JP6760450B2 - Automatic arrangement method - Google Patents

Description

The present invention relates to a device and a program for automatically arranging (arranging) music performance information, and particularly to a technique for performing high-quality automatic arrangement by considering the accent position of the original music.

Patent Document 1 below discloses a technique in which a user selects a desired part from MIDI-format automatic performance data of a plurality of parts of a given musical piece, and generates a score in a desired format for the selected part. ing. Then, as an example, the user selects a melody part, automatically generates accompaniment data suitable for the melody of the selected melody part, and the selected melody part and the accompaniment part based on the automatically generated accompaniment data. It has been shown to generate a score consisting of. As a method of automatically generating accompaniment data suitable for a melody, specifically, a plurality of accompaniment patterns according to the performance level are prepared in advance, and an accompaniment pattern corresponding to the performance level selected by the user is selected. , The accompaniment data is automatically generated based on the selected accompaniment pattern in consideration of the chord progression in the melody.

Japanese Unexamined Patent Publication No. 2005-202204

It can be said that the automatic generation of accompaniment data in Patent Document 1 described above automatically arranges (arranges) accompaniment based on a given melody. However, the automatic generation of accompaniment data in Patent Document 1 is only to change the pitch of the musical tones constituting the existing accompaniment pattern (chord backing, arpeggio, etc.) according to the chord progression of the melody. Therefore, it was not possible to arrange accompaniment in harmony with the rhythmic elements of the original song. Therefore, there is a disadvantage that the accompaniment added by the automatic arrangement is not in harmony with the rhythmic elements of the original song, so that the accent position originally possessed by the original song is canceled. In addition, when an accompaniment performance based on the accompaniment data automatically generated in this way is performed together with the melody performance of the original song, for example, on a keyboard instrument, the performer's right hand (melody performance) and left hand (accompaniment performance) There is an inconvenience that it becomes difficult to play because the accent positions are not in harmony.

The present invention has been made in view of the above points, and an object of the present invention is to provide an automatic arranging method capable of improving the quality of automatic arranging.

The automatic arrangement method according to the present invention is based on the step of extracting the performance information of at least one part including the melody part from the original performance information and the performance information of the at least one part extracted from the original performance information. The step of extracting one or more accent positions in the song represented by the performance information and the time position of one or more accompaniment sounds generated based on the accompaniment pattern data are matched with the extracted one or more accent positions. The processor executes the step of generating the arranged accompaniment data by adjusting.
According to another viewpoint, the automatic arrangement method according to the present invention is generated based on the step of extracting one or more accent positions in the song represented by the original performance information and the accompaniment pattern data based on the original performance information. A step of generating arranged accompaniment data by adjusting the time position of one or more accompaniment sounds to be adjusted so as to match the extracted one or more accent positions, and a performance of a specific part from the original performance information. The processor executes a step of separating and extracting the information and a step of synthesizing the performance information of the extracted specific part with the generated accompaniment data arranged.

According to the present invention, in generating (arranging) accompaniment data suitable for the original performance information, the time position of one or more accompaniment sounds generated based on the existing accompaniment pattern data based on the existing accompaniment pattern data. Is adjusted so as to match one or more accent positions extracted from the original performance information, thereby generating (arranging) the accompaniment data. Therefore, it is possible to generate accompaniment data that is in harmony with the accent position (rhythmic element) of the original performance information in the song, and automatically arranges the existing accompaniment pattern data and the characteristics of the original performance information (original song). Can be done In addition, when an accompaniment performance based on the accompaniment data automatically generated in this way is played together with the melody performance of the original song, for example, with a keyboard instrument, the performer's right hand (melody performance) and left hand (accompaniment performance) Since the accent position (rhythmic element) is in harmony, it is easy to play. Therefore, it is possible to automatically provide a high quality arrangement.

In one embodiment, the steps for generating the accompaniment data are as follows: (1) In one or more accompaniment sounds generated based on the accompaniment pattern data, the accompaniment sound exists at a time position corresponding to one of the extracted accent positions. In this case, one accompaniment sound existing at a time position matching the one accent position is included in the arranged accompaniment data, and (2) in one or more accompaniment sounds generated based on the accompaniment pattern data, the said If there is no accompaniment sound at the time position that matches one of the extracted accent positions, the accompaniment sound that exists at the time position close to the one accent position is shifted to the time position that matches the accent position, and the arranged accompaniment data. It may be configured to be included in. As a result, accompaniment data that is in harmony with the accent position of the original performance information can be generated.

In another embodiment, the step of generating the accompaniment data is as follows: (1) One accompaniment sound located at a time position finer than a predetermined note resolution among one or more accompaniment sounds generated based on the accompaniment pattern data. Is matched with one of the extracted accent positions, one accompaniment sound located at the fine time position is included in the arranged accompaniment data, and (2) is generated based on the accompaniment pattern data. If the accompaniment sound located at a time position finer than the predetermined note resolution among the one or more accompaniment sounds does not match the extracted accent position, the accompaniment data located at the fine time position is arranged. It may be configured not to be included in. According to this, according to the above-mentioned feature (1), it is possible to generate accompaniment data in harmony with the accent position (rhythmic element) in the song that the original performance information has, and the above-mentioned (2). ), Accompaniment sounds with a resolution finer than the predetermined note resolution are omitted from the arranged accompaniment data unless it corresponds to the accent position, providing an arrangement that is easy for beginners to play. can do.

Each step constituting the automatic arrangement method according to the present invention includes a program module configured to realize each function and a processor capable of executing the program module (a general-purpose processor such as a CPU or a dedicated processor such as a DSP). ) Can be configured.

Further, the present invention can be implemented not only as an invention of a method, but also as an invention of an apparatus provided with means configured to realize the function of each step, and the method can be implemented. It can be implemented as an invention of a program consisting of a group of instructions that can be executed by a processor for execution, and can also be implemented as an invention of a non-transient computer-readable storage medium that stores the program. it can.

The hardware block diagram of the apparatus for realizing one Example of the automatic arrangement method which concerns on this invention. The functional block diagram of the process which concerns on one Embodiment of this invention which is executed under the control of a CPU. As a first embodiment, it is a figure which shows an example of the accent position extraction result performed with respect to the 1st and 2nd measures of a certain original performance information, (a) is a figure of the 1st and 2nd measures of the original performance information. A figure showing an example of a rhythm image, (b) is a table showing an accent position extraction result extracted from the original performance information having the rhythm image of (a). It is a figure which shows the example according to 1st Example, (a) is a musical score which shows an example of the acquired accompaniment pattern data (template), (b) is based on the accompaniment pattern data of (a) according to the acquired chord information. A score showing an example of shifting the pitch of the generated accompaniment sound, (c) is a score showing an example of adjusting the accent position of the accompaniment sound shown in (b) according to the result of extracting the accent position shown in FIG. As a first embodiment, a flowchart showing a specific example of a process (arrangement process) for generating accompaniment data executed in FIG. As a second embodiment, it is a figure which shows an example of the accent position extraction result performed about the 1st and 2nd measures of another original performance information, (a) is the 1st and 2nd of the original performance information. The figure which shows an example of the rhythm image of a bar, (b) is the table which shows the accent position extraction result extracted from the original performance information which has the rhythm image of (a). It is a figure which shows the example according to 2nd Example, (a) is a musical score which shows an example of the acquired accompaniment pattern data (template), (b) is based on the accompaniment pattern data of (a) according to the acquired chord information. A score showing an example of shifting the pitch of the generated accompaniment sound, (c) is a score showing an example of adjusting the accent position of the accompaniment sound shown in (b) according to the result of extracting the accent position shown in FIG. As a second embodiment, a flowchart showing a specific example of a process (arrangement process) for generating accompaniment data executed in FIG.

FIG. 1 is a hardware configuration block diagram of a device (hereinafter referred to as an automatic arranging device) for realizing an embodiment of the automatic arranging method according to the present invention. This automatic arranging device does not necessarily have to be configured as a dedicated device for automatic arranging, and is an automatic device or device according to the present invention in any device or device having a computer function such as a personal computer, a mobile terminal device, or an electronic musical instrument. It suffices if the application program for arrangement is installed. The hardware configuration of this automatic arranging device is for inputting a configuration well known in a computer, for example, a CPU (central processing unit) 1, a ROM (read-only memory) 2, a RAM (random access memory) 3, characters and symbols. Input device 4 including keyboard and mouse, display 5, printer 6, hard disk 7 which is a non-volatile large-capacity memory, memory interface (I / F) 9 for portable medium 8 such as USB memory, sound source circuit board 10. Sound system 11 such as a speaker, communication interface (I / F) 12 for connecting to an external communication network, and the like. The application program for automatic arrangement and other application programs and control programs according to the present invention are non-temporarily stored in the ROM 2 and / or the hard disk 7.

FIG. 2 is a functional block diagram of a process according to an embodiment of the present invention, which is executed under the control of the CPU 1. First, the block 20 acquires music performance information (this is referred to as original performance information) to be arranged. The specific configuration for acquiring the original performance information may be any. For example, the data format of the original performance information to be acquired may be data encoded in a predetermined format such as a standard MIDI file (SMF), or may be image information of a musical score drawn on a five-line notation. It may be composed of well-audible audio waveform data, and may be composed of any other data format as long as it can express a musical score. When the original performance information consisting of the score image is acquired, the image is analyzed according to a known score analysis technique, and the pitch, beat position (time position), note value, etc. of the individual notes constituting the original performance information are coded. And, various symbols such as the strength symbol and the accent symbol are encoded together with the time position. Similarly, in the case of consisting of audio waveform data, the audio waveform is analyzed according to a known analysis technique such as pitch and volume, and the pitch, beat position (time position), and note value of each note constituting the original performance information are analyzed. Etc. may be encoded, and the volume may be encoded together with the time position. Further, the musical part composition of the original performance information to be acquired may be arbitrary, for example, it may consist only of a melody part, or a right-hand part (melody part) and a left-hand part (accompaniment or accompaniment or) like a piano score. It may consist of a chord part), or it may consist of a melody part and a plurality of accompaniment parts such as chord backing, arpeggio and rhythm (drum). For example, when the original performance information to be acquired consists only of the melody part, it is possible to make an arrangement to add the performance information of the accompaniment part according to the present invention. Alternatively, when the original performance information to be acquired consists of a melody part and other accompaniment parts, the performance content of the accompaniment part of the original performance information is different according to the present invention (for example, simplified for beginners or for advanced users). Arrangements can be made to provide accompaniment performance information (complexed to). Further, the configuration or route for acquiring the original performance information is arbitrary, and for example, the original performance information may be acquired from the portable medium 8 holding the desired original performance information via the memory I / F9, or may be acquired. , The desired original performance information may be selectively acquired from an external source or server via the communication I / F12. When the CPU 1 executes the process of the block 20, it functions as a means for acquiring the original performance information.

In the block 21, chord information indicating the chord progression of the musical piece in the acquired original performance information is acquired. If the acquired original performance information includes chord information, the chord information may be acquired. If the acquired original performance information does not include chord information, a chord is detected by analyzing the melody progression included in the original performance information using a known chord analysis technique, and the chord is detected based on this detection. All you have to do is get the information. Alternatively, the user may input chord information via the input device 4 or the like, and acquire the chord information based on the user input. The acquired chord information is used to shift the pitch of the accompaniment sound indicated by the accompaniment data when the accompaniment data to be harmony is generated later.

In blocks 22 and 23, the melody part and other parts (if any) are separated from the acquired original performance information, the original performance information of the melody part is acquired (block 22), and the parts other than the melody part are obtained. Acquire the original performance information (if any) (block 23). If the original performance information includes the part information or the identification information equivalent thereto, the original performance information for each part may be acquired by using the part information or the identification information. If the original performance information consists of a musical score image, if the musical score consists of a melody clef (treble clef) and an accompaniment clef (bass clef) like a piano staff, or by part. If it consists of the staff notation of, the original performance information of each part can be obtained based on it. If it does not consist of a staff notation for each part, the notes of each part such as melody, chord, and bass may be presumably extracted by analyzing the score.

In the block 24, one or more accent positions in the song represented by the original performance information are extracted based on the acquired original performance information. In this case, the accent position of the song may be extracted from the combination of all the parts included in the original performance information, or the accent position of the song may be extracted from some parts included in the original performance information. Also, for example, the user may be able to select from which part the accent position is to be extracted. The extraction of the accent position is performed over the entire song (or one chorus), and one or more extracted accent positions are specified (memorized) in relation to the temporal progress of the original performance information. When the CPU 1 executes the process of the block 24, it functions as a means for extracting one or more accent positions in the song represented by the original performance information based on the acquired original performance information.

The method (algorithm) for specifically extracting the accent position is not limited to a specific method (algorithm), but any method (algorithm) as long as the accent position can be extracted to some extent according to some standard. ) May be adopted. Some examples of methods (algorithms) for extracting accent positions are as follows. From these illustrated methods (algorithms), only one may be adopted, or a plurality of methods may be adopted in combination.

(1) When the original performance information includes a chord part, the number of notes that are simultaneously pronounced at each pronunciation timing in the chord part (or including the chord part and the melody part) is obtained, and the number of notes is pronounced at a predetermined threshold value or more. The timing (that is, the time position or the beat position) is extracted as an accent position. This takes into consideration the characteristic that, especially in the case of piano performance, the more emphasized parts are, the more sounds are played at the same time.

(2) When an accent symbol exists in the original performance information, the pronunciation timing (time position) in which the accent symbol exists is extracted as an accent position.

(3) When the original performance information is a MIDI file, the sounding timing (time position) of the note event whose velocity value is equal to or higher than a predetermined threshold value is extracted as an accent position.

(4) Extracted as an accent position in consideration of the position of the sound in one phrase in the original performance information (for example, melody). For example, since the first sound or the last sound in one phrase seems to have a strong accent, the pronunciation timing (time position) of those sounds is extracted as the accent position. Alternatively, since it seems that the sounds having the maximum or minimum pitch in one phrase also have strong accents, the sound generation timing (time position) of those sounds is extracted as the accent position.

(5) In the melody progression in the original performance information, the sounds whose pitch is significantly changed to the treble side or the bass side by a predetermined threshold value or more from the immediately preceding sound are considered to have a strong accent, so that the sound is pronounced. The timing (time position) is extracted as an accent position.

(6) Each sound of the melody (or accompaniment) in the original performance information is weighted in consideration of the beat position in one bar, and the sounding timing (time position) of the sound whose weighting value is equal to or more than a predetermined threshold is set as the accent position. Extract as. For example, the maximum weight is given to the note of the first beat in one bar, the second weight is given to the note of the front beat (on beat) after the second beat, and the note value of the note of the back beat (off beat) is given. Give the appropriate weight (for example, the eighth note has the third weight, the sixteenth note has the fourth weight).

(7) The note value or duration of each sound of the melody (or accompaniment) in the original performance information is weighted, and the sounding timing (time position) of the sound whose weighted value is equal to or higher than a predetermined threshold value is extracted as an accent position. That is, a sound with a long pronunciation time is considered to have a stronger accent than a sound with a short pronunciation time.

In block 25, the existing accompaniment pattern data is acquired. A large number of existing accompaniment pattern data (templates) are stored in an internal database (for example, hard disk 7 or portable medium 8) or an external database (for example, a server on the Internet), and the user tries to make arrangements. The desired accompaniment pattern data (template) is selected from the database in consideration of the time signature, rhythm, etc. of the original performance information. In response to such user selection, existing accompaniment pattern data is acquired in block 25. It is not necessary to select (acquire) the same accompaniment pattern data throughout the music of the original performance information, and it is possible to select (acquire) different accompaniment pattern data in sections of several measures. Further, a combination of a plurality of types of accompaniment pattern data (for example, a chord backing pattern and a drum rhythm pattern) to be played at the same time may be acquired at the same time. When the CPU 1 executes the process of the block 25, it functions as a means for acquiring the existing accompaniment pattern data.

In one embodiment, a bank of known accompaniment style data (automatic accompaniment data) may be used as a source of existing accompaniment pattern data. In the bank of known accompaniment style data (automatic accompaniment data), a plurality of sets of accompaniment style data are stored for each category (Pop & Rock, County & Blues, Standard & Jazz, etc.), and each set of accompaniment style data is an intro section. The accompaniment data set for each section such as the main section, fill-in section, and ending section is included, and the accompaniment data set for one section is rhythm 1, rhythm 2, bass, rhythmic chord 1, rhythmic chord 2, phrase 1, Contains accompaniment pattern data (template) for multiple parts such as phrase 2. The accompaniment pattern data (template) for each part of the lowest layer in the bank of such known accompaniment style data (automatic accompaniment data) is the accompaniment pattern data acquired in the block 25. In the block 25, the accompaniment pattern data of only one part of the accompaniment data set for a certain section may be acquired, or the accompaniment pattern data of all parts or a part of a plurality of parts may be acquired in combination. Good. As is known, in the accompaniment pattern data (template) of a part including pitch elements such as rhythmic chord 1, rhythmic chord 2, phrase 1, and phrase 2, information indicating a reference chord name (for example, C major chord). It also contains additional information that defines pitch conversion rules. The substance of the accompaniment pattern data (template) may be data distributed and encoded according to a MIDI standard or the like, or data along the time axis such as audio waveform data.

In block 26, accompaniment sound data (accompaniment data) is generated based on the accompaniment pattern data (template) acquired in block 25, and at that time, one or more accompaniment sounds generated based on the accompaniment pattern data. The accompaniment data arranged is generated by adjusting the time position (sounding timing) of the above so as to match one or more accent positions extracted in the block 24. For example, accompaniment pattern data (template) consisting of one or more measures is repeated once or multiple times to generate accompaniment data for a desired section, and the time position of one or more accompaniment sounds in the section ( Arranged accompaniment data is generated by changing the sounding timing) according to the extracted one or more accent positions. When the CPU 1 executes the processing of the block 26, the time position of one or more accompaniment sounds generated based on the acquired accompaniment pattern data is adjusted so as to match the extracted one or more accent positions. It functions as a means to generate arranged accompaniment data.

The block 27 is a process performed when the accompaniment pattern data (template) acquired in the block 25 includes an accompaniment sound having a pitch element such as a chord backing or an arpeggio, and the accompaniment data arranged in the block 26 is used. At the time of generation, the pitch of the accompaniment data to be generated is shifted according to the chord information acquired in the block 21. If the accompaniment pattern data (template) acquired in the block 25 consists of a drum rhythm pattern, the processing of the block 27 is omitted.

The block 28 provides the user with arranged performance information including the arranged accompaniment data generated in the block 26. In terms of design, the format in which the arranged performance information is provided may be arbitrarily selected. For example, whether to provide only the arranged accompaniment data generated in the block 26 in the form of electronic data encoded according to a predetermined format such as the MIDI standard, or to display it on the display 5 as a concrete musical score image. Alternatively, it may be printed out on paper by the printer 6 or provided as electronic image data. As another example, the original performance information of at least one part of the melody part of the original performance information separated by the blocks 22 and 23 and the other parts (if any) is appropriately selected (for example, according to the user's request). (Selected), the original performance information of at least one selected part and the arranged accompaniment data generated in the block 26 are combined, and the arranged performance information thus synthesized is encoded in the same manner as described above. It may be provided in the form of data or in the form of physical or electronic musical score image data.

<First Example>
Next, a specific example of the accompaniment data generation process in the block 26 will be described as a first embodiment. In this first embodiment, the accompaniment data generation process in the block 26 is, for example, 1) a time that matches one of the extracted accent positions in one or more accompaniment sounds generated based on the acquired accompaniment pattern data. When there is an accompaniment sound at the position, one accompaniment sound existing at the time position corresponding to the one accent position is included in the arranged accompaniment data, while 2) it is generated based on the acquired accompaniment pattern data. If there is no accompaniment sound at the time position that matches one of the extracted accent positions in one or more accompaniment sounds, the accompaniment sound that exists at the time position close to the one accent position is set to the time position that matches the accent position. It consists of shifting the accompaniment data and including it in the arranged accompaniment data. As a result, accompaniment data that is in harmony with the accent position of the original performance information can be generated. For example, an example of the accompaniment sound generated according to 1) above is the accompaniment sound at the sounding timing A3 in FIG. 4 (c) described later. Further, an example of the accompaniment sound generated according to the above 2) is the accompaniment sound at the sounding timing A4 in FIG. 4 (c) described later.

More specifically, in the first embodiment, the accompaniment data generation process in the block 26 is, for example, one or more accent positions extracted from the one or more accompaniment sounds generated based on the acquired accompaniment pattern data. One or more accompaniment sounds existing at a time position separated from the predetermined length (for example, one beat or more) are included in the arranged accompaniment data. As a result, it is possible to make an arrangement that retains the characteristics of the existing accompaniment pattern data selected by the user. An example of an accompaniment sound that retains the characteristics of such existing accompaniment pattern data is the accompaniment sound at the sounding timings A1, A2, A5, A6, and A7 in FIG. 4C, which will be described later.

Here, a specific example of the processing in the blocks 24 to 27 according to the first embodiment will be described with reference to FIGS. 3 and 4. FIG. 3 is a diagram showing an example of the accent position extraction result performed in the block 24 with respect to the first and second measures of a certain original performance information, and FIG. 3A is a diagram showing the first and second measures of the original performance information. A figure showing an example of a rhythm image of a bar, (b) is a table showing an accent position extraction result extracted from the original performance information having the rhythm image of (a). In this example, the accent position is extracted according to the method described in (1) above, that is, the method of obtaining the number of sounds that are simultaneously pronounced at each pronunciation timing including the chord part and the melody part in the original performance information. Is shown.

In FIG. 3B, the “bar number” indicates the bar number in the original performance information, “1” is the first bar, and “2” is the second bar. The "beat number" indicates the beat number in one bar, "1" is the first beat, "2" is the second beat, "3" is the third beat, and "4" is the fourth beat. It is assumed that the song related to the original performance information has a four-quarter time signature. The "difference clock" indicates the deviation of the position of the note from the beat position in the original performance information by the number of clocks in which the length of the quarter note is 480 clocks. The difference clock "0" indicates that the note (pronunciation timing) is located at the beat position, and the difference clock "240" indicates that the note (pronunciation timing) is located at a position deviated from the beat position by an eighth note length. It shows that there is. The "number of notes" indicates the number of notes that are simultaneously pronounced at each pronunciation timing. "Length" indicates the length of the note that is pronounced at each pronunciation timing. Note that "eighth note + quarter note" indicates that the eighth note at the end of the first bar and the quarter note at the beginning of the second bar are connected by syncopation. “Accent” indicates whether or not the symbol A was extracted as the accent position, indicates that the symbol A was extracted as the accent position, and indicates that the symbol − was not extracted as the accent position. In this example, it is assumed that the threshold value of the number of sounds extracted as the accent position is set to "3". After all, in the example of FIG. 3, the sounding timing of the dotted quarter note length of the third beat of the first bar and the syncopation extending from the back beat of the fourth beat of the first bar to the first beat of the second bar " The pronunciation timing of "8th note + quarter note" length is extracted as an accent position.

FIG. 4A is a musical score showing an example of the accompaniment pattern data (template) acquired in the block 25. It consists of a chord backing pattern in which C major chords are regularly carved with quarter note lengths. As is known, the accompaniment pattern data (template) including the pitch information has a reference key or root defined in advance, and in FIG. 4A, the reference key or root is C. FIG. 4B shows an example of the chord information acquired in the block 21, and the accompaniment sound generated based on the accompaniment pattern data in FIG. 4A according to the acquired chord information by the processing of the block 27. This is a score showing an example of shifting the pitch of. In this example, the front beat (eighth note length) of the first bar of the original performance information is the C major chord, and the back beat of the fourth beat (eighth note length) of the first bar. From to the end of the second bar is the F major chord. In FIG. 4B, the pitch of the chord backing pattern of the second bar is shifted to the F major chord, reflecting the chord of the original performance information.

FIG. 4C is a musical score showing the accompaniment data generation result performed in the block 26, and the accompaniment sound shown in FIG. 4B depends on the accent position extraction result shown in FIG. 3B. An example of adjusting the accent position is shown. Since there is a dotted quarter note length accent (rhythmic accent) on the third beat of the first bar, the C major chord of the third beat of the first bar in FIG. 4 (b) is shown in the figure. As shown in the sounding timing A3 of 4 (c), the note length is changed to a dotted quarter note. In addition, since the back beat (8th note length) of the 4th beat of the 1st bar has a syncopated accent (rhythmic accent) of "8th note + 4th note" length, the 1st bar in FIG. 4 (b). The sounding timing of the C major chord with the quarter note length of the 4th beat is moved to the time position of the back beat (8th note length) as shown in the sounding timing A4 of FIG. 4 (c), and the third beat. Change to syncopation of "8th note + 4th note" length that continues to the 1st beat of the 2nd bar, and switch from the back beat (8th note length) of the 4th beat of the 1st bar to the F major chord. Then, the pitch of the chord sounded at the sounding timing A4 is shifted to the F major chord. In FIG. 4C, the sounding timings A1, A2, A5, A6, and A7 are separated from the accent position by a predetermined length or more (for example, one beat or more), so that the accompaniment is not affected by the adjustment by the accent position. The chord sounding timing in the pattern data (template) is maintained.

The processing as shown in FIGS. 3 and 4 is executed not only in the first and second measures but also in all necessary measures. Thus, in block 26, the arranged accompaniment data is generated over all the necessary measures.

Next, an example of a specific procedure of the accompaniment data generation process (arrangement process) executed in the block 26 according to the first embodiment will be described with reference to FIG. In FIG. 5, in step S1, the arrangement storage area is set in the RAM 3. In step S2, the accompaniment pattern data selected by the user in the block 25 (see, for example, FIG. 4A) is repeated for the length of the section in which the accompaniment pattern data is used, and is added to the accompaniment pattern data. The pitch of the accompaniment based on the accompaniment is converted according to the chord progression of the original performance information (see, for example, FIG. 4B), and stored in the arrangement storage area. Hereinafter, the data stored in the arrangement storage area is referred to as the current arrangement data. In step S3, the current position for the arrangement process is set to the sounding timing of the first note of the current arrangement data in the arrangement storage area.

In step S4, it is determined whether or not the current position is the accent position with reference to the result of extracting the accent position performed in the block 24 (see, for example, FIG. 3). If NO, the process proceeds to step S5, and it is determined whether or not the accent position is within a predetermined range (for example, less than a quarter note length) from the current position. If step S5 is NO, the process proceeds to step S6, and it is determined whether or not the end of the current arrangement data in the arrangement storage area has been processed. If the processing has not been performed to the end, the process proceeds to step S7, and the sounding timing of the next note currently in the arrangement data is set as the current position. After step S7, the process returns to step S4. The processing relating to the sounding timings A1 and A2 in FIG. 4C is performed by a route that proceeds to step S7 via NO in step S4, NO in step S5, and NO in step S6, and returns to step S4.

For example, when the current position of the process is the accent position as in the sounding timing A3 in FIG. 4C, YES is determined in step S4, and the process proceeds to step S8. In step S8, a note (note group) having a sounding timing at the current position is extracted from the current arrangement data in the arrangement storage area. At the sounding timing A3 in FIG. 4C, a group of notes constituting the C major chord is extracted. In the next step S9, the length of the extracted note group is changed to the note length of the original performance information at the accent position. In the example of the sounding timing A3 in FIG. 4C, the length is changed to a dotted quarter note length. In this way, the length of the note corresponding to the accent position in the current arrangement data in the arrangement storage area is adjusted so as to match the rhythmic accent of the original performance information.

Next, in step S10, if the note length changed as described above is longer than the predetermined period, another note (note group) having a pronunciation timing overlapping with the changed note length is detected from the current arrangement data. Then, the detected note (note group) is deleted from the current arrangement data. For example, this predetermined period can be appropriately set by the user, such as an eighth note length and a quarter note length. The longer the predetermined period is set, the stronger the accent feeling of the original performance with a long duration is reflected in the arranged accompaniment data. On the other hand, the shorter the predetermined period is set, the notes from the current arrangement data. Since the probability of deleting is reduced, the beat feeling of the accompaniment pattern data can be easily maintained. For example, assuming that this predetermined period is set to the quarter note length, the note group of the third beat of the first bar of FIG. 4 (b) is shown in the sounding timing A3 of FIG. 4 (c). When the note length is changed to the dotted quarter note length, it is determined by the process of step S10 that the note length has been changed to be longer than the predetermined period, and the note has a sounding timing overlapping with the changed note length. The note group of the 4th beat of the 1st bar of 4 (b) is detected, and these note groups are currently deleted from the arrangement data.

Next, in step S11, it is determined whether the chord is changed in the middle of the note (note group) whose note length is changed as described above. Regarding the note group changed to the dotted quarter note length in the sounding timing A3 of FIG. 4 (c), since the chord is not changed in the middle, it is determined as NO in step S11 and jumps to the step S6. Then, the process returns to step S4 through the step 7. At this time, since the note group of the 4th beat of the 1st bar of FIG. 4B has already been deleted, the next sounding timing in the current arrangement data is the 1st beat of the 2nd bar of FIG. 4B. It's an eye. Therefore, in step S4, it is determined that the current position (the first beat of the second bar) does not correspond to the accent position.

From NO in step S4, the process proceeds to step S5, and it is determined whether or not the accent position is within a predetermined range (for example, less than a quarter note length) from the current position. When the first beat of the second bar of FIG. 4B is the current position, the position before the eighth note length is extracted as the accent position (see FIG. 3). Therefore, in step S5, A determination of YES is made, and the process proceeds to step S12. In step S12, when the corresponding accent position is ahead of the current position in time, it is determined whether or not there is a note event that is in the sounding state at the corresponding accent position in the current arrangement data. When the first beat of the second bar of FIG. 4B is the current position, there is no note event at the accent position corresponding to the eighth note length in the current arrangement data. Step S12 is determined to be YES, and the process proceeds to step S13.

In step S13, a note (note group) having a sounding timing at the current position is extracted from the current arrangement data. In the case of the above example, the constituent tones of the F major chord of the first beat of the second measure of FIG. 4B are extracted. Next, in step S14, the pronunciation timing of the notes (note group) extracted in the previous step is changed to the corresponding accent position. Next, in step S15, the length of the note (note group) changed to the corresponding accent position in the previous step is expanded or adjusted. As an example, as the sounding timing of the notes (note group) extracted in the previous step is moved to the accent position that precedes in time, the note length is expanded by that amount. For example, as shown in the sounding timing A4 of FIG. 4C, "eighth note + quarter note" that continues from the back beat (eighth note length) of the fourth beat of the first bar to the first beat of the second bar. Extend to a long syncopation.

After step S15, the process proceeds to steps S10 and S11. If the chord is changed in the middle of the note (note group) whose note length is changed as described above, the process proceeds from YES in step S11 to step S16. In step S16, the pitch of the note (note group) whose length has been changed is shifted according to the chord after the change.

As described above, according to the first embodiment, it is possible to automatically generate accompaniment data in consideration of the accent position (and rhythmic accent) of the original performance information, and therefore, a high-quality automatic arrangement can be performed. Can be done.

<Second Example>
Next, another specific example of the accompaniment data generation processing in the block 26 will be described as a second embodiment. In this second embodiment, as a general rule, accompaniment sounds that do not correspond to the extracted accent positions are not included in the arranged accompaniment data. In addition, when the accompaniment sound in the accompaniment pattern data is located at a time position finer than the predetermined note resolution, it is not included in the arranged accompaniment data unless it corresponds to the extracted accent position. That is, in the second embodiment, in relation to the predetermined note resolution, the accompaniment data generation process in the block 26 1) among one or more accompaniment sounds generated based on the acquired accompaniment pattern data. When one accompaniment sound located at a time position finer than the predetermined note resolution matches one of the extracted accent positions, the accompaniment data in which one accompaniment sound located at the fine time position is arranged. 2) Of the one or more accompaniment sounds generated based on the acquired accompaniment pattern data, the accompaniment sound located at a time position finer than the predetermined note resolution does not match the extracted accent position. In this case, the accompaniment sound located at the fine time position is not included in the arranged accompaniment data. The predetermined note resolution can be arbitrarily set by the user, such as a quarter note resolution or an eighth note resolution.

Specific examples of the processing in the blocks 24 to 27 according to the second embodiment will be described with reference to FIGS. 6 and 7. FIG. 6 is a diagram showing an example of the accent position extraction result performed in the block 24 with respect to the first and second measures of a certain original performance information, as in the case of FIG. 3, and FIG. 6A is a diagram showing the original performance. A figure showing an example of the rhythm image of the first and second measures of the information, (b) is a table showing the accent position extraction result extracted from the original performance information having the rhythm image of (a). The way of reading the table is the same as that in FIG. However, since the original performance information with respect to FIG. 6 is different from the original performance information with respect to FIG. 3, the accent position extraction results of the two are different.

FIG. 7A is a musical score showing an example of the accompaniment pattern data (template) acquired in the block 25 in the second embodiment. It consists of a combination of a chord backing pattern and a bass pattern in which C major chords are regularly carved with a quarter note length. FIG. 7B shows an example of the chord information acquired in the block 21 in the second embodiment, and the accompaniment pattern data of FIG. 7A according to the acquired chord information by the processing of the block 27. It is a musical score showing an example of shifting the pitch of accompaniment sounds (chords and bass sounds) generated based on the basis. In this example, the first bar of the original performance information is a C major chord, and the second bar is a G major chord. Therefore, in FIG. 7B, the pitch of the chord in the second bar is shifted to the G major chord, and the pitch of the bass note is also shifted in the same manner, reflecting the chord of the original performance information.

FIG. 7 (c) is a musical score showing the result of the accompaniment data generation processing performed in the block 26 according to the second embodiment, and is shown in FIG. 7 (b) according to the result of the accent position extraction shown in FIG. An example of adjusting the accent position of the accompaniment sound shown is shown. As an example, it is assumed that the predetermined note resolution is set to the eighth note resolution. In the accompaniment pattern data (template) shown in FIG. 7 (a), since there is originally no sound with a resolution finer than the eighth note resolution, the time finer than the predetermined note resolution as shown in 1) and 2) above. When the accompaniment sound located at the position matches the accent position, it is included in the arranged accompaniment data, but the process of not including it in the arranged accompaniment data that does not match the accent position is shown in the illustrated example. It has not appeared. On the other hand, as shown in FIG. 6, the numbers of the first beat, the third beat, the fourth beat of the first bar and the first beat, the third beat, and the fourth beat of the second bar in the original performance information are "4". , Or “3”, so these are extracted as accent positions in consideration of the threshold value “3”. On the other hand, since the number of notes in the second beat of the first bar and the second beat of the second bar is "2", these are not extracted as accent positions. Therefore, as shown in FIG. 7 (c), in the generated accompaniment data, the chords of the second beat of the first bar and the second beat of the second bar in the original accompaniment pattern data are deleted (pronunciation). Timing A12 and A16).

Next, an example of a specific procedure of the accompaniment data generation process (arrangement process) executed in the block 26 according to the second embodiment will be described with reference to FIG. In FIG. 8, in step S0, the note resolution for the arrangement process is set according to the user's selection. For example, it is assumed that the resolution is set to an eighth note length. Steps S1 and S2 perform the same processing as the processing of the same reference numerals in FIG. For example, the accompaniment pattern data as shown in FIG. 7A, which is repeated for the length of the section in which the accompaniment pattern data is used, is stored as the current arrangement data in the arrangement storage area, and the current arrangement data corresponds to the chord information. The pitch is shifted as shown in FIG. 7 (b). In the next step S20, the current position for the arrangement process is set to the first beat of the first measure of the current arrangement data in the arrangement storage area.

In step S21, it is determined whether or not there is a musical note event at the current position of the current arrangement data. In the example of FIG. 7B, when the current position is the first beat of the first bar, there is a musical note event, so it is determined as YES, and the process proceeds to step S22. In step S22, it is determined whether or not the current position is the accent position. In the example of FIG. 6, since the first beat of the first bar is the accent position, it is determined as YES, and the process proceeds to step S23.

In step S23, it is determined whether or not there is an accent position in the original performance information within the range after the current position and less than the note length (for example, eighth note length) corresponding to the set resolution. For example, if there is an accent position at a position 16th note length after the current position, YES is determined and the process proceeds to step S24. In step S24, it is determined whether or not there is a musical note event at this accent position in the current arrangement data. If there is a note event, the process proceeds to step S25, excluding the note at the corresponding accent position, and deleting the note within the range of the note length corresponding to the resolution after the current position from the current arrangement data.

Within the range after the current position and less than the note length (for example, eighth note length) corresponding to the set resolution, the accent position does not exist in the original performance information, or even if it exists, at the accent position. If no note event currently exists in the arrangement data, the process proceeds to step S26. In step S26, all notes within the range of less than the note length corresponding to the resolution after the current position are deleted from the current arrangement data.

In this way, in steps S23 to S26, 1) of the one or more accompaniment sounds generated based on the acquired accompaniment pattern data, one accompaniment sound located at a time position finer than a predetermined note resolution is the extracted accent. If it matches one of the positions, one accompaniment sound located at the fine time position is included in the arranged accompaniment data, and 2) one or more accompaniment sounds generated based on the acquired accompaniment pattern data. Of these, if the accompaniment sound located at a time position finer than the predetermined note resolution does not match the extracted accent position, the accompaniment sound located at the fine time position is not included in the arranged accompaniment data. , Is done.

After step S26, the process proceeds to step S6 to determine whether or not processing has been performed up to the end of the current arrangement data in the arrangement storage area. If the process has not been performed to the end, the process proceeds to step S27, and the current position is set to the beat position (for example, the back beat of the first beat) after the note length (for example, an eighth note) corresponding to the set note resolution. To do. After step S27, the process returns to step S21.

In the examples of FIGS. 6 and 7 (b), when the current position is the first beat of the first bar, there is a musical note event and it corresponds to the accent position, so YES in step S22 and NO in step S23. , Step S26, step S6, and step S27. Therefore, the notes (note group) of the first beat of the first bar of the arrangement data are left without being deleted, and as shown in the sounding timing A11 of FIG. 7 (c), one beat of these first bars is left. The eye notes (note groups) are included in the arranged accompaniment data.

Next, in the examples of FIGS. 6 and 7 (b), when the current position is the back beat of the first beat of the first bar, there is no note event, so NO in step S21, NO in step S23, and step S26. , NO in step S6 to step S27. In this case, in step S27, the current position is set to the second beat (front beat) of the first bar.

In the examples of FIGS. 6 and 7 (b), when the current position is the second beat (front beat) of the first bar, there is a musical note event, but since it does not correspond to the accent position, NO is displayed in step S22. The determination is made, and the process proceeds to step S28. In step S28, the note having the pronunciation timing at the current position is deleted from the current arrangement data. As a result, the notes (note group) of the second beat (front beat) of the first bar of the current arrangement data are deleted, and as shown in the sounding timing A12 of FIG. 7C, the second beat of the first bar is deleted. The notes will not be included in the arranged accompaniment data.

After step S28, the process proceeds to step S23, and the same processing as described above is repeated. In the subsequent processing, in the examples of FIGS. 6 and 7B, the third and fourth beats of the first bar, Since there are note events in the sound accompaniment pattern data at the 1st, 3rd, and 4th beats of the 2nd bar and they correspond to the accent positions of the original performance information, those accompaniment sound data are currently in the arrangement data. As shown in the sounding timings A13, A14, A15, A17, and A18 of FIG. 7 (c), those notes (note groups) are included in the arranged accompaniment data. On the other hand, in the case of the second beat of the second measure, since there is a note event but it does not correspond to the accent position, those accompaniment sound data are currently deleted from the arrangement data and are shown in the sounding timing A16 of FIG. 7 (c). As described above, those notes (note group) are not included in the arranged accompaniment data.

As described above, according to the second embodiment, the accompaniment data considering the accent position (and rhythmic accent) of the original performance information can be automatically generated as in the first embodiment. Therefore, high quality automatic arrangement can be performed. Further, since the accompaniment sound having a resolution finer than the predetermined note resolution is omitted from the arranged accompaniment data unless it corresponds to the accent position, it is possible to provide an arrangement that is easy for beginners to play. it can.

In the above embodiment, the strong accent position in the song represented by the original performance information is determined, and the time position of the accompaniment sound is adjusted according to the strong accent position. However, the present invention is not limited to this, and one or more weak accent positions in the song represented by the original performance information may be determined, and the time position of the accompaniment sound may be adjusted according to the weak accent positions. For example, based on the acquired original performance information, it is determined whether or not the current time corresponds to a weak accent position in the song, and when it is determined that the current time corresponds to a weak accent position, it is generated based on the acquired accompaniment pattern data. The arranged accompaniment data may be generated by adjusting the time position of one or more accompaniment sounds so as to match the extracted one or more accent positions. As a result, the accompaniment performance can be arranged so as to show a weak accent in synchronization with one or more weak accent positions in the song represented by the original performance information.

1 CPU (Central Processing Unit)
2 ROM (read-only memory)
3 RAM (random access memory)
4 Input device 5 Display 6 Printer 7 Hard disk 8 Portable medium 9 Memory interface 10 Sound source circuit board 11 Sound system 12 Communication interface

Claims (9)

The step of extracting the performance information of at least one part including the melody part from the original performance information,
Extracting said based on the extracted performance information of the at least one part from the original performance information, one or more accent position in a song represented by raw performance information,
A step of generating arranged accompaniment data by adjusting the time position of one or more accompaniment sounds generated based on the accompaniment pattern data so as to match the extracted one or more accent positions.
An automatic arrangement method consisting of executing by a processor. A step of separating and extracting the performance information of a specific part from the original performance information, and
A step of synthesizing the performance information of the extracted specific part with the generated accompaniment data arranged.
The automatic arrangement method according to claim 1 , wherein the above-mentioned processor further executes the above. Based on the original performance information, a step of extracting one or more accent positions in the song represented by the original performance information, and
A step of generating arranged accompaniment data by adjusting the time position of one or more accompaniment sounds generated based on the accompaniment pattern data so as to match the extracted one or more accent positions.
A step of separating and extracting the performance information of a specific part from the original performance information, and
A step of synthesizing the performance information of the extracted specific part with the generated accompaniment data arranged.
An automatic arrangement method consisting of executing by a processor . The steps for generating the accompaniment data are as follows: (1) In one or more accompaniment sounds generated based on the accompaniment pattern data, if the accompaniment sound is present at a time position corresponding to one of the extracted accent positions, the 1 One accompaniment sound existing at a time position corresponding to one accompaniment position is included in the arranged accompaniment data, and (2) in one or more accompaniment sounds generated based on the accompaniment pattern data, the extracted accent position If there is no accompaniment sound at the time position that matches one, the accompaniment sound that exists at the time position close to the one accent position is shifted to the time position that matches the accent position and included in the arranged accompaniment data. The automatic arranging method according to any one of claims 1 to 3, comprising. In the step of generating the accompaniment data, among the one or more accompaniment sounds generated based on the accompaniment pattern data, one or more accompaniment sounds existing at a time position away from the extracted one or more accent positions are referred to. The automatic arrangement method according to any one of claims 1 to 4, wherein the accompaniment data is included in the arranged accompaniment data. The steps for generating the accompaniment data are as follows: (1) Of the one or more accompaniment sounds generated based on the accompaniment pattern data, one accompaniment sound located at a time position finer than a predetermined note resolution is the accent extracted. If it matches one of the positions, one accompaniment sound located at the fine time position is included in the arranged accompaniment data, and (2) one or more accompaniment sounds generated based on the accompaniment pattern data. Of these, if the accompaniment sound located at a time position finer than the predetermined note resolution does not match the extracted accent position, the accompaniment sound located at the fine time position is not included in the arranged accompaniment data. The automatic arrangement method according to any one of claims 1 to 3, which is configured as described above. A means for extracting the performance information of at least one part including the melody part from the original performance information,
A means for extracting one or more accent positions in a song represented by the original performance information based on the performance information of at least one part extracted from the original performance information.
A means for generating arranged accompaniment data by adjusting the time position of one or more accompaniment sounds generated based on the accompaniment pattern data so as to match the extracted one or more accent positions.
An automatic arranging device equipped with. A means for separating and extracting the performance information of a specific part from the original performance information,
A means for synthesizing the performance information of the extracted specific part with the generated accompaniment data arranged.
7. The automatic arranging device according to claim 7. A means for extracting one or more accent positions in a song represented by the original performance information based on the original performance information, and
A means for generating arranged accompaniment data by adjusting the time position of one or more accompaniment sounds generated based on the accompaniment pattern data so as to match the extracted one or more accent positions.
A means for separating and extracting the performance information of a specific part from the original performance information,
A means for synthesizing the performance information of the extracted specific part with the generated accompaniment data arranged.
An automatic arranging device equipped with.

Images