JP4682375B2 - Simplified score creation device and simplified score creation program - Google Patents

Description

  The present invention relates to a simplified score creating apparatus and a simplified score creating program, and more particularly to, for example, a simplified score creating apparatus and a simplified score creating program for creating a simplified score that reduces a burden of playing an original score of a music piece as a note.

In the field of music recognition, there is a technology that attempts to analyze the structure of music using a computer. For example, Non-Patent Document 1 discloses the implementation of grouping structure analysis and rhythm structure analysis on a computer based on Musical Theory of Tonal Music (GTTM). GTTM is proposed by Lerdahl and Jackendoff and aims to formally describe content obtained by musical intuition common to an audience with experience in a certain musical language (tonic music). Grouping structure analysis hierarchically divides music into groups with a sense of musical unity, and rhythm structure analysis identifies strong and weak beats at each syllable level.
Masatoshi Hamanaka (2 others), Implementation of music structure analysis based on GTTM: Acquisition of grouping structure and rhythm structure, IPSJ Research Report, 2004-MUS-56 (1), pp. 1-8, 2004

  On the other hand, in recent years, the number of middle-aged and older people who have some experience in playing musical instruments as a child has increased, and it has become more frequent to enjoy playing musical instruments as entertainment. However, it is a big goal to play the song according to the notes, so if you try to try a song that is higher than your level, it will be difficult, you will not stop the practice, even abandon the challenge There are many examples of doing this. In this way, because of the high technical threshold of performance, it was difficult to enjoy the real pleasure of performing the musical performance expression of the art itself. It is desirable to make it easy for a person who wants to play a song to start practicing the performance even if the song is technically difficult.

  In addition, the related technique described above is intended to acquire a hierarchical structure of music, and the technical level and difficulty of playing the music are not considered.

  SUMMARY OF THE INVENTION Therefore, a main object of the present invention is to provide a simplified score creating apparatus and a simplified score creating program capable of creating a simplified score that makes it easy to work on musical performance expression by lowering the technical threshold. is there.

  A first invention (invention of claim 1) is an apparatus for creating a simplified score from an original score of a song, an original score storage means for storing score information described in the original score of the song, an abstract score A performance difficulty knowledge base that stores elements and performance difficulty in association with each other, a musical knowledge base that stores a group of rules that determine the musical importance of individual notes on a score, and judgment of note selection Musical note selection knowledge base for storing rules, musical score abstraction means for analyzing musical score information of music stored in the original musical score storage means to generate abstract musical score information of the music, music generated by the musical score abstraction means Performance difficulty evaluation that evaluates the performance difficulty of each abstract score element included in the abstract score information by comparing the abstract score information of the music with the knowledge base of performance difficulty The musical importance of each musical score included in the musical score information is evaluated by collating the musical score information of the music stored in the original musical score storage means with the rule group described in the musical knowledge base. The degree of playing difficulty of each abstract score element obtained by the degree judging means and the performance difficulty judging means and the musical importance of each note obtained by the musical importance judging means are described in the note selection knowledge base. A simplified score creating apparatus comprising: a note selection unit that determines selection of each note by collating with a rule group; and a simplified score generation unit that generates simplified score information by a note determined to be adopted by the note selection unit. is there.

  According to the first aspect of the present invention, the simplified score creating apparatus (10: reference numerals corresponding to corresponding parts in the embodiments described later) is a simplified score from the original score of the music stored in the original score storage means (14). Is for creating. In the performance difficulty level knowledge base (18), abstract score elements and performance difficulty levels are stored in association with each other. Here, the abstract score element is a certain sound string or one sound extracted from the abstract score information. In addition, the abstract music score information specifies attribute information such as the pitch and note value of individual notes in a normal music score, time-series arrangement information of multiple note chains, and simultaneous pronunciation information of multiple notes. This is musical score data obtained by abstracting so as not to depend on the music. In other words, the performance difficulty level knowledge base is associated with each of the abstract score elements such as the abstract pitch information, the abstract note value information, and the like when playing the abstract score element. Evaluation points indicating technical difficulties are stored. The musical knowledge base (20) stores a rule group including a plurality of rules for determining musical importance of individual notes on the score. The musical importance of each sound is set based on some music theory, for example. The rules for this musical importance are defined so that the musical flow is taken into account. The note selection knowledge base (22) stores a group of rules including a plurality of rules for determining selection of notes. The rule for selection is defined based on, for example, the performance difficulty level of the abstract score element and the musical importance level of the notes. In the rules with a low level of simplification, notes that have a low performance difficulty and a high musical importance are mainly adopted, and in the rules with a high level of simplification, in addition to the above, the performance difficulty is high and the music importance is high. Low notes are also used. The score abstraction means (12, S11) analyzes the original score information of the music and generates abstract score information of the music by abstract conversion. The performance difficulty level determination means (12, S13) refers to the performance difficulty level knowledge base and evaluates the performance difficulty level of each abstract score element included in the abstract score information of the music. The musical importance determination means (12, S15) refers to the musical knowledge base and evaluates the musical importance of each note included in the musical score information of the music. The note selection means (12, S17) determines the selection of each note of the music by comparing the performance difficulty of each abstract score element and the musical importance of each note with the note selection knowledge base. The simplified musical score generating means (12, S19) generates simplified musical score information of the music based on the notes determined to be adopted. When rules for each reduction level are stored in the note selection knowledge base, note selection means performs note selection based on the rules corresponding to each reduction level, and a plurality of types (reduction levels) are generated by the simplified score generation means. You may create a simplified score.

  According to the invention of claim 1, since it is determined whether or not each musical note can be adopted based on the difficulty of playing the abstract score element and the musical importance of each musical note, the technical threshold of performance This makes it possible to create a simplified score that keeps the music atmosphere low. Therefore, the burden on the performer is reduced, and it is possible to easily tackle the expression of musical performance expressions.

  The invention of claim 2 is dependent on the invention of claim 1, and the original score storage means stores the original score of a plurality of pieces of music, and the user can simplify the plurality of pieces of music stored in the original score storage means. A music designating unit for designating music for

  According to the second aspect of the invention, the music designating means (12, S5, S7) allows the user to designate music for reduction from the original music score information of a plurality of music stored in the original music score storage means. Therefore, a user such as a simplified score creator or a player can create a simplified score of a desired music piece.

  The invention of claim 3 is dependent on the invention of claim 1 or 2, further comprising a reduction level specifying means for the user to specify the degree of reduction, and the note selection means is for the reduction specified by the reduction level specifying means. The selection of each note is determined according to the degree.

  In the invention of claim 3, the user specifies the degree of reduction (reduction level) by the reduction level specifying means (12, S1, S3). The note selection means determines the selection of each note according to, for example, a rule for selection associated with the reduction level in accordance with the specified degree of reduction. Therefore, a user such as a creator of a simplified score or a player can create a simplified score of a desired reduction level.

  The invention of claim 4 is dependent on the invention of any one of claims 1 to 3, and is a performance data acquisition means for acquiring performance data of a music piece by a performer, a performance error position from the performance data acquired by the performance data acquisition means. Performance error position detection means for detecting the performance error position detected by the performance error position detection means, the original score information of the music stored in the original score storage means, and the abstract score information generated by the score abstraction means A performance error pattern determining means for determining a performance error pattern unique to the performer by collating, and based on abstract score information related to the performance error pattern specific to the performer determined by the performance error pattern determining means And a unique knowledge base generation means for constructing a knowledge base of performance difficulty specific to the performer. Degree determining means, and also refer to the specific performance difficulty knowledge base to player generated by a unique knowledge base generation means evaluates the performance difficulty of each abstract music elements.

  In the invention of claim 4, the performance data acquisition means (12, S31, S33) acquires the performance data of the music by the performer, and the performance error position detection means (12, S35) detects the performance error position from the performance data. . Performance error positions such as replay, insertion, pitch error, and omission are detected by, for example, comparing the pitch on the score with the pitch of the performance data and evaluating the weight of each sound. The performance error pattern determination means (12, S37) compares the detected performance error position with the original musical score information and the abstract musical score information of the music to determine a performance error pattern unique to the performer. The unique knowledge base generating means (12, S39) constructs a performance difficulty level knowledge base (24) unique to the player based on the abstract score information related to the performance error pattern specific to the player. The performance difficulty level of the abstract musical score information (that is, the abstract musical score element) related to the performance error pattern is evaluated according to, for example, the detection frequency as a performance error. The performance difficulty level judging means refers to the performance difficulty level knowledge base unique to the performer together with the performance difficulty level knowledge base and evaluates the performance difficulty level of each abstract score element in the musical composition to be reduced. Therefore, the performance difficulty of each abstract score element is evaluated based on the technical characteristics of the performer, so that simplification according to the technical characteristics of the performer can be performed and the performance is performed according to the score. You can create a simplified score that reduces the burden on the performer.

  The invention of claim 5 is dependent on the invention of any one of claims 1 to 3, and is a performance data acquisition means for acquiring performance data of a music piece by a performer, a performance error position from the performance data acquired by the performance data acquisition means. Performance error position detection means for detecting the performance error position detected by the performance error position detection means, the original score information of the music stored in the original score storage means, and the abstract score information generated by the score abstraction means A performance error pattern determining means for determining a performance error pattern unique to the performer by collating, and based on abstract score information related to the performance error pattern specific to the performer determined by the performance error pattern determining means Further, it comprises update means for updating the performance difficulty level in the performance difficulty level knowledge base to the player's information.

  In the invention of claim 5, the performance data acquisition means (12, S51, S53) acquires the performance data of the music by the performer, and the performance error position detection means (12, S55) detects the performance error position from the performance data. . Performance error positions such as replay, insertion, pitch error, and omission are detected by, for example, comparing the pitch on the score with the pitch of the performance data and evaluating the weight of each sound. The performance error pattern determination means (12, S57) compares the detected performance error position with the original score information and abstract score information of the music to determine a performance error pattern unique to the performer. The updating means (12, S59) updates the performance difficulty level based on the performance difficulty level knowledge base to the player information based on the abstract score information related to the performance error pattern unique to the player. The performance difficulty level of the abstract musical score information (that is, the abstract musical score element) related to the performance error pattern is evaluated according to, for example, the detection frequency as a performance error. Thus, the performance difficulty level knowledge base can be corrected or adjusted to a knowledge base that reflects the characteristics of the performer. Therefore, the evaluation of the performance difficulty of each abstract score element in the music is performed based on the technical characteristics of the performer, so that the simplification according to the technical characteristics of the performer can be performed, and according to the score. You can create a simplified score that reduces the burden of playing according to the performer.

  The invention of claim 6 is dependent on any one of the inventions of claims 1 to 5, wherein performance data acquisition means for acquiring performance data of a music piece by a performer, error position of performance from performance data acquired by the performance data acquisition means Performance error position detecting means for detecting the performance error position, and the performance error position detected by the performance error position detecting means, the original score information of the music stored in the original score storage means and the abstract score information generated by the score abstract means Performance error pattern determination means for determining a performance error pattern unique to the performer, and the original score storage means stores the degree of difficulty of each piece of music in addition to the score information. By checking the performance error pattern determined by the error pattern determination means with the performance difficulty level knowledge base, the player Based on the performance level determination means for determining the performance level, the performance level of the performer determined by the performance level determination means, and the difficulty level of the music stored in the original score storage means, the degree of music reduction is determined. A reduction level determination means for determining is further provided, and the note selection means determines the selection of each note according to the degree of reduction determined by the reduction level determination means.

  In the invention of claim 6, the performance data acquisition means (12, S71, S73) acquires the performance data of the music by the performer, and the performance error position detection means (12, S75) detects the performance error position from the performance data. . Performance error positions such as replay, insertion, pitch error, and omission are detected by, for example, comparing the pitch on the score with the pitch of the performance data and evaluating the weight of each sound. The performance error pattern determination means (12, S77) compares the detected performance error position with the original musical score information and the abstract musical score information of the music to determine a performance error pattern unique to the performer. The original score storage means stores the degree of difficulty of each song in addition to the score information. The performance level determination means (12, S79) compares the performance error pattern with the performance difficulty level knowledge base, for example, calculates the average difficulty level of the performer, and determines the performance level of the performer. The reduction level determination means (12, S81) determines the degree of reduction (reduction level) of the music based on the performance level of the performer and the difficulty level of the music. The note selection means determines the selection of each note according to the determined degree of reduction. Therefore, it is possible to create a simplified score corresponding to the reduction level based on the performance skill of the performer and the difficulty level of the music, that is, a reduced score corresponding to the skill level of the performer's performance. Accordingly, if the performance skill of the performer is improved, mistakes are reduced, and the performance level and the reduction level are highly evaluated, so that a high-reduction score with a reduced number of notes is created.

  A second invention (invention of claim 7) is a program for creating a simplified score from an original score of a song, an original score storage means for storing score information described in the original score of the song, abstract Difficulty knowledge base that stores and associates musical score elements with performance difficulty levels, a musical knowledge base that stores rules for determining musical importance of individual notes on a score, and note selection A score abstracting means for analyzing a score information of a music stored in an original score storage means and generating abstract score information of the music by a computer having a note selection knowledge base for storing a rule group for determining Individual abstract score included in the abstract score information by collating the abstract score information of the music generated by the converting means with the performance difficulty level knowledge base Each of the musical score information included in the musical score information is checked by comparing the musical score information stored in the musical score base with the rule group described in the musical knowledge base. Musical importance determination means for evaluating musical importance of a note, performance difficulty of each abstract score element obtained by performance difficulty determination means, and musical of each note obtained by musical importance determination means By comparing the importance with a rule group described in the note selection knowledge base, note selection means for determining selection of each note, and simplified score information is generated by the notes determined to be adopted by the note selection means. This is a simplified score creation program that functions as a simplified score generation means.

  The invention of claim 7 is a simplified score creating program corresponding to the reduced score creating apparatus of the first invention described above, and has the same effect as the first invention.

  According to the present invention, the abstract score information of the music is generated to evaluate the performance difficulty level of each abstract score element and the musical importance of each note on the score is evaluated. A simplified score is generated based on notes determined to be adopted based on the difficulty of playing and the musical importance of each note. Therefore, to create a simplified score that can reduce the burden of playing according to the notes, lower the technical threshold of the music and make it easier for the performer to express musical performance expressions. Can do.

  The above object, other objects, features and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

  Referring to FIG. 1, a simplified score creating apparatus 10 according to this embodiment includes a computer 12 that creates a simplified score corresponding to a technical level of performance from an original score of music. The computer 12 is, for example, a PC including a CPU, a memory, an input device, a display device, a communication device, and the like (not shown). The memory of the computer 12 includes an HDD, a ROM, and a RAM, and programs and data for creating a simplified score are stored in advance in the HDD and ROM.

  The computer 12 is connected to an original musical score database (DB) 14 in which original musical score data of one or a plurality of musical pieces is stored in advance, and a simplified musical score DB 16 in which created simplified musical score data is stored. The original score DB 14 and the simplified score DB 16 may be provided outside the computer 12 as illustrated, or may be provided in an HDD or the like in the computer 12. The DB 14 and DB 16 may be provided in another computer on a network that can communicate with the computer 12.

  The “original score” here includes not only the score of the original music as it is, but also the score of the original excerpt when it is famous or standardized. For example, the original “parting song”, which generally has a desire to “become able to play”, is an excerpt from Chopin's “Etude No. 3”, and the score of this type of song is also treated as the original score.

  In consideration of distribution on a network such as the Internet, that is, network distribution, the simplified score data to be created is text data in the MusicXML format instead of the PDF format in this embodiment. As a result, the amount of musical score data can be reduced. Also, the original score data is also prepared in the MusicXML format so that the simplified score data in the MusicXML format can be created without difficulty.

  Note that MusicXML is XML (extensible markup language) that describes information on a musical score. In MusicXML, key marks (C major, etc.), time signatures, clef marks (G clef, etc.) can be described first. Then, the pitch, tone value, tie, articulation symbol (slur, staccato, etc.), strength symbol, decoration sound, etc. of each sound can be described. In SMF (Standard MIDI File), it is possible to describe the pitch, tone length, and volume of performance data, but it is difficult to describe even symbols on musical scores such as MusicXML. In addition, MusicXML is a separate standard from Japanese Musical Plan (http://www.musicalplan.com/home.html) and US Recordare (http://www.recordare.com/xml.html). Has been proposed.

  Original music score data in MusicXML format is, for example, score recognition software such as “Score Maker” (http://www.kawai.co.jp/cmusic/products/scomwin/index.htm) from Kawai Gakki Seisakusho. It may be generated by image scanner recognition or input using software. Alternatively, the computer 12 may acquire original musical score data distributed over a network or stored in a storage medium in the original musical score DB 14.

  The HDD or the like of the computer 12 is further provided with a performance difficulty level knowledge base 18, a music knowledge base 20, and a note selection knowledge base 22.

  The performance difficulty level knowledge base 18 stores an abstract score element and a performance difficulty level in association with each other. The abstract score element is a certain sound string or one sound extracted from the abstract score information, and the performance difficulty level is a technical difficulty level of performance of the abstract score element.

  Specifically, first, the non-abstracted musical score information indicates a normal musical score. For example, in the case of a musical score of a kid's “bonfire” melody, the first measure includes four notes whose note values are quarter notes, and the pitches are “G5, A5, G5, E5” in order. . In the abstraction, for example, the pitch information is abstracted by indicating all the pitches with the pitch between adjacent sounds. In the example of the first measure of the song “bonfire”, it is expressed as “0, +2, -2, -3”. Here, “+2” means that the sound is two semitones higher than the previous sound, and “−3” means that the sound is three semitones lower than the previous sound. means. When abstracting the note value information, the note value information is indicated by the length occupied by each note in one measure. In the example of the first measure of the above song “bonfire”, if the length of one measure is 1.0, it is four quarter notes, so “0.25, 0.25, 0.25, 0.25”. It becomes. For example, an eighth note is 0.125, a half note is 0.5, and a dotted quarter note is 0.375.

  Abstract music score information is attribute information such as pitch and note value of individual notes in normal music score, time-series arrangement information of multiple note chains, and simultaneous pronunciation information of multiple notes. As described above, it refers to the whole score data converted (abstracted) so as not to depend on a specific music piece. Also, the abstract musical score element indicates the whole or part of the abstract pitch information and the abstract note value information. The unit of each abstract musical score element is variously selected such as one sound, two consecutive sounds, and three consecutive sounds. In the performance difficulty level knowledge base 18, each abstract score element is given an evaluation point of technical difficulty of performance.

  For example, it is assumed that there is a series of three sounds “Dorella” in a certain score, and the difficulty level “4 (maximum 5)” is given because it is determined that the performance is relatively difficult. In this case, a series of three sounds called “Sorami” should probably be as difficult as that. However, if the performance difficulty is given to the original non-abstracted musical score information called “Dorella”, when “Sorami” appears in another musical score, it is judged as a series of three unknown sounds. Therefore, the difficulty cannot be evaluated. Therefore, since both “Dorera” and “Sorami” become “*, +2, +7” by abstraction, the difficulty level 4 given to the score “Dorera” is also applied to “Sorami”. It becomes possible. In this way, instead of directly assigning a difficulty level to a sound string or sound of the original score, one knowledge can be obtained by assigning a difficulty level to a once abstracted sound string or sound. It can be applied to the case of The performance difficulty level knowledge base 18 is a knowledge base that stores the correspondence between the score information and the difficulty level.

  For example, a relatively high degree of difficulty is set for the following abstract score information. An abstract musical score element whose pitch is 4-7 degrees or 9 degrees or more away from the previous note, such that the note value is 1/2, 1/3, or 2 times the previous note Abstract musical score element, an abstract musical score element that is a note that should be sharpened or flattened by a key signature, and whose key signature is especially the third or later key signature, or a musical note that should be played at the same time other than that note Abstract score elements such as three or more.

  Such a performance difficulty level knowledge base 18 is mounted on the computer 12 with the knowledge obtained by subject test. Through experiments, we extract the characteristics of the notes or strings that are the cause of the difficulty of performing according to the notes. Specifically, in the experiment, a subject performs a plurality of commercially available practice songs (task songs) at first sight. The musical instrument is a piano in this embodiment. A total of about 100 subjects, for example, beginners and intermediate learners of piano learners. The task song is a song at a level slightly higher than the subject's level, and the first performance (both hands) is recorded. Record the location where the performance failed and the location where the performance stopped, using MIDI and video recording (hand and overview). Watch the video after the performance is complete and ask the subject to explain the reason for the failure or stoppage and the reason (retrospective protocol analysis).

  On the other hand, the characteristics of all the notes of all the task songs are extracted. For example, pitch (pitch of Doremi etc.), note value (quarter note etc.), key signature, accidental symbol, pitch range with previous sound, pitch range with sound just below (upper), Extract characteristics such as the sound value ratio and the number of sounds to be generated simultaneously. Here, difficulty in performance is extracted from the viewpoint of human interface. In other words, we focus on the relationship between the pitch range of the "previous note" and the current note, and the width of the keyboard, based on the keyboard's "horizontal" interface feature. For example, when a player plays a musical note string “de, les” on a piano, the following two flows can be assumed. (1) A note is determined to be “do”, and a keyboard corresponding to “do” is searched for and played. The next note is determined to be “L”, and the keyboard corresponding to L is searched for and played. (2) It is determined that the note is “do” and at the same time the pitch width with the next note is 2 degrees (adjacent). At the same time as searching for the keyboard corresponding to the keyboard, the keyboard that is two times above (right next to) the keyboard is recognized as the sound to be played next.

  The operation of hitting the keyboard corresponding to the note string has an inseparable relationship with the side-by-side interface called the keyboard. If you are a beginner or more, the pattern of (2) above associates the width of notes with the width of the keyboard. Come. However, the ability to instantly determine the width of 5 degrees (for example, the width of De and Seo) and move the finger on the keyboard, regardless of adjacent note strings (2 degrees wide), requires piano playing experience. . Therefore, when extracting difficulties in performance, the pitch range (pitch) with the previous note and the ratio of the note value are also included in the note element.

  Then, by using a multiple regression equation (variable subtraction method) or a decision tree, the characteristics of the notes or sound strings that are strongly related to the location that caused the failure or stop of the experimental result are extracted. The characteristic of the note or the string that causes the difficulty extracted in this way is adopted as the element of the note that is difficult to play. Then, the note or the sound string is abstracted and adopted as abstracted score information, and is taken into the performance difficulty level knowledge base 18 in association with the performance difficulty level.

  The musical knowledge base 20 stores a plurality of rules (rule groups) for determining musical importance of individual notes on the score based on music theory. One or more music theories are considered to evaluate and set the musical importance of each note. As an example, GTTM (Generative Theory of Tonal Music) proposed by Lerdahl and Jackendoff may be applied. The rules of the knowledge knowledge base 20 evaluate the importance of each note in consideration of musical flow. As will be described later, in the note selection knowledge base 22, note adoption rules based on this importance are determined. Therefore, a rule is set so that the importance of a note is evaluated relatively high so that the atmosphere of the music is kept as much as possible and the user (that is, the performer) does not lose the sense of time.

  For example, the following guidelines should be taken when creating rules. (1) In order to determine the nature of the chords and maintain the atmosphere (musical appeal) of the music as much as possible so that the performer does not lose the sense of time, the “bass sound (at each chord on the score) Increase the importance of the sound that is set as the lowest sound). (2) It is better to avoid the melody becoming fragmented as much as possible. First of all, melody and accompaniment are clearly separated (for example, from classical music to early romanticism in Western classical music). For voices other than bass sounds, at least the importance of notes at the beginning of a measure should be highly evaluated. In addition, the importance level is set so that it is continuously adopted in the horizontal direction unless the pitch is difficult to play. (3) The importance level is set so that the bass sound, the highest sound (which tends to be a melody), and other inner voice parts are alternately adopted at every stage as much as possible.

  Note that when creating a rule, a note is specified in consideration of ease of coding. It is difficult for a computer to determine to which voice part (soprano, alto, tenor, bass, etc.) a musical note in a musical piece belongs to each part because there is an increase or decrease in the voice part depending on the part. However, since it is relatively easy to instruct the computer to specify the “highest sound”, “lowest sound”, or “second from the highest sound” among the sounds that are generated simultaneously, such an instruction method is adopted. To do.

  By using the rules of the knowledge base 20, for example, the melody of a simplified score can be prevented from becoming fragmented, and “bass sounds” that influence the atmosphere of the music can be captured from an early stage of reduction. The importance of each note is determined.

  The note selection knowledge base 22 stores a plurality of rules (rule groups) for determining the selection of notes. The rules can be described in the form of a set of if-then rules or a probabilistic model such as a Markov model. Specifically, in order to determine whether or not to adopt a note at a certain location in the simplified score based on the performance difficulty of the abstract score element at that location, the musical importance of the note, and the reduction level The conditions are specified. For example, a rule with a low level of simplification mainly uses notes that have a low performance difficulty and a high musical importance. In addition, a rule with a high level of simplification has a high performance difficulty and a musical importance. Notes with a low degree are also adopted. Therefore, by using the rule group of the note selection knowledge base 22, it is possible to create a plurality of simplified musical scores in which the difficulty level of the performance changes stepwise according to the reduction level, and to keep the atmosphere of the music as much as possible. It is possible to create a simplified score that does not cause the player to lose the sense of time.

  A specific operation of the simplified score creating apparatus 10 will be described with reference to the flowchart of FIG. When the simplified score creation process is started, the CPU of the computer 12 first determines whether or not a reduction level has been selected in step S1. In this embodiment, a reduction level designation screen is displayed on the display device, and the degree of reduction (reduction level) of the music is selected by the operation of the input device by the user (creator or performer). If “YES” in the step S1, the selected reduction level is stored and designated in a step S3. On the other hand, if “NO” in the step S1, the process proceeds to a step S5 as it is. When there is no selection by the user, a default predetermined level may be designated as the reduction level. In another embodiment, the reduction level may be automatically specified in stages regardless of the user's input, or may be fixed to a predetermined level.

  In step S5, it is determined whether or not music selection has been performed. In this embodiment, a music selection screen including a list of music stored in the original score DB 14 is displayed on the display device, and the user selects a music by operating the input device. If “NO” in the step S5, the process returns to the step S1. On the other hand, if “YES” in the step S5, the music selected by the user is stored and designated as a reduction target in a step S7. In another embodiment, the music to be reduced may be automatically selected or designated regardless of user input. In step S9, the original musical score information of the designated music is retrieved from the original musical score DB 14 and read out to the memory of the computer 12.

  Subsequently, in step S11, the original score information is analyzed and converted into abstract score information. As described above, the pitch information, note value information, simultaneous pronunciation information, etc. of each note of the original score are abstracted. For example, the pitch information is converted into information indicating the pitch between all adjacent sounds. The note value information is converted into information indicating the ratio of the lengths of the notes for each measure.

  In step S13, the abstract score information is checked against the performance difficulty knowledge base 18 to evaluate the performance difficulty of each abstract score element included in the abstract score. In the performance difficulty level knowledge base 18, as described above, the abstract score element and the performance difficulty level are associated with each other. Therefore, each abstract score element in the music score abstract score information is extracted and each abstract score element is extracted. The performance difficulty level of the abstract score element can be evaluated by searching the score score element in the performance difficulty knowledge base 18. In this way, it is possible to calculate the performance difficulty level of each part of the designated music.

  In step S15, the musical score of each note is evaluated by comparing the original musical score information with the rule group of the musical knowledge base 20. As described above, the musical knowledge base 20 stores a plurality of rules for determining the musical importance of each note based on the music theory, and the importance of each note in consideration of the musical flow. Degree can be rated.

  In step S17, the selection of each note is determined according to the reduction level by comparing the difficulty level of each abstract score element and the musical importance of each note with the rule group of the note selection knowledge base 22. . As described above, the note selection knowledge base 22 stores a plurality of rules determined based on the reduction level, the difficulty of playing, and the musical importance, so that each note of the original score is designated. It is possible to determine whether or not the reduction level can be adopted.

  In step S19, simplified musical score information corresponding to the designated reduction level is generated based on the notes determined to be adopted. The simplified score is composed only of the notes determined as “adopted” among the notes of the original score, and any notes are not added. It should be noted that an appropriate rest is inserted as needed at a location where a note has not been selected.

  In step S21, the generated simplified score is output. In this embodiment, the generated simplified score data is stored in the simplified score DB 16. The simplified score data may be stored in association with, for example, simplified score identification information, a reduction level, original music identification information, and the like.

  In the output process of step S21, the created simplified score may be displayed on the display device of the computer 12, or may be distributed to another computer.

  In this way, the sound corresponding to the reduction level is selected from the original score of the music based on the performance difficulty of the abstract score element of the portion including the sound and the musical importance of the sound, Reduced score data corresponding to the reduction level is generated. Further, by manually inputting a plurality of reduction levels or automatically sequentially specifying a plurality of reduction levels, it is possible to create a reduced score of a plurality of stages (reduction levels) for the music.

  According to this embodiment, abstract score information is generated from the original score of a musical piece, the performance difficulty level of each abstract score element is evaluated, the musical importance of each note on the score is evaluated, and each abstract score is evaluated. A simplified score is generated from notes that are determined to be adopted based on the difficulty of playing the score information and the musical importance of each note. Therefore, it is possible to create a simplified score that reduces the burden of operating the musical instrument according to the notes. In addition, since musical notes are adopted in consideration of the musical flow, it is possible to maintain the atmosphere of the music and prevent the player from losing the sense of time. Such a simplified score can lower the technical threshold of the music and make it easier for the performer to express the musical performance expression.

  Further, when performing reduction according to each reduction level using the note selection knowledge base 22 having selection rules for a plurality of reduction levels, one content, that is, the original score can be edited to a plurality of levels. . Specifically, it is possible to diversify the level of content by arranging the musical score in a plurality of stages using only the notes of the original musical score. Since a plurality of types (stages) of simplified music scores are created for one piece of music, the performer can start practicing from a simplified music score at a stage suitable for his / her level. In addition, each simplified score can gradually incorporate the burden of playing according to the musical note at each stage according to the reduction level, so that the technical threshold of the music can be lowered step by step and the burden on the performer can be reduced. .

  In the past, famous music has been arranged and marketed so that even beginners can easily play it. However, this kind of arrangement score has the following problems: it doesn't always fit the learner's level, it's too easy, it's too motivated, it's too difficult, and it gets frustrated. There is also. The notes displayed on the original score are rarely used, and even if it becomes possible to perform with the arranged score, it is necessary to practice again from the beginning when performing with the original score. It is dissatisfied with the learner when the musical quality is significantly lower than the original music or when it is similar but not similar.

  On the other hand, the musical quality of the simplified score created in this embodiment may be lower than the arrangement by a professional composer because the emphasis was placed on reducing the difficulty of playing. However, in this embodiment, a simplified score of several stages is created using only the notes displayed on the original score, and the original score is placed on the top after the simplified score stage. Therefore, the learner can practice with a simplified score with the expectation that he can challenge a real score that is not inferior to anything that has improved.

  FIG. 3 shows an example of an original score, and FIGS. 4 and 5 show an example of a simplified score created by the simplified score creation device 10. The original musical score in FIG. 3 is Debussy's “Moonlight”. Comparing FIG. 3 with FIG. 4 or FIG. 5, it can be seen that the simplified musical score maintains the atmosphere and the time signature of the original music. Further, it can be seen that the simplified score of level 1 in FIG. 4 can be challenged by beginners, and the simplified score of level 5 in FIG. 5 is suitable for beginners to intermediates.

  In this way, even if the score of the simplified score is a little higher, the simplified score is suitable for the skill level of the user (player), so the threshold for the user to start practicing is lowered. be able to. Even if it is difficult to perform according to the notes, you can easily challenge the song you want to play, and achieving the performance with the score of each stage becomes a subgoal, and it is easy to continue the practice until you can practice with the original score It is considered to be. In addition, since the burden of operating musical instruments (keyboards, etc.) according to the notes is reduced, it is possible to make it easier for the performer to express musical performance expressions.

  Note that the simplified score data and the original score data created by the simplified score creating apparatus 10 may be taken into a performer's computer or electronic musical instrument via a storage medium. Alternatively, the simplified score data and the original score data may be distributed to a performer's computer or electronic musical instrument from a server provided on the network. In that case, the simplified score creating apparatus 10 may function as the server. On the performer's computer or electronic musical instrument, the simplified score and the original score are displayed on the display based on the obtained simplified score data and the original score data. Since the musical score data is described in the MusicXML format, communication problems due to the amount of data hardly occur.

  Alternatively, the player's computer may function as the simplified score creating apparatus 10 by installing the simplified score creating program and data (including each knowledge base) of this embodiment on the performer's computer. The original score data may be acquired from a storage medium or a server on the network, as in the above-described embodiment. In this case, it is possible to select a desired music piece and a reduction level by the player's own input.

  In the above-described embodiment, the performance difficulty level knowledge base 18 constructed based on the general or standard performance difficulty levels is used. However, in another embodiment, the simplified score creating apparatus 10 may perform reduction based on the performance difficulty inherent to the performer so as to correspond to the technical characteristics of each performer.

  In the reduced score creating apparatus 10 of the embodiment shown in FIG. 6, a specific performance difficulty level knowledge base 24 based on the performance difficulty level specific to the performer is constructed, and this specific performance difficulty level knowledge base 24 is used as the performance difficulty level knowledge base 18. A simplified score is created in conjunction with

  In order to create the player's unique performance difficulty level knowledge base 24, the computer 12 of the simplified score creating apparatus 10 is connected to the player's electronic musical instrument 26 and acquires the player's performance data. The performance data is, for example, MIDI data. Alternatively, the computer 12 may obtain performance data from another computer connected to the performer's electronic musical instrument 26 via a network or the like. The computer 12 creates the inherent performance difficulty level knowledge base 24 for the performer based on the acquired performance data.

  The operation of generating the inherent performance difficulty level knowledge base 24 will be described with reference to the flowchart of FIG. When the unique base generation process is started, the CPU of the computer 12 first determines whether or not the performance data has been acquired in step S31. If “NO”, the generation process ends.

  On the other hand, if “YES” in the step S31, the inherent performance difficulty level knowledge base 24 is constructed. Specifically, the performance data is read in step S33, and the performance error position is detected in step S35.

  As a performance error position detection method, for example, a performance position determination method disclosed in Japanese Patent Application Laid-Open No. 2005-250053 by the present applicant can be used. This performance position determination method extends the DP matching method proposed by Dannenberg, and can handle not only insertion, pitch error, and omission but also replaying. In this method, the performance position is determined by matching only the pitch, ignoring the note value of each sound on the score and the length of the performance sound.

  Specifically, N is the number of sounds included in the score. However, when a plurality of sounds such as chords are pronounced at the same time, the number of sounds at that point is set to “1” regardless of the number of sounds that are sounded simultaneously, and only the highest sound is targeted for matching. When determining the j-th sound Pj from the start of the performance (this time is referred to as “performance time tj”), the performance position on the score at the performance time tj-1 in the previous matching is Si (1 ≦ i ≦ N). ). At this time, it is determined by the following algorithm which sound on the musical score Pj corresponds to.

(1) The pitch Pitch (Sk) of all the sounds Sk (1 ≦ k ≦ N) on the score is compared with the pitch Pitch (Pj) of Pj, and the weight W () at the performance time tj for all sounds Sk, tj) is evaluated by the following method.
(A) When Pitch (Sk) is equal to Pitch (Pj), the weight W (Sk, tj) of the note Sk at the performance time tj is used as the weight W (of the note Sk-1 at the previous time tj-1. A value obtained by adding a predetermined value (for example, “1”) to Sk−1, tj−1) is set.
(B) Otherwise, as the weight W (Sk, tj) of the note Sk at the time tj, the weight W (Sk-1, tj-1) of the note Sk-1 at the previous time tj-1 A value obtained by subtracting a predetermined value (for example, “1”) is set. However, when W (Sk-1, tj-1) is smaller than "0", "0" is set as the weight W (Sk, tj) of the note Sk at the time tj.

  If the weight W is evaluated for all positions, whether the weight at the performance position Si + 1 at the time tj is equal to the weight at the performance position Si at the time tj-1 plus "1", that is, W ( It is determined whether or not Si + 1, tj) = W (Si, tj-1) +1. That is, it is determined whether or not the performance position is correctly updated.

  Here, this condition determination will be described in detail. This is a state in which determination of the performance position at the performance time tj is being attempted. The performance position at the performance time tj-1 is Si. Therefore, W (Si, tj-1) is the weight of the portion determined as “this is the performance position” in the determination at the previous determination (tj-1). W (Si + 1, tj) is the weight obtained as a result of the weighting process at this time (at time tj) of the sound Si + 1 next to Si on the score. When the value of W (Si + 1, tj) is a value obtained by adding “1” to the value of W (Si, tj−1), Si + 1 is determined as the performance position at the time point tj. To do.

  If W (Si + 1, tj) = W (Si, tj-1) +1, the performance position is correct, and the next performance position Si + 1 is set as the performance location at time tj.

  (2) On the other hand, if W (Si + 1, tj) ≠ W (Si, tj-1) +1, a performance error (including replay) is dealt with. That is, when the performance is wrong, a process for identifying the correct performance position is executed.

  First, the correspondence to replaying will be explained in detail. As a “place where replaying is likely to occur” such as measures, phrases, columns and page heads in the range of “1” to “i”. The weights of all the locations specified in advance on the musical score data are W (Si + 1, tj) -m (m is a positive constant). That is, the weights of the performance positions for which the replay position marker is set in the score data are all set to a value “m” smaller than the maximum. However, if the original weight at that location is larger than the weight obtained by this calculation, the value is not changed. In the experiments by the inventors, the value of m is empirically set to “2”. Decreasing the value of m improves the ability to follow replays. On the other hand, in the case of music in which similar patterns appear repeatedly, there is a high possibility of inaccurately recognizing performance points with minor errors (such as missing sounds). Become. In other words, when there are m sounds from a part that seems to be a “replayed part”, it is determined as “replayed” for the first time when m pieces of sound from a part that is considered to be “replayed part” are played continuously. This prevents excessive reaction to minor errors such as missing or insertion of about one sound.

  Next, how to handle insertion, pitch error, and dropout will be described. For all sounds in the range from the position Si-r to Si + r, the weight is made the same as W (Si + 1, tj). In other words, up to r insertions, pitch errors, and omissions are dealt with by highly evaluating the possibility that r sounds before and after the current performance position at time tj-1 are used as performance positions at time tj. . Note that the number r of sounds is a positive constant, and the value of r is empirically set to “2” in the experiment. Increasing this value can deal with errors over a larger number of sounds, but increasing it too much widens the range of possibilities, and conversely increases the risk of finding incorrect matching points.

(3) The position where the weight W (Sk, tj) (where 1 ≦ k ≦ N) takes the maximum value is the current performance position. If multiple locations have the same weight, priority is given to the following order.
(A) If there is only one place where the weight W (Sk, tj) takes the maximum value, the position Sk at which the weight W (Sk, tj) at that time becomes the maximum value is set as the current performance position.
(B) If there are a plurality of places where the weight W (Sk, tj) takes the maximum value, the current performance position is determined as follows.

  (I) If the location Sk having the maximum value closest to Si is unique, that location Sk is set as the current performance position.

  (Ii) When there are a plurality of locations that take the maximum value closest to Si, that is, when a maximum value occurs at positions Si-d and Si + d equidistant from Si, Si-d is set as the current performance position. To do.

  The above algorithm realizes a robust performance tracking process capable of following the “replay” that frequently occurs in beginner performances, in addition to the three types of performance errors that have been handled conventionally. Then, performance error positions such as insertion, pitch error, dropout, and replay in the acquired performance data are detected.

  In step S37, the performance error pattern is determined by comparing the performance error position with the original score data and the abstract score data. For example, when a performer often makes a mistake in the arrangement of sounds such as “Domire”, “Sosila”, “Faraso”, etc., these abstract pitches “*, +4, -2” are It is found that the error pattern is peculiar to the performer. In this way, an abstract musical score element in which a large or small number of errors occur is detected as a performance error pattern from the original musical score information and the abstract musical score information corresponding to the detected performance error position.

  In step S39, a performance difficulty level knowledge base 24 unique to the player is generated based on the abstract score information related to the performance error pattern. The unique performance difficulty level knowledge base 24 stores a performance difficulty level specific to the performer in association with each abstract score element. In the evaluation of the performance difficulty level of the performance error pattern, for example, the technical difficulty level of performance is set to a higher value for an abstract score element having a higher detection frequency as a performance error. In this way, the unique performance difficulty level knowledge base 24 is constructed in which the score pattern and the performance difficulty level that are easily mistaken by the performer are stored in association with each other. This unique performance difficulty level knowledge base 24 is a performance difficulty in which technical features of the performer are aggregated, such as an abstract score element that the performer is not good at or weak at, and an abstract score element that the performer has not mastered. Is a knowledge base. When step S39 ends, the unique base generation process ends.

  When the simplified score creating apparatus 10 of this embodiment creates a simplified score, a general performance difficulty level knowledge base 18 and a specific performance difficulty level knowledge base 24 are used in combination, and individual abstract score elements are used. Evaluate the difficulty of playing. That is, when evaluating the performance difficulty level in step S13 of FIG. 2, the computer 12 also refers to the inherent performance difficulty level knowledge base 24 together. As a result, the performance difficulty level of each abstract score element in the music is evaluated based on the technical characteristics of the performer, and the reduction is performed in consideration of the technical characteristics of the performer.

  In the music specifying process in steps S5 and S7, a music corresponding to the acquired performance data is specified by an input of a user such as a creator or a player. When music performance identification information is given to the performance data, designation is made based on the music identification information.

  According to the embodiment of FIG. 6, it is possible to create a simplified score that is specific to the performer in accordance with the technical characteristics of the performer. Since the burden of operating the musical instrument according to the notes can be incorporated into the simplified score in accordance with the technical characteristics of the performer, the burden on the performer can be further reduced. Therefore, this simplified score makes it easier for the performer to start and continue the practice, and to make it easier to express musical performance expressions.

  In the above-described embodiment of FIG. 6, the inherent performance difficulty level knowledge base 24 is separately constructed so as to correspond to the technical characteristics of the performer. However, in another embodiment, instead of constructing the specific performance difficulty level knowledge base 24, the performance data of the performer is obtained and the general performance difficulty level knowledge base 18 is adjusted to the technical characteristics of the performer. It may be modified and individualized.

  In this embodiment, the performance difficulty level knowledge base 18 is corrected by individualization processing as shown in FIG. When the personalization process is started, the CPU of the computer 12 determines whether or not performance data has been acquired in step S51. If “NO” in the step S51, the individualization process is ended as it is.

  On the other hand, if “YES” in the step S51, the performance difficulty level knowledge base 18 is corrected. That is, the performance data of the performer is read in step S53, and the performance error position is detected in step S55. In step S57, the performance error pattern is determined by comparing the performance error position with the original score data and the abstract score data. The specific processing contents of steps S55 and S57 are the same as those of steps S35 and S37 in FIG.

  In step S59, the performance difficulty level knowledge base 18 is updated based on the abstract score information related to the performance error pattern. As a result, in the performance difficulty level knowledge base 18, the performance difficulty level associated with each abstract score element is overwritten with information unique to the performer. The technical difficulty level of performance of each abstract score element is updated by adding a higher value to the original value as the number of performance errors detected is increased, for example. When step S59 ends, the personalization process ends. According to this embodiment as well, similar to the embodiment of FIG. 6, it is possible to create a simplified score specialized for the player in accordance with the technical characteristics of the player.

  Further, in each of the above-described embodiments, a simplified score is created based on a reduction level selected according to a user's input or a reduction level that is set stepwise or automatically to a predetermined value. . However, in another embodiment, a simplified score may be created based on the reduction level automatically determined based on the performance data of the performer.

  In this embodiment, the reduction level for the performer is determined by the reduction level determination process as shown in FIG. When the reduction level determination process is started, the CPU of the computer 12 first determines whether or not the performance data has been acquired in step S71. If “NO”, the CPU 12 ends the reduction level determination process.

  On the other hand, if “YES” in the step S71, the performance data is read in a step S73, and a performance error position is detected in a step S75. In step S77, the performance error pattern is determined by comparing the performance error position with the original score data and the abstract score data. The specific processing contents of steps S75 and S77 are the same as those of steps S35 and S37 in FIG.

  Subsequently, in step S79, the performance level of the performer is determined by comparing the performance error pattern with the performance difficulty level knowledge base 18. As an example of the performance level determination method, a determination method based on the statistics of the appearance frequency of the performer's performance error pattern can be considered. For example, let Q = {qi; 1 ≦ i ≦ n} be a set of performance error patterns (abstracted musical score elements) detected based on the performance data of the performer. Further, the subset of Q is Q ′ (Q′ＱQ), and Q ′ = {qi; 1 ≦ i ≦ k, k ≦ n}. A performance error pattern included in the subset Q ′, that is, an abstract score element is registered in the performance difficulty level knowledge base 18. Further, the appearance frequency of each performance error pattern is f (qi), and the performance difficulty level value of each performance error pattern registered in the general performance difficulty level knowledge base 18 is w (qi).

The average difficulty Wave. Is calculated by the following equation (1).
[Equation 1]
Wave. = {Σf (qi) · w (qi)} / {Σf (qi)}
Note that the sum (Σ) is a sum from 1 to k. That is, a performance error pattern not included in the subset Q ′ cannot be taken into consideration because a general difficulty level cannot be determined.

  The higher the value of Wave., The more likely the performer is to make mistakes with more difficult patterns. That is, the player can determine that “the level is high” because “the simple pattern does not make a mistake”. The performance level of the performer is determined by the value of Wave.

  In step S81, the reduction level is determined based on the performance level of the performer and the difficulty level of the music. The degree of difficulty of each piece of music is stored in the original score DB 14 in association with, for example, the identification information of the song and the original score information. For example, table data in which combinations of performance level values and music difficulty level values and reduction levels are associated with each other is stored in advance in the memory of the computer 12, and the reduction level is evaluated with reference to the table data. When step S81 is ended, the reduction level determination process is ended.

  When the simplified score creating apparatus 10 of this embodiment creates a simplified score, it generates simplified score data corresponding to the reduction level determined based on the performance data of the performer. That is, in step S17 of FIG. 2, note selection is performed according to the rule corresponding to the determined reduction level. In this way, it is possible to determine the reduction level based on the performance level of the performer and the difficulty level of the music, and to create a simplified score corresponding to the reduction level. That is, it is possible to create a simplified score that matches the skill level of the performer's performance. Therefore, it is possible to allow the performer to practice efficiently and to make it easier to tackle the expression of musical performance expressions by using a simplified score that is just appropriate for the skill level of the current performer. Also, if the player's performance skills improve, mistakes will be reduced, and the performance level and reduction level will be highly evaluated, so a simplified score with a high level (ie adopted from the original score) will have fewer notes to be reduced. (Scores that are close to the original score with few unplayed notes) are created.

It is an illustration figure which shows the structure of the reduced score production apparatus of one Example of this invention. It is a flowchart which shows an example of operation | movement of FIG. 1 Example. It is an illustration figure which shows an example of an original score. It is an illustration figure which shows an example of the level 1 reduced score. It is an illustration figure which shows an example of the reduction score of level 5. FIG. It is an illustration figure which shows the structure of the reduced score production apparatus of another Example. It is a flowchart which shows an example of the operation | movement of the specific base production | generation process in another Example. It is a flowchart which shows an example of operation | movement of the individualization process in another Example. It is a flowchart which shows an example of operation | movement of the reduction level determination process in another Example.

Explanation of symbols

10 ... Simplified sheet music creation device 12 ... Computer 14 ... Original sheet music DB
16 ... Simplified sheet music DB
18 ... Performance difficulty level knowledge base 20 ... Musical knowledge base 22 ... Note selection knowledge base 24 ... Specific performance difficulty level knowledge base

Claims (7)

A device for creating a simplified score from the original score of a song,
Original score storage means for storing score information described in the original score of a song;
A performance difficulty knowledge base that stores and stores abstract score elements and performance difficulties,
A musical knowledge base that stores rules that determine the musical importance of individual notes on a score.
A note selection knowledge base that stores a group of rules for determining the selection of notes;
A score abstraction means for analyzing the score information of the music stored in the original score storage means and generating abstract score information of the music;
The performance score of each abstract score element included in the abstract score information is evaluated by comparing the score score information of the music generated by the score abstract means with the performance difficulty knowledge base. Performance difficulty determination means,
Music that evaluates the musical importance of each note included in the musical score information by comparing the musical score information of the music stored in the original musical score storage means with the rule group described in the musical knowledge base Importance evaluation means,
The performance difficulty of each abstract score element obtained by the performance difficulty determination means and the musical importance of the notes obtained by the musical importance determination means are described in the note selection knowledge base. Simplified score creation comprising note selection means for determining selection of each note by collating with the rule group, and simplified score generation means for generating simplified score information by the notes determined to be adopted by the note selection means apparatus. The original score storage means stores original scores of a plurality of songs,
The reduced score creating apparatus according to claim 1, further comprising a song designating unit that allows a user to designate a song for reduction from a plurality of songs stored in the original score storing unit. A reduction level specifying means for the user to specify the degree of reduction;
3. The reduced score creating apparatus according to claim 1 or 2, wherein the note selection means determines selection of each note according to the degree of reduction designated by the reduction level designation means. Performance data acquisition means for acquiring performance data of the music by the performer,
A performance error position detecting means for detecting a performance error position from the performance data acquired by the performance data acquiring means;
Collating the performance error position detected by the performance error position detection means with the original score information of the music stored in the original score storage means and the abstract score information generated by the score abstraction means; Based on the abstract score information related to the performance error pattern specific to the performer determined by the performance error pattern determination means determined by the performance error pattern determination means for determining the performance error pattern specific to the performer. And a unique knowledge base generating means for constructing a performance difficulty level knowledge base specific to the performer,
The performance difficulty level determination means refers to the performance difficulty level knowledge base unique to the performer generated by the specific knowledge base generation means, and evaluates the performance difficulty level of each abstract score element. 4. A simplified score creating apparatus according to claim 1. Performance data acquisition means for acquiring performance data of the music by the performer,
A performance error position detecting means for detecting a performance error position from the performance data acquired by the performance data acquiring means;
Collating the performance error position detected by the performance error position detection means with the original score information of the music stored in the original score storage means and the abstract score information generated by the score abstraction means; Based on the abstract score information related to the performance error pattern specific to the performer determined by the performance error pattern determination means determined by the performance error pattern determination means for determining the performance error pattern specific to the performer. The simplified score creating apparatus according to claim 1, further comprising updating means for updating the performance difficulty level of the performance difficulty level knowledge base to information of the performer. Performance data acquisition means for acquiring performance data of the music by the performer,
A performance error position detecting means for detecting a performance error position from the performance data acquired by the performance data acquiring means, and the performance error position detected by the performance error position detecting means are stored in the original score storage means. A performance error pattern determining means for determining a performance error pattern unique to the performer by comparing with the original score information of the music and the abstract score information generated by the score abstraction means;
The original score storage means stores the degree of difficulty of each song in addition to the score information,
By comparing the performance error pattern determined by the performance error pattern determination means with the performance difficulty level knowledge base, the performance level determination means for determining the performance level of the performer, and the performance level determination means And a reduction level determination means for determining a degree of reduction of the music based on the performance level of the performer and the difficulty level of the music stored in the original score storage means,
6. The reduced score creating apparatus according to claim 1, wherein the note selection unit determines whether or not each note is selected according to the degree of reduction determined by the reduction level determination unit. A program for creating a simplified score from the original score of a song,
Original score storage means for storing score information described in the original score of a song;
A performance difficulty knowledge base that stores and stores abstract score elements and performance difficulties,
A musical knowledge base for storing a group of rules for determining musical importance of individual notes on a score, and a computer having a note selection knowledge base for storing a group of rules for determining selection of notes;
A score abstraction means for analyzing the score information of the music stored in the original score storage means and generating abstract score information of the music;
The performance score of each abstract score element included in the abstract score information is evaluated by comparing the score score information of the music generated by the score abstract means with the performance difficulty knowledge base. Performance difficulty determination means,
Music that evaluates the musical importance of each note included in the musical score information by comparing the musical score information of the music stored in the original musical score storage means with the rule group described in the musical knowledge base Importance evaluation means,
The performance difficulty of each abstract score element obtained by the performance difficulty determination means and the musical importance of the notes obtained by the musical importance determination means are described in the note selection knowledge base. A simplified musical score functioning as a simplified musical score generating means for generating simplified musical score information based on the notes determined to be adopted by the note selecting means by means of matching with the rule group, and a note selecting means for determining the selection of each note. Creation program.