JP2005173492A - Code progress data generating system and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a code progress data generating system for obtaining code progress of user taste without a feeling of physical disorder when generating the code progress corresponding to a given melody. <P>SOLUTION: The system comprises registering user-desired code progress on a code progress database DBb, and customizing a tool when generating the code progress by a template matching system(R1 and R2). When inputting melody data (R3), the code progress suitable for the input melody data is generated by using the customized code progress database DBb (R4). When selecting the user-desired code progress from the code progress (R5 and R6), accompaniment data is generated (R8), and music data is completed by matching with the input melody (R9). In addition, when selecting the code progress, necessary code progress is registered to the code progress database DBb to prepare next code progress generation (R10). <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a chord progression data generation system suitable for generating chord progression data suitable for an input melody, and more particularly to a chord progression data generation system that can use a customized chord progression database.

Conventionally, what extracts chord progression data suitable for melody data from a database is known. For example, in Patent Document 1, a plurality of chord progressions are registered in a database in advance, the input melody data is divided into predetermined sections, a chord sound is detected from the melody data in each divided section, and the detected sum is detected. Chord progression data including voice sounds as chord constituent sounds is extracted from the database.
JP 2001-142462 A

  However, in the method of previously registering the chord progression data used for the template in the database in advance, it is necessary to prepare all combinations of expected chords, so chord progression data corresponding to all the combinations. Is registered in the database, it takes enormous memory capacity and enormous registration work time to construct the chord progression database. Further, since there is a limit to the chord progression (chord combination) prepared in the chord progression database, optimal chord progression data cannot be obtained.

  Further, as in the prior art, in the method of extracting chord progression from the registered data of the database corresponding to the melody data of a predetermined divided section, it is only musically because the chord progression suitable for each divided section is obtained. A chord progression without undulations is generated.

  Furthermore, extracting chord progression data that matches the input melody from chord progression data registered in advance in the chord progression database has the disadvantage that it is difficult to obtain chord progressions that the user likes.

  SUMMARY OF THE INVENTION In view of these disadvantages, the present invention has an object to provide a chord progression data generation system capable of obtaining a chord progression of a user's preference without a sense of incongruity when generating a chord progression suitable for given melody data. And

  According to the main feature of the present invention, chord progression registration means (R1, R10) for registering chord progression data in the chord progression database (DBb) based on user operation, melody input means (R3) for inputting melody data, Chord progression data generating device (Claim 1) comprising chord progression data generating means (R4) for generating chord progression data according to the registered chord progression data corresponding to the input melody data, and chord progression Steps (R1, R10) of registering chord progression data in the chord progression database (DBb) based on a user operation for an information processing apparatus having a chord progression database (DBb) in which data is registered, and melody data In response to the input step (R3) and the input melody data, Chord progression data generation program for executing the steps consisting the steps (R4) for generating a chord progression data [Claim 3] is provided in accordance with the chord progression data registered in over de progression database (DBb). Note that parentheses indicate reference symbols in the embodiments.

  According to a further feature of the present invention, a chord progression database (DBb) in which chord progression data including standard chord progression data (DBo) prepared in advance is registered, and chord progression data (DBu) based on a user operation. Are registered in the chord progression database (DBb), the chord progression registration means (R1, R10), the melody input means (R3) for inputting melody data, and the chord progression database (DBb) corresponding to the inputted melody data. A chord progression data generating device (Claim 2) comprising chord progression data generating means (R4) for generating chord progression data according to the chord progression data registered in (1) is provided.

  In the chord progression data generation system according to the present invention, the chord progression desired by the user is registered in the chord progression database (DBb) based on a user operation (R1), and when melody data is input (R3), the chord progression database (DBb) The chord progression that matches the input melody data is generated in accordance with the chord progression registered in () (R4). Furthermore, the user's desired chord progression is selected from the generated chord progressions (R5, R6), and at that time, the necessary chord progressions are registered in the chord progression database (DBb) (R10) to prepare for the next chord progression generation. Can do. Further, accompaniment data is generated from the chord progression desired by the user (R8), and the song data can be completed together with the input melody (R9).

  When extracting the chord progression using the chord progression of the database as a template, the chord progression is output when the input melody tone string can take the chord progression of the template, and is prepared in advance in the system. The chord progressions generated using the database may match the melody, but often feel uncomfortable. This is because the registered chord progression does not match the user's preference.

  Therefore, in the present invention, as described above, the chord progression generation tool can be customized so that the user can arbitrarily register a desired chord progression in the database and the user's registered chord progression is easily generated. Can do. Therefore, the chord progression of the user's preference can be obtained, and the above-mentioned uncomfortable feeling is eliminated.

  Further, the chord progression data registered in the chord progression database (DBb) may be only chord progression data (DBu) by user registration, but the standard chord progression data (DBo) prepared in advance in the apparatus is included. The standard chord progression data (DBo) and user registration chord progression data (DBu) can be used together to generate desired chord progression data that matches the melody. That is, if only the user registered chord progression data (DBu) is registered in the chord progression database (DBb) according to the user's own preference, chord progression data may not be obtained depending on the input melody. This can be dealt with by generating desired chord progression data from the progression data (DBo). In this case, when candidate data comes out from both chord progression data (DBu, DBo) corresponding to the input melody, the user registration chord progression data (DBu) should be given priority.

  In addition to the chord progression database (DBb) that can be registered by the user, the chord progression generation conditions used for generating chord progression data include preset chord candidate generation and chord progression candidate generation rules (CDa, CDb). included. Specifically, according to the chord candidate generation rule (CDa), chord candidate data for each section of the input melody data is generated (Pa; P41), and the chord candidate data for each section is combined for a plurality of sections. Then, chord progression candidate data (Pb; P42) is created, and chord progression candidate data (Pd; Pd; which matches the chord progression data stored in the chord progression database (DBb) from the created chord progression candidate data. P44) or chord progression candidate data (Pc; P43) matching the chord progression candidate extraction rule (CDb) is extracted, the extracted chord progression candidate data is displayed (P5), and the displayed chord is displayed based on the user operation. A desired chord progression candidate data is selected from the progression candidate data (P6). (CDa) is applied to obtain chord candidates for each section from the input melody, and chord progression candidate generation rules (CDb) are used from chord progression candidates based on combinations of these chord candidates as well as chord progression database (DBb). Thus, chord progression candidates to be presented to the user can be obtained.

  In other words, when it is desired to obtain special chord progression data desired by the user, if such chord progression data is registered in the database, the template matching method using the chord progression database (DBb) (user registration is performed). The chord progression candidate that matches the chord progression is extracted as a template), and special chord progression data suitable for the input melody can be easily obtained. On the other hand, when it is desired to obtain general chord progression, a predetermined musical rule is applied using chord progression candidate generation rules (CDb), and general chord progression that matches the input melody is applied. It can be obtained with an algorithm. The chord progression database (DBb) can be configured as a user registration dedicated database (DBu), or can be configured to perform user registration during chord progression data generation work. Therefore, a large-scale memory for the chord progression database is not required, and user registration work is reduced.

  Furthermore, regardless of whether the chord progression candidate is extracted by the rule application method or the template matching method, the chord progression candidate variation is generated by performing further re-harmonization on the extracted chord progression candidate, and more Various chord progressions can be obtained.

  In addition to setting the chord progression generation condition (CDc) corresponding to the whole melody, it is set so that a special chord (specific code) is designated for the melody paragraph part. When input (Q1), chord progression data is generated according to the set chord progression generation condition (CDc) corresponding to the input melody data (Q4), but the musical paragraph part of the input melody data Is detected (Q2), a specific code is specified in the vicinity of the detected paragraph part (Q3), and the specified specific code is given priority over the chord progression generation condition (CDc) for the vicinity of the paragraph part. By assigning (Q4), a special code is assigned to the melody paragraph detected from the given melody data, and the others The partial grants chord progression based on the chord progression generation conditions, can generate a chord progression with a musically relief.

  A preferred embodiment of the present invention will be described below with reference to the drawings. However, this is merely an example, and the present invention can be variously modified without departing from the spirit of the invention.

[System Overview]
FIG. 1 is a block diagram showing the hardware configuration of a chord progression data generation system according to an embodiment of the present invention. An information processing device such as an electronic musical instrument or a personal computer (PC) having a music information processing function is used as a chord progression data generation device as a main body of the system. Device (CPU) 1, random access memory (RAM) 2, read only memory (ROM) 3, external storage device 4, performance operation detection circuit 5, setting operation detection circuit 6, display circuit 7, tone generator circuit 8, effect circuit 9 , A MIDI interface (I / F) 10, a communication interface (I / F) 11, and the like, and these devices 1 to 11 are connected to each other via a bus 12.

  The CPU 1 executes various music data processing including chord progression data generation using a clock generated by the timer 13 in accordance with a predetermined control program. The RAM 2 is used as a work area for temporarily storing various data necessary for these processes. The ROM 3 stores various control programs including a chord progression data generation program, various data, parameters, and the like in order to execute these processes.

  The external storage device 4 is not only a built-in storage medium such as a hard disk (HD) but also a compact disk read only memory (CD-ROM), a flexible disk (FD), a magneto-optical (MO) disk, a digital multipurpose disk ( Various portable external storage media such as a small memory card such as a DVD, a smart media (registered trademark), and the like are included. For example, a chord progression database (DBb) can be constructed in the ROM 3 or the external storage device 4 to store a large number of chord progression data.

  The performance operation detection circuit 5 detects the performance operation contents of the performance operation elements 14 such as a keyboard and a wheel, and the setting operation detection circuit 6 detects the setting operation contents of the setting operation elements 15 such as a key switch and a mouse. The information corresponding to the detected contents is introduced into the system main body. The display circuit 7 includes a display for displaying various screens and various indicators (lamps), and controls the display / lighting contents in accordance with instructions from the CPU 1 to provide display assistance for the operations of the operators 14 and 15. Note that an acoustic input means for inputting acoustic information from a microphone or the like is provided as necessary.

  The tone generator circuit 8 generates a musical tone signal corresponding to musical tone data obtained by processing performance data from the performance operator 14 and the storage means 3 and 5. An effect circuit 9 having an effect-providing DSP imparts a predetermined effect to the tone signal, and a sound system 17 connected to the effect circuit 9 includes a D / A converter, an amplifier, and a speaker, and the tone to which the effect is imparted. A musical tone based on the signal is generated.

  Further, another electronic musical instrument (MIDI device) ED is connected to the MIDII / F10, and performance data including chord progression is exchanged between this chord progression data generation device and another electronic musical instrument ED. Can be used. Further, the communication I / F 11 is connected to a communication network CN such as the Internet or a local area network (LAN), and downloads various information such as control programs and performance data from an external server computer SV or the like to external storage device. 4 can be used in this system.

[First method (first embodiment)]
In the chord progression data generation system according to the first embodiment of the present invention, the chord candidate data is generated for each section from the input melody data by the “first chord progression generation method”, and the chord candidate data of a plurality of sections is obtained. After generating chord progression candidate data in combination, chord progression candidate data matching the input melody can be generated from this chord progression candidate data. In the first chord progression generation method, a “rule application method” is applied, in which a predetermined rule is applied to generation of chord candidate data and extraction of chord progression candidate data, and chord progression candidate data that meets a predetermined condition is extracted. In addition, the extraction of chord progression candidate data can be used in combination with a “template matching method” that extracts chord progression candidate data that matches a chord progression database template from chord progression candidate data.

  The chord progression candidate data extracted by the rule application method and / or the template matching method is further presented to the user, and desired chord progression data is generated through the user's selection operation. In this way, the technique of creating song data using the first chord progression generation method is called a “first method” in this specification.

  FIG. 2 is a flowchart showing an operation example by the process of the “first method”, and FIG. 3 is a flowchart showing a specific processing operation example in step P4 of FIG. FIG. 4 shows an example of the input melody and the chord candidates obtained in the process of the “first method”. Hereinafter, referring to the specific example of FIG. 4 as appropriate, FIG. 2 and FIG. The process flow will be explained.

  In FIG. 2, when the processing operation of the first method is started by the operation of the predetermined setting operator 15, first, in step P1, the CPU 1 sets the melody data of the predetermined portion including the motif on the RAM 2 and sets the melody. Enter.

There are several ways to enter this melody:
(1) Direct recording = as shown in Patent Document 1, actual performance data input from the performance operation detection circuit 5 based on the performance operation of the performance operator 14 such as a keyboard or information for each sound of the setting operator 15 Based on the input, melody data in a predetermined format (such as MIDI) is generated in accordance with step performance data input from the setting operation detection circuit 6, and the generated melody data is loaded onto the RAM 2. Alternatively, the pitch of nasal song input from the sound input means is detected to generate melody data in a predetermined format and load it into the RAM 2.
(2) Recall of stored data = Melody data generated by the method of (1) or melody data in a predetermined format (MIDI etc.) sent from an external MIDI device ED such as a sequencer or server computer SV The melody data stored in the storage device 4 and stored in the external storage device 4 is loaded onto the RAM 2.
(3) Direct reception = The melody data created by the external MIDI device ED is received and loaded directly onto the RAM 2 or the melody data received from the server computer SV is downloaded onto the RAM 2. And so on.

  In the melody input step P1, the input melody can be displayed as a score on the display 16. Furthermore, the input melody may be configured to be listened to through the tone output units 8, 9, and 17. In this case, since the input melody does not yet have chord information, the input melody may be auditioned without accompaniment, or may be auditioned with a metronome or drums accompaniment.

In the next step P2, the key of the input melody is detected. The key detection is performed, for example, according to the following rules 1 to 3 (key detection rule):
[1] Rule 1 = “The key that composes the scale from all notes is the key candidate”
For example, when a melody with a motif as shown in the upper part of FIG. 4 is input, since the third sound has “F” (fa), it is not a # system key. Since “B” does not appear, if it is natural (no # ♭), it is “C major” or “A minor”, “F major” with B in the habit or “D minor” "Is the key candidate. In the case where two or more wrinkles are attached, at least “E” has wrinkles, but “E” is natural in this motif.
From the above, in the example in the upper part of FIG. 4, the key candidates are “C major” / “A minor” and “F major” / “D minor”.
[2] Rule 2 = “If there are multiple key candidates, select key with less # and wrinkles”
In the case of the example in the upper part of FIG. 4, “C major” / “A minor” is obtained. [3] Rule 3 (determination of major / minor) = “minor if the first note is 1 or 5 degrees minor, otherwise Is a major. ''
In the case of the example in the upper part of FIG. 4, the first sound “G” (So) is not 1.5 degrees of the A minor, so it is determined as “C major”.

In the following step P3, based on the detected key, the diatonic chord, that is, a chord composed only of the detected key scale (for example, seven chords and four chords are included) Furthermore, +7 types) are determined. The diatonic chord determined here is used to narrow down chord (progress) candidates in the next chord progression candidate generation step P4. The diatonic code is determined according to the following rule 4, for example:
[4] Rule 4 = “C major, 3 chords, C, Dm, Em, F, G, Am, Bm ♭ 5 (decrease 5 degrees), 7; In the case of the example in the upper part of FIG. 4, since it is a major, it becomes the former seven.

  Next, in step P4, chord progression candidate data is generated by the first chord progression generation method. That is, a chord candidate is obtained for each input melody section by the rule application method (P41), the obtained chord candidates are arranged in time series from the top, and a chord progression candidate is created by combining the arranged chord candidates for a plurality of sections. (P42) And from the created chord progression candidates, chord progression candidates suitable for the melody are obtained by the rule application method or the template matching method (P43, P44). Further, rehabilitation is performed as necessary (P45). Details of the chord progression candidate data generation in step P4 will be described later with reference to FIG.

  The chord progression candidate data generated in step P4 is displayed on the chord progression candidate screen on the display 16 in step P5. Next, in step P6, when desired chord progression data is designated by the user operation of the operator 15 from the chord progression candidate data displayed on this screen, the CPU 1 inputs the designated chord progression candidate data. Select chord progression data that matches the melody.

  In this case, the chord constituting the selected chord progression data may be editable so that the user can manually change the chord progression data. The selected chord progression data or manually changed chord progression data may be additionally registered in the chord progression database (DBb) by user operation to customize the template (parallel user registration function). ). Furthermore, the chord progression chord data displayed on the screen may be output to the tone output units 8, 9, and 17 so that the user can select a desired chord progression while listening.

  On the other hand, for accompaniment data, a large number of pattern data (relative frequency) of accompaniment sounds called “style data” is stored in the style database DBa of the external storage device 4 in advance. Therefore, the user-desired style data is selected from the style data in the style database DBa at step P7 with respect to the chord progression data selected at step P6, and the selected style data is added at step P8. To generate accompaniment data.

  In step P9, the generated accompaniment data is merged with the input melody data to complete music data composed of melody data and accompaniment data. The completed song data may be stored in a song database of the external storage device 4, or may be delivered to another MIDI device ED or distributed to other communication terminals via the communication network CN. Good.

  The chord progression data obtained by this system is not only used for generating accompaniment data as described above, but also displayed on the display 14 to present the chord progression that matches the input melody to the user by a chord name or the like. It can be used in various ways, such as for the production of sub-melody for another part. Further, the input melody may be the entire music piece or just the motif part.

[Details of the first chord progression generation method]
Now, the chord progression candidate generation step P4 of FIG. 2 functionally includes a rule application chord candidate detection function unit Pa (P41), chord progression candidate creation function unit Pb (P42), chord progression candidate as shown in FIG. The functions are divided into extraction function parts Pc, Pd (P43, P44) and reharmonication function part (P45). Among these, the function parts Pa to Pc sequentially apply predetermined rules CDa, CDb and perform chord progression in an algorithmic manner. A rule application code progress generation block PA for generating candidates is configured.

  Here, the characteristics of the first chord progression generation method will be briefly described with reference to FIG. This system generates chord candidates for each section from the input melody (Pa), combines these code candidates for a plurality of sections (Pb), and chord progression that matches the input melody from the chord progression candidates by these combinations Candidates are extracted (Pc). At this time, the chord candidate detection rule CDa and the chord progression candidate extraction rule CDb based on a predetermined musical rule are applied to the generation of chord candidates and extraction of chord progression candidates (algorithm method PA). In addition to the algorithm method PA, the template matching method (Pd) is also used for extracting chord progression candidates, and the chord progression stored in the chord progression database DSb is used as a template. It is also possible to extract chord progression candidates that match

  First, in step P41 functioning as the rule application code candidate detection unit Pa, code candidates are detected for each section under a code candidate detection condition CDa prepared in advance. Here, the section is a section (chord section) corresponding to one chord, and is set in advance according to the time signature of the melody. For example, in the case of 4 beats like the melody in the upper part of FIG. 4, one section is 1 to 2 beats, such as “1 bar 1 chord” to “2 beat 1 chord” (or “1 bar 2 chord”), etc. Called. When a plurality of types of code sections can be set as described above, priorities may be determined in advance for these sections, or the user may set them. For example, automatic priority may be set so that either one bar 1 chord or two beats 1 chord is set, or priority may be extracted with an operator. However, in the following example, it is assumed that both 1 bar 1 chord and 2 beat 1 chord are set.

The chord candidate detection condition CDa includes rules 5 to 7 (referred to as chord candidate generation rules). First, non-chord sounds are removed by the following rule 5:
[5] Rule 5 (removal of non-chord sound) = “If three consecutive sounds are moved up and down twice by two adjacent sounds, the middle sound is removed as a non-chord sound”
In the case of the example in the upper part of FIG. 4, the third sound “F” (F) and the sixth sound “D” (Le) indicated by “x” correspond to non-code sounds and are removed.

After removing non-chord sounds, code candidates are found according to the following rule 6:
[6] Rule 6 (Detection of Chord Candidate) = “A Chord with the Most Containing Sound after Non-Cord Sound Removal as Chord Constituent Sound”
When multiple types of code sections are set, each code candidate is detected for each type of section. For example, in the case of 4 beats, “1 bar 1 chord” and “2 beats 1 chord” are obtained.

  When rule 6 is applied to the example in the upper part of FIG. 4, in the case of “one measure and one chord”, the chord constituent sounds are made to match two notes, and as shown in the middle part of FIG. C, Am7, A7, etc. are required as chord candidates including sounds “E” (mi) and “G” (so) (note that all chord candidates are actually listed, but here the explanation is simplified) For these reasons, these three are used.) In the second measure, the code candidates F, Am, Dm7, F # dim, etc., in which the two sounds in the measures “A” (ra) and “C” (do) coincide with each other, are similarly obtained.

  In the case of two beats and one chord, as shown in the lower part of FIG. 4, for the first measure, in the first two beats, the chords that coincide with each sound “E” (Mi) and “G” (So) Candidates Em, Am7, and the like are obtained, and in the second half of the second beat, C, G, E ♭, Gm, and the like that coincide with the fourth sound “G” (So) are obtained. For the second measure, chord candidates C, Cm, etc. that match the fifth note “C” (do) by one note are obtained in the first two beats, and in the second half, each note “A” (la), “ Code candidate F, Am7 ♭ 5, which matches two notes with C ″ (do), is obtained.

Next, with respect to the code candidate obtained in rule 6, the code type is further specified in accordance with the next rule 7, the type of code candidate is limited by the diatonic code determined in step P3, and the code in the next step P42 Prepare for narrowing down combinations of types:
[7] Rule 7 (designation of chord type) = “The chord to be used is a triad and a diatonic chord”

  If this rule 7 is further applied to the example in the upper part of FIG. 4, the next code candidate is detected as the corresponding data among the code candidates obtained in step P41. That is, in the case of 1 bar 1 code, code candidate C is obtained at the 1st bar, and code candidates F and Am are obtained at the 2nd bar. In the case of two beats and one chord, the chord candidate Em is obtained in the first two beats of the first bar, and the chord candidates C and G are obtained in the second two beats. The chord candidate C is obtained in the first two beats of the second bar. However, the chord candidate F is obtained in the second half of the second beat.

The chord candidates whose types are restricted by the rule 7 are used in combination in the next step P42 to create chord progression candidates. Therefore, in the example of FIG. 4, the chord candidates used for creating chord progression candidates are as shown in the following Table 1.

In the next step P42 functioning as the chord progression candidate creation unit Pb, the obtained chord candidates are arranged in time series from the top, and chord progression candidate data is created by combining the arranged chord candidates for a plurality of sections. Accordingly, in the example of FIG. 4, chord progression candidate data for two measures created by combining all the obtained chord candidates is nine candidates as shown in (a) below. Indicates the boundary of the bar, and the symbol “/” indicates the boundary between the first half (2 beats) and the second half (2 beats) in one bar]:
(A) = “C−F”, “C−Am”, “C−C / F”, “Em / C−F”, “Em / C−Am”, “Em / C−C / F”, "Em / G-F", "Em / G-Am", "Em / GC-F"

  In addition, the chord candidates for each section obtained before and after the application of the rules 6 to 7 (chord candidate generation rules) in step P41 and the chord progression candidates created in step P42 may be displayed on the display 16. Good.

  Subsequently, the process proceeds to Step P43 or Step P44, and the chord progression candidate data suitable for the melody out of the chord progression candidate data for the plurality of sections created in Step P42 according to the rule application chord progression extraction method or the template matching chord progression extraction method. Ask for.

First, in step P43 functioning as the chord progression candidate extraction unit Pc according to the rule application method, the chord progression candidate extraction condition CDb prepared in advance for the chord progression candidate data for a plurality of sections obtained in step P42. The chord progression candidate data suitable for the input melody is extracted. The rule based on the chord progression candidate extraction condition CDb (called chord progression candidate extraction rule) includes, for example, the following rule 8:
[8] Rule 8 = “The chord at the beginning of the song (and at the end of the song in some cases) is a key tonic”.
In the case of the example of FIG. 4, by applying rule 8, three chord progression candidates “C-F”, “C-Am”, and “CC / F” are obtained, and these become chord progression candidates. .

  The chord progression candidate extraction rule used for the chord progression candidate extraction condition CDb includes, in addition to rule 8, the same chord usage frequency is not increased, the same chord does not continue for a predetermined number (for example, two or three), etc. General music rules for chord progression are included.

  When extracting chord progression candidates using the rule application method, chord progression candidate data combining all chords may be temporarily stored in a memory such as the external storage device 4 and read out for processing in step P43. However, the same processing may be performed directly without storing the chord progression candidate data in the memory.

  On the other hand, in the chord progression candidate extraction unit Pd according to the template matching method, the short chord progression data is registered in the chord progression database DBb as a template. In step P43, chord progression candidate data for a plurality of segments obtained in step P42. Are compared with the chord progression data registered in the database DBb, and matching chord progression candidate data is extracted.

  For example, if the chord progressions “Em / GC / F”, “Em7 ♭ 5-Am”, “C-A7”, “Em / C-C / F” are registered as the chord progression database DBb, FIG. From the nine chord progression candidates (a) obtained in the above example, two chord progression candidates “Em / GC / F” and “Em / CC / F” that coincide with these registered chord progressions are obtained. .

  Thus, by using not only the rule application code progression candidate extraction unit Pc but also the template matching code progression candidate extraction unit Pd, more various chord progressions can be obtained. For example, the effect can be improved by using a configuration in which a normal chord progression is generated by the rule application method and a special chord progression is generated by the template matching method. In this case, both or either of the chord progression candidate extraction units Pc and Pd may be operated according to the input melody, or both or one of the chord progression candidate extraction units Pc and Pd may be activated according to the input melody section. It may be operated.

  Note that the chord progression candidate extraction technique itself by template matching is already known from Patent Document 1 and the like, and the template matching chord progression candidate extraction unit Pd can be added to the system as necessary. For example, if a system including only the rule application chord progression candidate extraction unit Pc is completed and then added later, various chord progressions can be increased. In addition, when applying template matching (Pd), the chord candidate detection unit (stage) excludes some rules (for example, rule 7) in the chord candidate detection condition CDa or extracts chord progression candidates. A partial rule in the condition CDb (for example, “first code is tonic” of rule 8) may be set.

  After obtaining chord progression candidate data suitable for a melody in steps P43 and P44, rehabilitation is performed in step P45 as necessary. This step P45 functions as a rehearsalization execution unit, generates variations of chord progression candidates, and obtains more various chord progressions. For example, the chord progression candidate obtained in steps P43 and P44 is replaced with a substitute chord (C → Em), or a certain chord “V7” is divided into “IIm” and “V7” to become chord progression “IIm + V7”. replace. However, the code to be replaced needs to be a code that exists in the code candidates obtained in step P41.

  The chord progression candidates thus obtained in step P4 (P41 to P45) are displayed in step P5 (FIG. 2) as described above, and the user is made to select a desired one in step P6. In the case of the example of FIG. 4, in the rule application method, three extracted chord progression candidates are displayed on the display 16, and in the template matching method, two extracted chord progression candidates are displayed, and both types are used together Displays a total of five chord progression candidates. When a desired one of these chord progressions is selected in response to a user operation, a melody with chord accompaniment is played in step P9 after step P8.

[Second method (second embodiment)]
In the chord progression data generation system according to the second embodiment of the present invention, a paragraph is detected from input melody data, a special chord is assigned to the paragraph portion, and chord progression generation conditions are set for the other portions. Based on the “second chord progression generation method” in which the normal chord progression data is given, it is possible to generate chord progression candidate data with undulating input melody. A method of creating song data using the second chord progression generation method is called a “second method”.

  FIG. 5 is a flowchart showing an operation example by the “second method” process, and FIG. 6 shows an example of the input melody and the process by the “second method”. Here, the characteristics of the second chord progression generation method will be briefly described with reference to FIG. In this system, a chord progression generation condition CDc corresponding to the entire melody is set in advance, and a special chord [specific code (also referred to as a paragraph special code)] is designated for the musical paragraph portion of the melody. ing. In the chord progression candidate generation (Q4), chord progression data is generally generated in accordance with the chord progression generation condition CDc corresponding to the input melody, but when a musical paragraph part of the given melody is detected (Q2) The specific code is designated for the detected paragraph part (Q3), and the designated specific code is assigned near the paragraph part (paragraph part) in preference to the chord progression generation condition CDc. That is, a musical paragraph is detected from a given melody, a specific chord is assigned to the paragraph portion, and chord progression based on the chord progression generation condition CDc is assigned to the other portions.

  In FIG. 5, when the processing operation of the second method is started by the operation of the predetermined setting operator 15, first, in step Q1, melody data is input as in step P1 of the first method (FIG. 2). In the next step Q2, the paragraph part of the melody is detected from the input melody data. Here, the paragraph means that the melody is divided by one. A melody of a song has paragraphs not only at the beginning and end of the song, but also in the middle of the song, and these paragraphs are very musically important in that they give relief as a song.

Therefore, in this step Q2, for example, the division of the melody is examined in accordance with the following paragraph detection conditions, and the paragraph part of the melody is detected (note that the beginning and end of the song can be easily started and stopped from the data format). End point can be detected):
[9] Paragraph detection condition = “When there is no note between 1 and 3 beats of a certain bar (this includes the case where there is no key-on event but the note of the previous bar is extended to this bar with a tie. ) Or, if there is only one sound, make it a paragraph "
If there is no sound in the 1st to 3rd beats or there is only 1 note in a certain bar, but there is a note in the 4th beat, this note will be used as the pickup for the next phrase and will not interfere with the paragraph ( (Refer to bar 7 in FIG. 6 described later).

  In the next step Q3, it is specified that a paragraph part special code is assigned to the vicinity of the paragraph part detected in step Q2. A portion near the detected paragraph portion is called a “paragraph portion”, and a portion from one paragraph portion to the next paragraph portion is called a “paragraph section”.

  Step Q4 corresponds to steps P2 to P4 of the “first method” processing flow (FIGS. 2 and 3), and generates chord progression candidates based on the chord progression generation condition CDc and the paragraph special chord designation. The processing operation in step Q4 is executed in parallel with the paragraph part detection and the paragraph part special chord designation in steps Q2 and Q3, and the chord generation condition CDa and chord progression generation condition used in the “first method” for all melodies. In accordance with the chord progression generation condition CDc including CDb and chord progression database DBb (FIG. 3), the chord progression candidate generation function similar to steps P2 to P4 (P41 to P45) is executed. However, for the paragraph portion, the paragraph part special chord designation in step Q3 is applied in preference to the chord progression generation condition CDc, and chord progression candidates are generated according to the paragraph part special chord designation.

The conditions for obtaining the chord progression candidate from the melody are generally “3 chords” (rule 7) according to the chord progression generation condition CDb (FIG. 3) or “the chord at the beginning or end of the song is tonic” (rule 8). However, for the paragraph part, the following paragraph code assignment conditions 1 and 2 based on the paragraph part special code designation are given priority, and codes are assigned according to these paragraph code assignment conditions 1 and 2 (the explanation is simplified). In the following, the key is C major)
[9] Paragraph code assignment condition 1 (when ending at a paragraph, that is, at the front of a paragraph)
= "In the case of complete termination, go back from the paragraph and set the two codes to GC,
In the case of false termination, G-Am is the two codes that go back from the paragraph,
In the case of half-end, the end of the paragraph section (immediately before the paragraph part) is G ”
[10] Paragraph code assignment condition 2 (when moving to the next paragraph section, that is, after the paragraph)
= “Use a code that gives a sense of the next break. For example, go up four times from just before a paragraph (at the end of a paragraph) to immediately after (at the beginning of a paragraph) (Am → Dm, etc.)”
A code that can be assigned to a melody with priority given to these paragraph code assignment conditions 1 and 2 is preferentially adopted in the paragraph portion.

  FIG. 6 shows a specific example of an input melody and chord assignment thereto. In this example, paragraph parts indicated by double vertical lines are detected at the third and tenth measures of the input melody. In this case, since the note of the third measure bar is G, the chord is half-stopped as shown below the staff notation representing the input melody (other stops cannot be made). Also, for the paragraph part of the seventh measure, tracing back to the sixth measure, it can be completely terminated with the immediately preceding chord G (the chord G is possible because the notes are “A” (la) and “B” (si)). As shown, the code is GC. Except for the paragraph part, for example, chords as illustrated are extracted according to the normal chord progression generation condition CDc.

  In this way, in step Q4, a specific chord progression candidate can be generated in the paragraph part. Next, a supplementary explanation will be given for the generation of chord progression candidates for a paragraph section.

  In the case of the second method in which the paragraph part detection is performed, the paragraph section is not constant and has various measures. In general, there are 4 bars, but there are 3 bars and 5 bars, and 7 bars when long. When chord progression candidates are generated for a paragraph section by the rule application method (FIG. 3: Pa to Pc), priority should be given to the paragraph code assignment conditions 1 and 2 for the paragraph portion, and processing can be performed relatively easily. it can. In addition, the chord progression generation condition CDb itself may be configured to incorporate the assignment of code candidates to the paragraph portions corresponding to the paragraph code assignment conditions 1 and 2 in accordance with the paragraph part special code designation. For example, when a paragraph section of five bars is detected, the code string GC may be generated preferentially from the end of the five bars.

  On the other hand, when short chord progression data is registered in the database DBb and chord progression candidates that match the melody of the paragraph section are extracted by the template matching method (FIG. 3: Pd), consideration is given to paragraph sections of various lengths. is necessary. In other words, since there are various lengths in the paragraph section, it is difficult to prepare a database for all lengths. For example, if one to seven bars are registered in the database, the number of chord progressions becomes enormous.

  Therefore, chord progression data having various bar lengths are registered in the database DBb and used in combination. For example, if chord progressions of 1, 2, 3, 4 bars are prepared, and the paragraph section length is 3 bars, chords based on paragraph code allocation conditions 1 and 2 from the chord progression database (DB) of 3 bars Progress candidates are extracted (for example, a chord progression candidate having GC from the end of the three bars is preferentially extracted from the three bar chord progression DB). Also, when the paragraph section length is 7 bars, first, the chord progression of 4 bars is assigned (in this case, the processing of the paragraph portion is unnecessary), and the paragraph code is extracted when the remaining bars are extracted from the 3 bar chord progression DB. Allocation conditions 1 and 2 are applied (for example, a chord progression candidate having GC from the end of three measures is preferentially extracted from the three measure chord progression DB).

  FIG. 7 is a diagram for explaining a paragraph configuration example and a chord progression candidate extraction mode corresponding thereto. For example, when paragraph parts are detected at the bars and 10th bar of the input melody 3, the paragraph section lengths are 3 bars, 7 bars,... As shown in FIG. Therefore, 7 bars (7 = 4 + 3) in the second paragraph section are divided into 4 bars and 3 bars as shown in FIG. 7 (2), and chord progression candidates of 4 bars and 3 bars are assigned to them respectively. On the other hand, in the chord progression database DBb, as shown in FIG. 7 (3), a four bar and a three bar chord progression DB are prepared in advance. Therefore, the three bars of the first paragraph section and the divided three bars of the second paragraph section are compared with the chord progression of the three bar chord progression DB, and the four bars divided of the second paragraph section are the four bar chord progressions. Each chord progression is extracted in comparison with the chord progression in the DB.

  In this case, with respect to the second paragraph section, there is no paragraph part code assignment condition in the first four bars section, and the chord progression is extracted by preferentially applying the paragraph part code assignment condition to the latter three measures section. Therefore, the chord progression of “Am-GC” including the specific chord string “GC” is extracted from the three-bar chord progression DB in the three bars of the first paragraph section and the latter three bars of the second paragraph section. . However, in the first place, in the template matching method, when such a specific code cannot be added to the melody, a code that can be added is extracted, and when it can be added, it is preferentially extracted.

  In the above description, only a specific code is assigned to the paragraph part in preference to the chord progression generation condition CDc. However, on the condition that a section that is not a paragraph part is not included in the paragraph part, The chord progression candidate extraction rule of the chord progression candidate extraction condition CDb is set so as not to extract chord progression candidates having a sense of termination or a break like a specific chord (see paragraph code assignment conditions in [9] and [10]). You may employ | adopt the structure which adds or restrict | limits the template of chord progression database DBb.

  In step Q4, as in step P4 (P45) of the “first method”, the chord progression that is reharmoniced as necessary, conforms to the melody, and is given a specific code in the paragraph portion. Candidate data is required.

  Thereafter, steps Q5 to Q9 are sequentially executed. That is, the chord progression candidates generated in step Q4 are displayed on the display 16 (Q5) and the user selects a desired chord progression (Q6) in the same manner as in steps P5 to P9 of the “first method”. Accompaniment data is generated by joining the desired style data (Q7) selected from the prepared style database DBa (Q8), and merged with the inputted melody data to complete the song data (Q9).

  The paragraph part information detected in step Q2 may be used not only for generating chord progression candidates in step Q4 but also for accompaniment generation (Q8) such as inserting a fill-in in the paragraph part.

[Third Method (Third Embodiment)]
In the chord progression data generation system according to the third embodiment of the present invention, when chord progression data suitable for an input melody is generated, a chord progression database customized by the user is used as a template when the template matching method is applied. According to the “third chord progression generation method”, chord progression candidate data desired by the user can be generated. A method of creating song data using the third chord progression generation method is called a “third method”.

  FIG. 8 is a flowchart showing an operation example by the process of the “third method”. Here, the characteristics of the third chord progression generation method will be briefly described with reference to FIG. In this system, a chord progression desired by the user is registered in the chord progression database DBb, and a tool for generating chord progressions by the template matching method can be customized (R1, R2). When the melody data is input (R3), a chord progression that matches the input melody data can be generated using the customized chord progression database DBb (R4). When the user-desired chord progression is selected from the generated chord progressions (R5, R6), accompaniment data is generated from the user-desired chord progression (R8), and the song data is completed together with the input melody (R9). Also, when chord progression is selected (R5, R6), the chord progression required by the user can be registered in the chord progression database DBb to prepare for the next chord progression generation (R10).

  Hereinafter, it will be described in detail with reference to FIG. In the “third system” system, while the system is operating, the user operates the predetermined setting operator 15 for user registration at any time and inputs desired new chord progression data to enter the chord progression database. You can register in DBb sequentially. That is, as shown in steps R1 and R2 in FIG. 8, the CPU 1 accepts this registration operation (R1) and stores the chord progression data desired by the user in the user registration code progression database DBu in the chord progression database DBb ( R2).

  The chord progression database DBb includes a standard chord progression database DBo possessed from the beginning as a system and a user registration chord progression database DBu for accumulating chord progression data input by user registration. It is updated as needed by registering the chord progression data in the registered chord progression database DBu.

  Now, when a predetermined setting operator 15 for starting chord progression generation is operated by the user and the processing operation of the third method starts, first, in step R3, the melody is the same as in step P1 of the first method (FIG. 2). Data is input and the process proceeds to the next step R4.

  Step R4 corresponds to steps P2 to P4 of the “first method” (see FIGS. 2 and 3). That is, in step R4, in the same manner as steps P2 to P4 (P41 to P45), according to the chord progression generation condition including chord generation condition CDa and chord progression generation condition CDb (see FIG. 3) and chord progression database DBb (FIG. 8). In operation, chord progression candidates that match the input melody are extracted by the rule application method (PA; Pa to Pc) to the template matching method (Pd).

  When the template matching method (Pd) is applied to generate the chord progression candidate, the chord progression database DBb including the user registration chord progression database DBu is used as described above. At this time, if there is a chord progression candidate that matches both the standard chord progression database DBo and the user-registered chord progression database DBu for the same melody, the standard chord progression registered in the standard chord progression database DBo The user registration code progression registered in the user registration code progression database DBu is preferentially adopted rather than (standard chord progression).

  In step R4, similarly to step P4 (P45) of the “first method”, rehearsalization is performed as necessary to obtain chord progression candidate data that matches the melody and suits the user's preference. It is done. Thereafter, similarly to Steps P5 to P9 of the “first method”, the processes of Steps R5 to R8 are sequentially executed, and finally, the music data is completed by merging with the melody data in Step R9.

  Here, the chord progression candidate generated in step R4 is displayed on the display 16 (R5). When selecting the chord progression desired by the user from the displayed chord progression candidates (R6), the chord progression candidate is displayed in step R10. The selected chord progression candidate and the selected desired chord progression can be newly registered in the chord progression database DBb.

  Note that the chord progression database DBb may be configured only by the user registration chord progression database DBu. However, if only the user registration code progression (DBu) is registered in the chord progression database DBb according to the user's own preference, there is a possibility that the corresponding chord progression data cannot be obtained depending on the input melody. If the standard code progression database DBo is used in combination as described above, the user registration code progression database DBu can be supplemented to cope with such a situation.

This chord progression database user registration function is particularly preferably used in the following situations:
(1) When you want to add chords to a melody that you played freely without being conscious of chords, make your favorite chord progression appear.
(2) When letting a child use song data, the teacher and parents are set so that a chord progression suitable for the child (for example, simple) will appear.
(3) When adding chords to a melody created by a friend, please make settings so that elaborate chord progressions can be generated.

It is a hardware block diagram of a chord progression data generation system according to an embodiment of the present invention. It is a flowchart showing the music preparation procedure of the "1st system" using the 1st chord progression data generation system by 1st Embodiment of this invention. It is a flowchart showing the processing operation of the principal part of the 1st chord progression data generation system in "1st system". It is a figure showing an example of the input melody and chord candidate in a 1st chord progression data generation system. It is a flowchart showing the music preparation procedure of the "2nd system" using the 2nd chord progression data generation system by 2nd Embodiment of this invention. It is a figure showing an example of the input melody and chord provision to which the 2nd chord progression data generation system is applied. It is a figure showing an example of the paragraph structure of an input melody in a 1st chord progression data generation system, and chord progression candidate extraction. It is a flowchart showing the music preparation procedure of the "3rd system" using the 3rd chord progression data generation system by 3rd Embodiment of this invention.

Explanation of symbols

PA rule application code progression candidate generation block,
Pa rule application code candidate detection unit (stage),
Pb chord progression candidate creation part (stage),
Pc rule application code progress extraction part (stage),
Pd template matching code progression extraction unit (stage),
DBa, DBb style database and chord progression database,
CDa, CDb chord candidate detection condition and chord progression candidate extraction condition,
CDc chord progression candidate generation conditions (including CDa, CDb, DBb functions),
DBo standard chord progression database,
DBu User registration code progress database.

Claims (3)

Chord progression registration means for registering chord progression data in the chord progression database based on a user operation;
A melody input means for inputting melody data;
A chord progression data generating device comprising chord progression data generating means for generating chord progression data in accordance with the registered chord progression data corresponding to the input melody data. A chord progression database in which chord progression data including standard chord progression data prepared in advance is registered;
Chord progression registration means for registering chord progression data in the chord progression database based on a user operation;
A melody input means for inputting melody data;
A chord progression data generating device comprising chord progression data generating means for generating chord progression data in accordance with chord progression data registered in the chord progression database in response to the input melody data. For an information processing apparatus having a chord progression database in which chord progression data is registered,
Registering chord progression data in the chord progression database based on a user operation;
Inputting melody data; and
A chord progression data generation program for executing a procedure consisting of a step of generating chord progression data in accordance with chord progression data registered in the chord progression database in response to input melody data.

Images