JP5678343B2 - Phrase data search device and program - Google Patents

Description

  The present invention relates to a phrase data search system for searching for desired phrase data with reference to a phrase data group collected from original music data in order to compose and arrange a certain section of music.

  2. Description of the Related Art Conventionally, an apparatus that divides music data into a plurality of sections and extracts data of each section as phrase data is known. For example, in the music information analyzing apparatus disclosed in Patent Document 1, the melody information of a desired song input as music information is analyzed for the length of the note information and the length of the rest to perform phrase division. In addition, a technique is also known in which phrases are collected from existing music pieces and converted into a database for use in music generation. For example, in the music generation system disclosed in Patent Document 2, each phrase extracted from a ready-made music in bar units has a key, chord, start, end, first phrase sound, last phrase sound, time signature, allowable tempo. A database is created by assigning attributes such as range.

JP-A-8-292762 JP 2006-178104 A

  However, in the conventional phrase dividing technique, when a phrase is extracted with a predetermined number of measures or a note string delimiter, the positional relationship in which each phrase is used in the original song cannot be known. Moreover, when the attribute is given to the extracted phrase like patent document 2, it can be discriminate | determined whether it is suitable to use for the start or the end of a music, About the phrase to which the attribute is not given The positional relationship used in the original music is not known, and the position within the bar cannot be determined from the attributes of the start and end of the music. In addition, when using a phrase from the database to compose and arrange a section of music, notes that were strong beats in the original music are used in weak beats, or vice versa. When a note with a weak beat is used as a strong beat, the sense of beat may be misaligned before and after the phrase or during playback, resulting in an uncomfortable hearing.

  In view of such circumstances, the present invention performs the phrase search in consideration of the deviation from the bar line or the beat position of the phrase data in the original music without impairing the beat feeling that the original phrase data had. An object of the present invention is to provide a phrase data search system that can be used for composition and arrangement.

  According to the main feature of the present invention, phrase storage means (PB) for storing a plurality of phrase data (Pr: Pr1, Pr2,...) Extracted from the original song data (Mr), and song data (Ms) to be edited ) And a planned position designation unit (R2, for designating a planned note position at the beginning of the editing range in the music data (Ms) acquired by the target music acquisition unit (R1). R6 to R7; B) and the deviation of the scheduled note position designated by the scheduled position designation means (R6 to R7) from the reference timing [the bar line or beat position (timing) immediately after the scheduled note position in the music data Ms) is detected. The deviation detection means (R8 to R10; B) for generating the designated deviation information (Gs) representing the detected deviation and the phrase storage means (PB) A reference representing a deviation from the reference timing (tr) in the original music data (Mr) for the phrase data reading means (P1) for reading data (Pr) and the phrase data (Pr) read by the phrase data reading means (P1) Deviation information extraction means (P2 to P5) for extracting deviation information (Gr), and reference deviation information (Gr) extracted by the deviation information extraction means (P2 to P5) are generated by deviation detection means (R8 to R10). The phrase data (Pr) read by the phrase data reading means (P1) when the two pieces of deviation information (Gs, Gr) match (P9, P11, P12 = YES). Candidate detection means (R11: P5 to P13; C) for detecting as editing area candidate phrase data (Prs) Phrase data search device (electronic music device) to perform (claim 1), and phrase storage means (PB) for storing a plurality of phrase data (Pr: Pr1, Pr2,...) Extracted from the original music data (Pr) Is acquired in a target music acquisition step (R1) for acquiring music data (Ms) to be edited and a target music acquisition step (R1) in a computer (electronic music device) that functions as a phrase data search device. In the music data (Ms), the planned position designation step (R2, R6 to R7) for designating the planned note position to be the head of the editing range and the planned note designation position designated in the planned position designation step (R6 to R7). A deviation from the reference timing [a bar line or a beat position (timing) immediately after the planned note position in the music data Ms) is detected and detected. A deviation detection step (R8 to R10) for generating designated deviation information (Gs) representing a deviation, a phrase data reading step (P1) for reading phrase data (Pr) from the phrase storage means (PB), and a phrase data reading step ( A deviation information extraction step (P2 to P5) for extracting reference deviation information (Gr) representing deviation from the reference timing (tr) in the original music data (Mr) for the phrase data (Pr) read in P1); The reference deviation information (Gr) extracted in the information extraction steps (P2 to P5) is compared with the designated deviation information (Gs) generated in the deviation detection steps (R8 to R10), and both deviation information (Gs, Gr) is obtained. If they match (P9, P11, P12 = YES), the phrase read in the phrase data reading step (P1) Candidate detecting step of detecting Zudeta the (Pr) as a candidate phrase data editing range (Prs) (R11: P5~P13) phrase data search program for executing the steps consisting a [Claim 3] is provided. Note that the parentheses indicate reference symbols, terms, and the like of the examples added for convenience of understanding, and the same applies to the following.

  The phrase data search device according to the present invention further comprises level setting means (R5; B) for setting a level for determining coincidence of the deviation information (Gs, Gr), and candidate detection means (P5 to P13; C) determines whether the deviation information (Gs, Gr) matches according to the level set by the level setting means (R5; B) (P8 to P12), and determines that the deviation information matches ( The phrase data (Pr) of P9, P11, P12 = YES) is set as candidate phrase data (Prs) (P13) [Claim 2].

  In the phrase data search system according to the main feature of the present invention (Claims 1 and 3), a plurality of phrase data (Pr: Pr1, Pr2,...) Previously extracted from the original music data (Mr) is stored in the phrase storage means (PB). To be prepared. When music data (Ms) to be edited is acquired (R1) and an edit range to be searched for phrase data is set for this (R2), a planned note position that is the head of the edit range is specified (R6). To R7), a deviation from the reference timing of the designated scheduled note position (a bar line or a beat position (timing) immediately after the scheduled note position in the music data Ms) is detected, and designated deviation information (Gs) indicating the detected deviation. Are generated (R8 to R10) (B). Next, the phrase data (Pr) is sequentially read from the phrase storage means (PB) (P1), and the reference deviation information (Gr) representing the deviation from the reference timing (tr) in the original music data (Mr) for the read phrase data. ) Is extracted (P2 to P5), and when the extracted reference deviation information (Gr) matches the designated deviation information (Gs) (P9, P11, P12 = YES), the read phrase data (Pr) is edited. It is detected as candidate phrase data (Prs) in the range (P5 to P13) (C). That is, when searching for a desired phrase from the phrase storage means (PB) in order to compose and arrange a certain section (editing range) in the target music data (Ms), the first scheduled note position in the set editing range Is detected from the reference timing (designated deviation Gs) (R8 to R10), and the phrase data (Pr) having the same deviation (reference deviation Gr) as the detected deviation (Gs) is the desired candidate phrase (Prs). (P5 to P13).

  Therefore, according to the present invention, the beat feeling of the original phrase (strength, how to make space, etc.) is taken into account by taking into account the deviation from the bar line or beat position (reference timing) of the phrase in the original music during phrase search. Can be reproduced and more natural phrase synthesis can be performed, and can be used for composition and arrangement without losing the beat feeling that the original phrase data had. Moreover, since the relationship between the strong beat and the weak beat that the original phrase had is maintained after the composition and arrangement, a stable audibility can be obtained. Furthermore, when making an aftertact song, it becomes easier to search for the phrase that was originally an outfact, based on the deviation information from the bar line.

  In the phrase data search system according to the present invention (Claim 2), a function (R5; B) for setting a level for determining coincidence of the deviation information (Gs, Gr) is provided. According to the set level, coincidence determination of the deviation information (Gs, Gr) is performed (P8 to P12). Therefore, according to the present invention, the degree of deviation determined as coincidence can be arbitrarily adjusted according to the level setting.

1 shows a configuration example of a phrase data search system according to an embodiment of the present invention. The example of the image of phrase data by one Example of this invention is shown. It is a part of example of a phrase data search process flow by one Example of this invention. It is the other part of the example of a phrase data search processing flow by one Example of this invention. It is a part of example of a candidate phrase extraction process flow by one Example of this invention. It is the other part of the example of a candidate phrase extraction process flow by one Example of this invention.

[Overview of system configuration]
FIG. 1 is a diagram showing a configuration example of a phrase data search system according to an embodiment of the present invention, and FIG. 1 (1) shows a hardware configuration example of a phrase data search device according to an embodiment of the present invention. This phrase data search device is a kind of computer having an electronic music information processing function. For example, an electronic music device such as an electronic musical instrument is used, and a phrase data search system is constructed in the phrase data search mode. This phrase data search device includes a central processing unit (CPU) 1, a random access memory (RAM) 2, a read only memory (ROM) 3, an external storage device 4, an input operation unit 5, a display unit 6, a sound source 7, a communication interface. (I / F) 8 and the like, and these elements 1 to 8 are connected to each other via a bus 9.

  The CPU 1 constitutes a data processing unit together with the RAM 2 and the ROM 3, and executes various music information processing including phrase data search processing according to a predetermined control program including a phrase data search processing program. The RAM 2 is used as a work area for temporarily storing various data necessary for these processes. For example, in the phrase data search mode, the RAM 2 is a target music storage area for holding music data (Ms) to be edited, A reference phrase storage area for holding phrase data (Pr) referenced for candidate extraction, a candidate phrase storage area called a “candidate list” for storing extracted candidate phrase data (Prs), and the like Provided. In addition, various control programs and various control data necessary for executing these processes are stored in the ROM 3 in advance.

  The external storage device 4 includes a storage medium such as a semiconductor memory such as an HD (hard disk), an FD (flexible disk), a CD (compact disk), a DVD (digital multipurpose disk), and a flash memory, and a driving device thereof. Arbitrary information such as a control program and various music data can be stored in an arbitrary storage medium, and phrases for managing various phrase data (Pr: Pr1, Pr2,...) Are stored in a predetermined storage medium. The database PB is constructed, and the phrase data can be stored in the phrase database PB in the phrase data creation mode. Further, the storage medium may be detachable or may be built in the phrase data search device, and the detachable recording medium includes a USB memory.

  The input operation unit 5 detects performance operations such as a keyboard, setting operations such as a key switch, and the operation of these operations and introduces operation information corresponding to the detected contents to the data processing units (1 to 3). And an operation detection circuit. The data processing unit controls each part of the phrase data search device based on the operation information, and for example, sends performance information corresponding to the performance operation input to the sound source 7 or in the target music storage area of the RAM 2 according to the operation of the performance operator. The operation is performed in the phrase data search mode or the like according to the writing or the operation of the mode setting button in the setting operator. The display unit 6 includes a display such as an LCD for displaying various information necessary for performance and setting, a built-in lamp attached to the switches in the input operation unit 5, and the like. Controls with commands, provides display support for performance and setting operations.

  The sound source 7 includes a sound source unit and a DSP, and generates a musical sound signal based on actual performance information obtained from performance operation information from the input operation unit 5 or based on music data read from the storage device 4. At the same time, a predetermined effect is given to the musical tone signal. The sound system 10 connected to the sound source 7 includes a D / A converter, an amplifier, a speaker, and the like, and generates a musical sound based on a musical sound signal output from the sound source 7.

  The communication I / F 8 includes a music wired I / F such as MIDI, a general-purpose network I / F such as USB, or a general-purpose short-range wireless I / F such as wireless Lan, and can communicate with the external electronic music apparatus MM. it can. For example, MIDI format music data can be acquired from the external electronic music device MM and stored in the storage device 4.

[Outline of phrase data search system]
FIG. 1 (2) is a functional block diagram of a phrase data search system according to an embodiment of the present invention. The data processing units (1-3) including the CPU 1 of the phrase data search device (electronic music device) include a phrase data extraction unit A that operates in the phrase data creation mode and a search condition setting unit B that operates in the phrase data search mode. , Can function as a candidate phrase extraction unit C and a phrase selection unit D. The phrase data storage unit PB provided in the external storage device 4 stores a large number of phrase data Pr: Pr1, Pr2,... That are referred to when searching for phrase data, and a phrase database is constructed. The reference symbol “Pr” generically represents the phrase data, and the phrase database is described with the same reference symbol “PB” as the phrase data storage unit.

  The phrase data extraction unit A, in the phrase data creation mode, from the song data (referred to as original song data) Mr that is the phrase data creation source, the note string information and the reference timing information that represent the note strings that constitute the phrase in time series Phrase data Pr including (tr) is extracted, and attribute information (also referred to as phrase attribute information) Ar representing various attributes of the extracted phrase data Pr is added and sent to the phrase database PB. The phrase database PB stores the phrase data Pr from the phrase data extraction unit A, and stores attribute information Ar in association with the phrase data Pr so that it can be searched according to the attribute.

  Here, the note string information of the phrase data Pr is composed of a plurality of note events composed of sounding timing and pitch information, and the reference timing information (tr) is information indicating the bar line and beat timing immediately after the head note event. It is. The attribute information Ar may be added to the main body of the phrase data Pr and stored, or may be stored in association with (linked to) the phrase data Pr separately from the phrase data Pr main body. Further, the attribute information (also referred to as phrase attribute information) Ar can include deviation information Gr indicating “deviation” with respect to the reference timing (tr) in the original music data Mr, and this deviation information Gr is also referred to as reference deviation information. be called.

<Example of phrase data extraction>
FIG. 2 is a diagram for conceptually explaining an example of phrase data according to an embodiment of the present invention. FIGS. 2 (1) and 2 (2) show original music data Mr and extracted phrase data, respectively. The note string which comprises Pr is shown. 2 (1) and 2 (2), the horizontal axis indicates a time axis in which time moves in the direction of the right arrow, and “1: 1”, “2: 1”,... Indicate “bar: beat”. (1 bar = 4 beats), a solid line represents bar line position or bar line timing, and a broken line represents beat position or beat timing. The vertical axis roughly indicates the tendency of pitch change, and note events constituting the original music data Mr or the phrase data Pr are represented by blocks N1, N2,. In FIG. 2 (1), the notation of preceding and following notes is omitted.

  The original music data Mr is read from the predetermined storage device 4, and the start and end of a note string constituting a desired phrase to be extracted from the original music data Mr by user operation (for example, note event N1 to note event N7). [FIG. 2 (1)], the phrase data extraction unit A converts the data of the instructed note sequence (for example, note events N1, N2,..., N7) to the start end (note event). N1) The phrase data Pr is extracted together with the reference timing tr representing the immediately following bar line timing or beat timing [FIG. 2 (2)], and the extracted phrase data Pr is stored in the phrase database NB. At that time, phrase attribute information Ar is also stored corresponding to the phrase data Pr.

  Phrase attribute information Ar, as shown in FIG. 2 (3), includes deviation information Gr from bar line and beat position, time signature, original music information (item information such as name related to original music data), resolution, and other phrase search. Required item information [for example, chords, keys (keys), time signature, tempo, etc.]. For example, in FIG. 2, when the phrase data Pra including the note strings N1 to N7 shown in (2) is extracted from the original music data Mr shown in (1), it corresponds to the reference timing tr as shown in (3A). Deviation information from the second bar line = “1 beat (480 clocks)”, deviation information from the same beat position “2: 1” = “1 beat (480 clocks)”, the time signature is “4/4”, The original music information is “XX”, the resolution is “16th notes (120 clocks)”, and so on. In this example, when there is no note N1 and the phrase data Prb consisting of the note strings N2 to N7 is extracted, as shown in (3B), the deviation information from the first bar line corresponding to the reference timing tr is shown. = “1/2 beat (240 clocks)”, deviation information from the same beat position “2: 1” = “1/2 beat (240 clocks)”, the time signature is “4/4”, and the original music information is “ OO "and the resolution is" 16th notes (120 clocks) ".

  Now, returning to FIG. 1 (2), the search condition setting unit B acquires the music data Ms as an editing target from the storage device 4 or the input operation unit 5 in the phrase data search mode, and acquires the acquired music data (target music) Various search conditions are set for a section to be edited in Ms (referred to as data). Here, the target music data Ms is music data stored in the storage device 4 that has been stored in advance or in the middle of production, or performance information that is input in response to a performance operation from the input operation unit 5. It can be obtained by reading into the area. Further, the search condition can include “designated deviation information” Gs that represents a deviation from the reference timing in the target music data Ms in preparation for a search by a beat feeling reproduction method. It is extracted by the same method as the deviation information Gr. In other words, when the planned note position that is the head of the editing range is designated by the user operation, it represents a deviation from the reference timing (the bar line or beat timing immediately after the planned note position in the music data Ms). Designated deviation information Gs is generated.

  The candidate phrase extraction unit C accesses the phrase database PB, refers to the phrase data Pr: Pr1, Pr2,... According to the search conditions set by the search condition setting unit B, and selects phrase data that matches the search conditions from these. The extracted phrase data Pr is stored in the candidate list in the RAM 2 as candidate phrase data Prs. For this phrase search, in addition to a general key search method that searches for candidates with search keys such as chord, key (tone), time signature, tempo, etc., a beat feeling reproduction method using deviation information can be used, When searching with the beat feeling reproduction method, candidate phrase data Prs having reference deviation information Gr that matches the designated deviation information Gs is extracted, and the difference “Gs−Gr” (reference symbol “ Gs ”and“ Gr ”may represent“ deviation ”itself), and are stored in association with the extracted candidate phrase data Prs. When the specified deviation Gs is smaller than the reference deviation Gr and there is a preceding note event in the range from the “position going back by the reference deviation Gr from the reference timing” to the scheduled note position in the target music data Ms, Since the phrase data Pr having the reference deviation Gr overlaps the preceding note event, it is deleted from the candidate phrase Prs (see step R12 in FIG. 4).

  The phrase selection unit D selects the optimal phrase data Prd from the candidate phrase data Prs accumulated in the candidate list, and the selected optimal phrase data Prd is applied to the editing range of the target music data Ms and is the target. It is incorporated in the music Ms. In this case, the optimum phrase data Prd is selected according to the user operation. However, when the beat feeling reproduction method is used, an automatic selection method according to the magnitude of the difference “Gs−Gr” is used. The number of candidate phrase data Prs can be narrowed down, and one candidate phrase data narrowed down or one candidate phrase data selected according to user operation from a plurality of narrowed down candidate phrase data Prs is optimal. Phrase data Prd is determined. Then, the determined optimum phrase data Prd is overwritten on the target music data Ms in accordance with the reference timing.

  As described above, in the phrase data retrieval system of the beat feeling reproduction method according to one embodiment of the present invention, a plurality of phrase data Pr: Pr1, Pr2,... Previously extracted from the original music data Mr are prepared in the phrase storage unit PB. When the editing range is set for the target music data Ms, the pronunciation position of the first note in the editing range is specified as the planned note position, and the reference timing of the specified planned note position (immediately after the planned note position) Designated deviation information Gs representing a deviation from a bar line or beat position appearing in (b) is generated (B). Next, the phrase data Pr is sequentially read from the phrase storage unit PB, and for each read phrase data, reference deviation information Gr representing a deviation with respect to the reference timing tr in the original music data Mr is extracted, and both deviation information Gr, Gs is obtained. If they match, the phrase data is detected as candidate phrase data Prs suitable for the set editing range (C).

[Example of processing flow]
3 to 4 are flowcharts showing an operation example of phrase data search processing according to one embodiment of the present invention. This phrase data search device operates in the phrase data search mode by operating the search mode setting button when operating in the standby mode, actual performance mode, automatic performance mode, etc., and the phrase data search of FIG. Processing can be performed. When the phrase data search process starts, the CPU 1 first prepares music data to be edited, that is, target music data Ms, in the target music storage area of the RAM 2 in step R1. That is, the target music data Ms is read from the storage device 4 or the like, and necessary display processing is performed. If the target music data Ms is composed of a plurality of parts, a part to be edited (referred to as a target part) is set, and as necessary. Then, the target music data Ms is quantized or the note information is input from the input operation unit 5, and the search screen is displayed on the display unit 6.

  In the next step R2, for example, in response to a user operation on the target music data Ms displayed on the search screen with a note string as shown in FIG. Then, an interval for performing phrase search in the target music data Ms, that is, an edit range is designated, and a phrase search method is selected in the subsequent step R3, and the process proceeds to step R4. In step R4, it is determined whether or not the selected phrase search method is a beat feeling reproduction method, and if it is a beat feeling reproduction method (R4 = YES), the beat feeling reproduction method in steps R5 to R12 (FIG. 4). The candidate phrase data Prs is extracted from the phrase database PB by executing the search processing routine according to the above. On the other hand, when it is determined in step R4 that the selected search method is not the beat feeling reproduction method (R4 = NO), the process proceeds to step R13, and the candidate phrase data Prs is followed according to another phrase search method selected in step R3. To extract. For example, the phrase database PB is searched using a search key such as chord, key, time signature, or tempo, and desired candidate phrase data Prs is extracted.

  Regarding the search processing by the beat feeling reproduction method, first, in step R5 (FIG. 4), the beat feeling reproduction search screen is displayed on the display of the display unit 6, and the search level is set according to the level setting operation for the screen. In this system example, search levels include “complete match”, “within 1/2 beat”, and “no level”. If there is no level setting operation, “no level” is set. The phrase data Pr having the reference deviation information Gr within a predetermined allowable upper limit value (for example, one beat) is extracted.

  In the next step R6, the input of the note event data to the section (editing range) designated in the target music data Ms (or target part) is accepted as necessary. That is, in the beat feeling reproduction method, the phrase data Pr starting from the position of the designated note is extracted. Therefore, if there is no note at the start position (start) of the designated section (editing range), the user is prompted to input the note, If there is a note input operation, note event data input is accepted accordingly. Subsequently, in step R7, the first note event data in the designated section is extracted, and the sound generation timing is designated as the note planned position.

  Next, in step R8, according to the user operation, it is selected whether the reference line of the planned note position is based on the bar line or the beat position, and in the next step R9, the bar line is used as a reference. Searches for the bar position immediately after the planned note position, and when using the beat position as a reference, searches for the beat position immediately after the planned note position and sets the searched bar line position or beat position as the reference timing. Further, in step R10, a deviation (number of clocks, etc.) of the scheduled note position with respect to the timing reference information is detected, and deviation information on the target music data Ms side indicating the detected deviation (designated deviation), that is, designated deviation information Gs is obtained. Then, the process proceeds to Step R11.

  In step R11, the phrase database PB is accessed, and the “candidate phrase extraction process” is executed by searching for phrase data Pr in which the deviation information Gs and Gr match (according to the level) and setting it as a candidate Prs. After the candidate phrase extraction process, the process proceeds to step R12, and the reference timing in the target music data Ms is referred to with reference to the deviation information Gr from the bar line or beat position for each phrase data Pr determined as the candidate Prs in the candidate phrase extraction process. It is checked whether or not there is a note event having a sounding timing earlier than the expected note position within a range that is traced back by the shift information Gr referred to from the position, and if such a note event exists, the phrase data Pr Is deleted from the candidate Prs.

  After the process of step R12 (FIG. 4) or step R13 (FIG. 3), the process proceeds to step R14 (FIG. 3) to ask the user whether to change the search method and to change the search method. (R14 = YES), the process returns to step R3, the processes of steps R3 to R14 are repeated, and if the search method is not changed (R14 = NO), the process proceeds to step R15.

  In step R15, desired phrase data Prd is selected from the candidate phrase data Prs obtained by the processing of R3 to R13 according to the user's selection operation or automatic predetermined selection method, and the desired phrase data Prd is selected from the target song data Ms. Applies to the specified section (edit range). Here, when the desired phrase data Prd is selected from the candidate phrase data Prs obtained by the beat feeling reproduction method (R5 to R12), the bar line or beat position immediately after the opening note event in the desired phrase data Prd is determined. The desired phrase data Prd is applied by overwriting in accordance with the position of the reference timing in the target music data Ms.

  After the phrase data selection and application process in step R15, in step R16, the user is asked whether or not to perform phrase search for another section (editing range), and the user performs a phrase search for another section. When there is an input (R16 = YES), the process returns to step R2, and the processes of steps R2 to R16 are repeated. If it is determined in step R16 that there is a user input indicating that the phrase search is not performed (R16 = NO), the phrase data search process is terminated and the original mode is restored.

<Candidate phrase extraction processing>
5 and 6 are flowcharts showing an example of candidate phrase extraction processing according to one embodiment of the present invention. This candidate phrase extraction process is executed in step R11 of the phrase data search process. In the first step P1, the CPU 1 refers to the phrase database PB and newly sets the phrase data Pr that matches the current resolution of the target music data to 1 The data is read into the reference phrase storage area of the RAM 2 as an object to be referred to. In the next step P2, it is determined whether or not the reference deviation information Gr is included in the attribute information Ar of the phrase data Pr to be referred to. When the reference deviation information Gr is not included in the attribute information Ar (P2 = NO), the process proceeds sequentially to steps P3 to P4, and the deviation (reference deviation) Gr of the phrase data Pr from the reference timing is detected. That is, in step P3, the time from the first note to the immediately following bar line is extracted based on the sounding timing information and the reference timing information tr of the first note event in the phrase data Pr. Is stored in the first reference shift storage area of the RAM 2 as “shift information Gr”. In step P3, the time from the first note to the immediately following beat position is extracted based on the sounding timing information and the reference timing information tr of the first note event in the phrase data Pr, and the phrase data “from the beat position” is extracted. It is stored in the second reference shift storage area of the RAM 2 as “shift information Gr”.

  When it is determined in step P2 that the reference deviation information Gr is included in the attribute information Ar (P2 = YES) or after the reference deviation detection process in step P4, the process proceeds to step P5, where the current bar line is used as a reference. It is determined whether it is selected (set). Here, when the bar line is set as a reference (P5 = YES), in step P6, the designated deviation information Gs extracted on the target music data Ms side and the phrase data Pr "deviation information Gr from beat position". After the difference is detected, the process proceeds to step P8, and when the beat position is set as a reference (P5 = NO), in step P7, the designated beat information Ms on the target music data Ms side and the phrase data Pr After detecting the difference from the position deviation information Gr ", the process proceeds to Step P8.

  In step P8, it is determined whether or not “complete match” is set as the search level. If “complete match” is set (P8 = YES), the difference between the deviation information Gs and Gr is determined in step P9. Is “0” or not, and when the difference in the deviation information = “0” (P9 = YES), the process proceeds to Step P13 (FIG. 6). If “complete match” is not set (P8 = NO), it is determined in step P10 (FIG. 6) whether “within 1/2 beat” is set as the search level. If "/ 2 beats or less" is set (P10 = YES), it is determined in step P11 whether or not the difference between the deviation information Gs and Gr is within 1/2 beat, and the difference between the deviation information ≦ “ When it is “1/2 beat” (P11 = YES), the process proceeds to Step P13. Further, when “no level” is not set for the search level (P10 = NO), in step P12, the difference between the deviation information Gs and Gr is an allowable upper limit value (for example, 1 beat). ) Or not, and if the difference in deviation information ≦ the allowable upper limit value (P12 = YES), the process proceeds to Step P13. In step 13, the phrase data Pr to be referenced is selected as a candidate Prs and recorded in the candidate list in the RAM 2. When the candidate phrase data Prs is recorded in the candidate list, the difference between the deviation information Gs and Gr is also recorded in association with each candidate phrase data Prs.

  In step P9, the difference between the deviation information Gs, Gr is not “0”, in step P11, the difference between the deviation information Gs, Gr exceeds 1/2 beat, or in step P12, the difference information Gs, Gr When it is determined that the difference exceeds the allowable upper limit value (P9 = NO, P11 = NO or P12 = NO), or after the list recording process in step 13, the process proceeds to step P14 and can be referred to. It is determined whether or not the data Pr still remains. Here, when the referenceable phrase data Pr remains (P14 = YES), the process returns to step P1 and the processes of steps P1 to P14 are repeated. If the referenceable phrase data Pr does not remain (P14 = NO), if there is a difference in the deviation information Gr of the candidate phrase data Prs in step P15, the candidate Prs are rearranged in the order of the small difference. The candidate phrase extraction process is terminated, and the process returns to step R12 of the phrase data search process.

[Various Embodiments]
The preferred embodiment of the present invention has been described above 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. For example, the user may specify the scheduled note position at the beginning of editing in the target music data, each time, for example, when the music data is automatically or manually divided into several sections. You may make it detect automatically, for example using the note position of the head in each area. For example, it is effective to utilize this function at key points in the composition of music, such as the start of rust and the start of the main melody.

  When the time signature is different between the target music data (Ms) and the reference phrase data (Pr), for example, the bar time or beat position as the reference timing of the phrase data is matched with the corresponding position of the target music data, and the time signature on the target music data side Adjust with.

  The shift information may be the number of clocks or may be expressed by a note length such as a quarter note or an eighth note. Alternatively, it may be a beat number. When the reference timing is a bar line, a combination of the number of beats and the number of clocks may be used. When referring to the deviation information, the match may be checked with the actual number of clocks, but the target music data and each phrase data to be searched are previously quantized with the same resolution, and the deviation in that state is determined. If the information is referred to, the types of phrase data extracted as candidates by the search become rich.

  The phrase data may be stored for each quantizing resolution, or may be quantized with a specified resolution for each search. It is sufficient that a plurality of types of resolutions can be selected. Further, when any phrase data is selected as the optimum phrase after the candidate phrase data is extracted, a note data string is generated by matching the reference bar line or beat position based on the phrase data. Each pitch conversion at that time is performed in accordance with the chord prepared on the target musical piece side effective at each note position, in accordance with the range of sound before and after each note, or in accordance with the user-specified pitch. Or the method of performing according to tonality etc. is taken.

  The phrase data group to be referenced may be stored in the device, or may be stored in an external storage device connected to the network. Each phrase data includes not only a note event but also an event related to expression such as pitch bend data.

  If the phrase data stores deviation information from the bar line and deviation information from the beat position as attribute information, the attribute information of each phrase may be referred to when searching for the phrase. Also, if no deviation information is stored in the attribute information, the first note event data in the phrase data and the deviation information from the bar line and beat position that comes immediately after it are calculated and stored as attribute information. Also good. You may make it memorize | store together the resolution | decomposability at the time of deviation | shift information calculation. Furthermore, if the shift information is not added to the phrase data as attribute information, if the music data that is the source of the phrase data can be identified and referenced, the original music data is referred to and the phrase data of the original music data is referred to. Deviation information from the first note position from the bar line or beat position immediately after the note may be calculated and stored as phrase data attribute information.

PB Phrase database or phrase data storage unit,
Mr, Ms Original music data and target music data,
Pr: Pr1, Pr2, ...; Pra, Prb Phrase data (also called phrase),
Prs, Prd candidate [phrase data] and optimum phrase data,
Ar; Ara, Arb [phrase] attribute information,
Gr, Gs [reference] deviation information and designated deviation information (may indicate deviation itself),
N1, N2, ... Note events,
tr Reference timing [information].

Claims (3)

Phrase storage means for storing a plurality of phrase data extracted from the original music data;
Target music acquisition means for acquiring music data to be edited;
A planned position designating unit for designating a planned note position at the beginning of the editing range in the music data acquired by the target music acquisition unit;
Deviation detection means for detecting a deviation of a planned note position designated by the planned position designation means with respect to a reference timing which is a timing of a bar line or a beat position immediately after the planned note position, and generating designated deviation information representing the detected deviation. When,
Phrase data reading means for reading phrase data from the phrase storage means;
With respect to the phrase data read by the phrase data reading means, deviation information extracting means for extracting reference deviation information representing deviation from the reference timing which is the bar line and beat timing of the first note in the original music data immediately after the first note ; and
The reference deviation information extracted by the deviation information extracting means is compared with the designated deviation information generated by the deviation detecting means, and when both deviation information matches, the phrase data read by the phrase data reading means is selected as the editing range candidate. Candidate detection means for detecting as phrase data ;
A phrase data search device comprising: further,
Comprising level setting means for setting a level for determining the coincidence of both deviation information;
The candidate detecting means performs matching determination of both deviation information according to the level set by the level setting means, and uses the phrase data when it is determined that the deviation information matches as the candidate phrase data. The phrase data search device according to Item 1. A computer having a phrase storage means for storing a plurality of phrase data extracted from the original music data, and functioning as a phrase data search device,
A target music acquisition step for acquiring music data to be edited;
A planned position designation step for designating a planned note position that is the beginning of the editing range in the music data acquired in the target music acquisition step;
A deviation detection step of detecting a deviation of the planned note position designated in the planned position designation step with respect to a reference timing that is a timing of a bar line or a beat position immediately after the planned note position and generating designated deviation information representing the detected deviation. When,
A phrase data reading step for reading phrase data from the phrase storage means;
About the phrase data read in the phrase data reading step, a deviation information extraction step for extracting reference deviation information indicating deviation from the reference timing which is the bar line and beat timing of the first note in the original music data , and the beat timing immediately after the first note ;
The reference deviation information extracted in the deviation information extraction step is compared with the specified deviation information generated in the deviation detection step. If both deviation information matches, the phrase data read in the phrase data reading step is used as a candidate for editing range. Candidate detection step for detecting as phrase data ;
A phrase data search program that executes a procedure consisting of:

Images