the following processing. That is, in determining the key, first, the key including the tone name in the order of appearance of the pitch name is extracted, and thereafter, every time a new pitch name appears, the key candidates are narrowed down. After the last note is processed, if only one candidate key is left, the key is left as undecided. However, at this stage, the method of deciding is undecided, so that there is one possibility of adjustment for each of the long and short merits. In this state, the deciding method is determined as follows. That is, the melody is determined so that the last sound becomes the main sound. That is, if the last note of the melody is “do” by “moving de”, it is determined to be a major key, and if “la”, it is determined to be a minor key. In other cases, the main sound is determined so that the appearance frequency of the main sound, the subordinate sound, and the subgeneric sound increases by looking at the appearance frequency of each pitch name. If no candidates are found as a result of narrowing down the key, it is checked whether any key candidate remains by removing any one note name, and if it remains, it is determined as a key. Usually, in that case, it is in minor key, and the removed sound corresponds to the guided sound (semitone below the main sound). If a key candidate still does not remain, the appearance frequency of the pitch name is examined, and the same determination as the above-mentioned method of determination is combined with the key determination. In this case, the mode and key are determined at the same time. Confirm that the last note of the melody is the one in the main chord. If candidates remain as a result of narrowing down the key, the appearance frequency of the pitch name is examined, and after narrowing down the candidates, the melody and key are determined in the same manner as above.
[0023]
The approach note determination means 33 determines whether or not each constituent sound of the melody corresponds to the approach note for the melody whose melody and key are determined. This approach note determination means 33 stores music theory knowledge information related to the approach note in the storage means 30 in advance, and determines whether or not each constituent sound corresponds to the approach note in the common chord area based on this information. To do. Specifically, music theory knowledge information that can be determined as follows is stored in the storage means 30, and an approach note is determined based on this information. Note that the following music theory knowledge information represents a part of the music theory knowledge information, and stores a plurality of music theory knowledge information in addition to this.
[0024]
(1) Judgment of elapsed sound
If there are 3 sounds and 5 sounds in succession, if there are 3 sounds following the “Doremi” with the moving sound set to “Do”, the second “Re” is determined as the elapsed sound.
[0025]
(2) Judgment of embroidery sound
If there are three sounds following “Dredo”, the second “Dole” is determined as the embroidery sound.
[0026]
(3) Judgment of stuttering
For two notes with strong beats and two pitches (for example, “red” from the first beat of a measure), “re” may be interpreted as stuttering.
[0027]
(4) Judgment of noise
It is assumed that two sounds of two pitches (for example, “sodore ...”) that follow the jumping progress, one of “do” and “le” may be interpreted as an unusual sound.
[0028]
(5) Determination of mooring sound
In a common chord area, a sound that persists from the previous common chord area may be a chord component in the previous common chord area, and therefore may be interpreted as a mooring sound in the common chord area. Suppose there is.
[0029]
(6) Judgment of preceding sound
In a certain common chord area, a sound that continues to the next common chord area may be a chord constituent sound in the next common chord area, and therefore may be interpreted as a preceding sound in the common chord area. Suppose there is.
[0030]
In the judgment of the above approach note, all the judgments of stuttering, anomalous sound, mooring sound and preceding sound are interpreted as “possible”, but whether these sounds fall under the approach note. First, the combinations of interpretation of approach notes for all sounds are listed in a state where it is assumed that the note is an approach note. Then, an arbitrary pattern is selected randomly or arbitrarily from among them, the interpretation is interpreted as a temporary approach note, and the chord generation means 34 described below determines the approach note together with the chord generation process.
[0031]
The chord generating means 34 lists a plurality of chord candidate groups based on the constituent sounds of each melody excluding the constituent sounds interpreted as the temporary approach note. The default of the common chord area is set to one measure, but the user can arbitrarily change it. For chords that can be set for the common chord region, the absolute lengths of the sounds included in that region are used as an index, and the chord names that make up the sound having a long absolute time are listed. Then, when generating this chord, it is determined whether or not the chord candidate group is suitable based on the music genre transmitted from the portable information terminal 2.
[0032]
Specifically, if “Doremifa” exists in a certain chord setting section, either “Do” or “Mi” or “Le” or “Fa” is interpreted as an approach note for the interpretation of the approach note. Cannot be determined in the chord setting section. For this reason, “do” and “mi” are interpretation notes (a), and “re” and “fa” are interpretation notes (b). First, interpretation (a) Is determined based on a chord candidate in the immediately preceding chord setting section. If it is based on interpretation (a), “Le” and “Fa” are chord constituent sounds, and the second minor (Dm for C major) that has “Le Fa La” as the constituent sounds is listed as a chord candidate. . However, if V7 (G7 in C major) is set in the chord setting section immediately before, if the progression from V7 to Dm is not used in a certain music genre, Dm is not suitable as a chord candidate. That is, the interpretation (b) is adopted on the assumption that the interpretation (a) is adopted in itself.
[0033]
Then, the approach note determined as “possible” in this way is determined using the chord progression of the music genre, and then a chord is generated using the determined approach note. For this chord generation, the chord progression for each music genre used previously is not used, but a chord corresponding to the melody of the common chord region is independently generated. Here, when generating chords, a plurality of chords can be generated in the common chord region, but the chords are generated in consideration of the degree of matching that indicates the degree of commonness between the melody and each component of the chord.
[0034]
Then, one of chords allowed between the common chord areas is selected. The criteria for selection follows genre statistics of the music database based on the second order Markov model. That is, the partial chord progression that is used most frequently in terms of probability is selected as the first candidate in consideration of the continuation of the front and back at each time point. This usage frequency is based on the music genre input by the user. Further, the degree of matching obtained previously is also taken into consideration in the chord progression determination. If there are candidates below the next point in the order of appearance frequency in the database, about three candidates are retained, and in preparation for a case where the user requests later the possibility of another chord progression. The reason for using the second order Markov model is that the progression of music is considered to depend on the progression of chords before and after a certain point in time. However, the starting chord is a main chord in principle, and if the first sound is a sound other than the main chord, it is determined that it is a decoration sound. The second chord is determined based on the first-order Markov model, and the third and subsequent chords are determined based on the second-order Markov model. The final chord that arrives at the last note of the melody is the main chord, and before that is the genus chord.
[0035]
The accompaniment data generating means 35 extracts accompaniment data such as drum sounds for the chord progression selected in this way. Accompaniment patterns have their own characteristics depending on the music genre. Accompaniment patterns such as drum sounds and rhythms characteristic of each music genre are held in the storage means 30 and are selectively used from among them.
[0036]
The integration means 36 integrates the chords and accompaniments extracted in this way into a melody and combines them into a single piece of music data as a whole. Then, this music data is output to the media data generating means 37 so as to be viewable on the transmission source portable information terminal 2.
[0037]
The media data generation unit 37 converts the music data into a file format that the portable information terminal 2 as the transmission source can output. The storage means 30 stores file formats that can be output for each portable information terminal 2, reads out the model information transmitted from the portable information terminal 2, and converts the format into a file format corresponding to the model information. To do.
[0038]
The transmission means 38 transmits the music data generated in this way to the portable information terminal 2 that is the transmission source via the network 4, and outputs the music data so that it can be viewed there. When this music data is transmitted to a third-party portable information terminal 2 as a melody mail, the music data is transmitted to a mail address related to the transmission destination.
[0039]
Next, a music data generation flow in the music data generation system 1 configured as described above will be described with reference to FIGS.
[0040]
First, when a melody mail is transmitted to another person, as shown in the flow of the portable information terminal 2 in FIG. 5, a predetermined program is started on the portable information terminal 2 to input a melody or music genre using operation keys or the like. Acceptance and the input of the mail address of the melody mail transmission destination are accepted (step S1). Then, on condition that a transmission instruction is input (step S2), the melody, music genre, and e-mail address of the transmission destination are transmitted. At this time, transmission destination information such as the e-mail address of the portable information terminal 2 is transmitted. And model information are also read out and transmitted together (step S3).
[0041]
When the music data generating device 3 receives the information (step T1), first, the melody and key are determined based on the received melody (step T2). In the determination of the melody and key, first, keys that include them as intones are extracted in the order in which the pitch names appear, and thereafter, key candidates are narrowed down every time a new pitch name appears. After the last note is processed, if only one candidate key is left, the key is left as undecided. However, at this stage, since the method of deciding is undecided, there remains a possibility of adjustment for each of the long and short merits. In this state, the deciding method is determined as follows. That is, the melody is determined so that the last sound becomes the main sound. That is, if the last note of the melody is “do” by “moving de”, it is determined to be a major key, and if “la”, it is determined to be a minor key. In other cases, the main sound is determined so that the appearance frequency of the main sound, the subordinate sound, and the subgeneric sound increases by looking at the appearance frequency of each pitch name. If no candidates are found as a result of narrowing down the key, it is checked whether any key candidate remains by removing any one note name, and if it remains, it is determined as a key. Usually, in that case, it is in minor key, and the removed sound corresponds to the guided sound (lower semitone of the main sound), but if there is still no key candidate, check the appearance frequency of the pitch name and make the same judgment as the above-mentioned melody judgment Is combined with key judgment. In this case, the mode and key are determined at the same time. Confirm that the last note of the melody is the one in the main chord. If candidates remain as a result of narrowing down the key, the appearance frequency of the pitch name is examined, and after narrowing down the candidates, the melody and key are determined in the same manner as above.
[0042]
Then, a chord generation process is performed on the melody whose melody and key are determined in this way (step T3). In this chord generation process, first, as shown in FIG. 6, a melody approach note determination process is performed (step T30). In this approach note determination process, first, a chord candidate group is generated based on the approach note to be confirmed and the assumed approach note (step T31). Then, it is determined whether or not the chord progression by the generated chord candidate group is suitable for the chord progression of the selected music genre (step T32). The assumed approach note of the chord area is corrected to determine approach notes of all the common chord areas (step T33). Next, when the approach notes of all the common chord areas are determined in this way, the approach notes are removed from the melody and a chord is generated based on the remaining constituent sounds (step T34). The chord generation is performed separately from the chord progression of the music genre. When a plurality of chords are generated in the common chord area, a degree of consistency with the melody is given (step T35). Then, using this degree of matching, one chord is selected according to the music genre statistics based on the second order Markov model (step T36).
[0043]
Next, the accompaniment data added to the generated chord and the like is read from the storage means 30 (step T4). The accompaniment data is extracted by grasping the transmitted music genre and reading the accompaniment data corresponding to the music genre from the storage means 30.
[0044]
Then, music data is generated by integrating these melody, chords, and performance pattern data (step T5), and this music data corresponds to the model information so that it can be viewed on the portable information terminal 2 that is the transmission source. The file data is converted into media data (step T6). Then, the converted music data is confirmed to be transmitted to the transmission source portable information terminal 2 together with the identification information corresponding to the music data (step T7), and the music data which is the transmission result and the music data corresponding to the music data are confirmed. The identification information is stored in the storage means 30 (step T8).
[0045]
In the portable information terminal 2 that has received the music data (step S4), the music data is output for viewing (step S5), and if the user determines that the melody and the chord do not match, The input to that effect is permitted (step S6) and transmitted to the music data generating device 3 (step S7). In this transmission, identification information corresponding to the previously assigned music data is transmitted.
[0046]
Based on this, the music data generating device 3 generates music data having a chord different from the chord included in the music data corresponding to the identification information (steps T3 to T8). Then, after repeating these viewing and correction processes, when music data desired by the user is generated, the music data is transmitted as a melody mail to the third-party portable information terminal 2 or the like (step T9).
[0047]
As described above, in the above embodiment, the approach note is first determined from the melody, and the chord is generated based on the constituent sound excluding the approach note. Therefore, the chord can be generated with a small amount of calculation. In addition, it is possible to generate music data that more reliably matches the melody.
[0048]
Also, in this embodiment, when it is not possible to determine whether or not it is an approach note, first, a chord candidate group is generated by setting an approach note in a tentative form, and this chord candidate group is a general chord progression. If not, the approach note interpretation of the non-conforming common chord region is corrected, so the accuracy of approach note determination can be increased, and a chord that conforms to the melody can be generated. Will be able to.
[0049]
Furthermore, based on the approach note corrected in this way, a plurality of chords in the common chord region are generated, and in addition to this, a chord that considers the characteristics of the music genre is generated. It is possible to generate optimal music data in consideration of the chord setting (vertical chord setting) and the chord setting based on the chord progression (horizontal chord setting).
[0050]
In addition, this invention can be variously changed without being limited to the said embodiment.
[0051]
For example, in the above-described embodiment, the music data generation system 1 including the portable information terminal 2 and the music data generation device 3 has been described. However, the present invention is not limited to this, and it is possible to input a melody and generate chords with a single device. Processing may be performed. In this case, the “system” in the present invention is interpreted as a single device. In addition, the device in this case is not limited to a mobile phone or a personal computer, but can be applied to an electronic piano or the like.
[0052]
In the above embodiment, the case where an incoming melody, a melody mail, or the like is generated has been described. However, the present invention is not limited to this, and music data for appreciation or music data for composition support may be generated.
[0053]
Furthermore, in the above embodiment, when generating a chord, the melody and key are first determined. However, when inputting the melody, the melody is input after the user determines the melody and key in advance. Alternatively, the melody and key may be determined by default, and melody input may be accepted in this state.
[0054]
In addition, when a chord is generated, a plurality of chords are generated based on the finally determined approach note, and a chord according to the music genre is generated from the secondary Markov model. Of the plurality of chords generated in the common chord area, the chord having the highest degree of matching may be extracted.
[0055]
【The invention's effect】
In the present invention, the approach note is determined from the input melody, and the chord is generated from the constituent sound excluding the approach note. Therefore, the chord can be generated with a small amount of calculation, and the melody is adapted. Chords can be generatedAlso, since the approach notes that cannot be determined only within the common chord area are determined based on the preceding and following common chord areas, it is possible to determine the approach notes more reliably and generate a chord adapted to the melody. Will be able to.
[Brief description of the drawings]
FIG. 1 is a schematic diagram of a music data generation system according to an embodiment.
FIG. 2 is a functional block diagram of the music data generation system according to the embodiment.
FIG. 3 is a memory map showing information stored in storage means of a portable information terminal
FIG. 4 is a memory map showing information stored in storage means of the music data generation device
FIG. 5 is a flowchart of a music data generation system according to the embodiment.
6 is a detailed flowchart of chord generation processing in FIG.
[Explanation of symbols]
1 ... Music data generation system
2 ... Mobile information terminal
3. Music data generator
20 ... Input acceptance means
24 ... Output means
33 ... Approach note judging means
34 ... Chord generating means
36. Integration means

Claims (4)

In a music data generation device that generates music data that is chorded based on an input melody,
An approach note determining means for determining an approach note in the common chord area from the given melody; a chord generating means for generating a chord in the common chord area based on the constituent sound excluding the determined approach note; An integration means for integrating the generated chord and the melody and an output means for outputting the integrated music data, and the approach note determination means takes a temporary approach note from the given melody. A temporary chord candidate is generated by determination, and when the chord progression including the temporary chord candidate is not suitable for the chord progression stored in the storage unit in advance, the common chord that is not suitable for the chord progression A music data generation device for correcting the interpretation of a temporary approach note in a domain . Said chord generating means, based on the component notes except for the approach note is determined by the approach note determination means generates a plurality of candidate chords in a common chord region, considering matching degree between melody from the The music data generation device according to claim 1 , which extracts chords. In a music data generation method for generating music data that is chorded based on an input melody,
Determining an approach note in the common chord region from the given melody, generating a chord in the common chord region based on the constituent sound excluding the determined approach note, and the generated chord And the step of integrating the melody and the step of outputting the integrated music data ,
The step of determining the approach note includes a step of determining a temporary approach note from a given melody, generating a chord candidate based on the temporary approach note, and including the generated temporary chord candidate The step of determining whether or not the chord progression matches the chord progression stored in the storage means in advance, and if it is determined that the chord progression is not suitable for the chord progression based on the determination result, the chord progression is not suitable A music data generation method comprising a step of correcting an interpretation of a temporary approach note in a common chord region . The step of generating the chord generates a plurality of chords in the common chord region based on the constituent sounds excluding the approach note determined in the approach note determination step, and considers the degree of consistency with the melody from the plurality of chords music data generation method according to claim 3 extracts a chord.

Images