JP2017078829A - Performance analysis device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To transmit, to a player, a performance pattern when the player executes a wrong performance.SOLUTION: When the number of mistakes is counted for every performance pattern, performance patterns causing the number of mistakes equal to or more than a predetermined threshold are regarded as mistakable performance patterns for the player. For example, when the threshold is 3 and Figure 5 (A) is cited, the number of mistakes of performance patterns in performance numbers x to x+2 in an upper column of Figure 5 (A) is more than the threshold, so that it is defined to be mistakable, however, the number of mistakes of performance patterns in performance numbers y to y+2 in a lower column of Figure 5 (A) is less than the threshold, so that it is defined not to be mistakable. Further, when the upper column and the lower column are compared, only difference between the performance patterns is string crossing from D to A (D→A) and string crossing from D to G (D→G), so that it is estimated that a cause of the mistake of the performance is string crossing from D→A.SELECTED DRAWING: Figure 5(A)

Description

  The present invention relates to a technique for analyzing a performance of a performer.

  For example, Patent Document 1 proposes a mechanism for learning how to play a musical instrument. In order for a performer to improve his performance, it is important to know the content of the performance he is not good at and try to improve that point.

JP 2006-284638 A

According to the mechanism described in Patent Document 1, it is possible to specify the performance location that was performed incorrectly in the music performed by the performer, but the suggestion from the viewpoint of why the performer is not good at the performance location. Can not give.
Therefore, an object of the present invention is to convey a performance pattern to the performer when an incorrect performance is performed.

  In order to solve the above problems, the present invention provides performance data acquisition means for acquiring performance data generated based on a performance by a performer, the performance data acquired by the performance data acquisition means, and an example of the performance. Comparison means for comparing the reference data, and a statistical processing means for calculating the performance pattern of the performance at the erroneous location specified by the comparison means by statistical processing. Provided is a performance analysis apparatus characterized by comprising:

  The statistical processing means may specify a combination of fingers used for a performance identified as an error location as the performance pattern.

  The statistical processing means may identify a combination of strings used for a performance identified as an error location as the performance pattern.

  The statistical processing means may identify a combination of positions in a performance identified as an error location as the performance pattern.

  The statistical processing means may identify the performance pattern from a performance identified as an error location to a performance that is temporally prior to the performance.

  The statistical processing means may identify the performance pattern from a performance identified as an error location to a performance later in time than the performance.

  The statistical processing means may determine the time range of the performance pattern based on the performance period or the number of pronunciations by performance.

  The statistical processing means may change the performance period for determining the time range of the performance pattern or the number of pronunciations by the performance according to a predetermined condition.

  Further, the present invention provides a performance data acquisition means for acquiring performance data generated based on a performance by a performer, the performance data acquired by the performance data acquisition means, and an example of the performance. Comparison means for comparing the reference data with each other to specify the wrong place of the performance by the performer, and statistical processing means for calculating the performance pattern of the performance at the wrong place specified by the comparison means by statistical processing Provide a program.

  According to the present invention, it is possible to inform a performer of a performance pattern when an incorrect performance is performed.

It is a block diagram which shows the structure of the whole analysis system which concerns on one Embodiment of this invention. It is a block diagram which shows the hardware constitutions of an analyzer. It is a block diagram which shows the function structure of an analyzer. It is a flowchart which shows operation | movement of an analyzer. It is a figure which shows an example of the analysis content. It is a figure which shows an example of the analysis content. It is a figure which shows an example of the analysis content. It is a conceptual diagram for demonstrating a statistical process.

An embodiment of the present invention will be described.
[1. Constitution]
FIG. 1 is a block diagram showing the overall configuration of an analysis system 1 according to this embodiment. The analysis system 1 includes a plurality of musical instruments 2a, 2b, and 2c, an analysis device 3, and a network 4 that connects these devices so that they can communicate with each other. The network 4 is, for example, ISDN (Integrated Services Digital Network) or the Internet, and includes a wired section or a wireless section.

  The musical instruments 2a, 2b, and 2c may be electronic musical instruments such as electronic keyboard musical instruments, electronic stringed musical instruments, and electronic wind instruments, or may be acoustic instruments such as ordinary pianos, violins, and various wind musical instruments that are not electronic musical instruments. . When the musical instruments 2a, 2b, and 2c are electronic musical instruments, performance data (for example, MIDI format) generated based on the performance by the performer is generated and transmitted to the analysis device 3 via the network 4. When the musical instruments 2a, 2b, and 2c are ordinary musical instruments, for example, WAVE format or MP3 (MPEG) that has been subjected to A / D conversion through voice processing from a microphone (not shown) arranged in the vicinity of the musical instrument itself or the musical instrument. Audio layer-3) format audio data is transmitted to the analysis device 3 via the network 4. Although three musical instruments are illustrated in FIG. 1, the number of musical instruments included in the analysis system 1 is not limited to three, and may be more or less. Since the musical instruments 2a, 2b, and 2c all have the same configuration and operation, they are simply referred to as “musical instrument 2” when they do not need to be distinguished from each other.

  The analysis device 3 is a computer that functions as a performance analysis device that analyzes performance data acquired from the musical instrument 2. More specifically, the analysis device 3 identifies the wrong place where the wrong performance is performed based on the performance data acquired from the musical instrument 2c, and calculates the performance pattern of the performance at the identified wrong place by statistical processing. And information on the performance pattern is output.

  FIG. 2 is a block diagram illustrating a hardware configuration of the analysis device 3. The control unit 31 is, for example, a CPU, and controls each unit of the analysis device 3 by reading and executing a computer program stored in the storage unit 32. The storage unit 32 is a large-capacity storage unit such as a hard disk. The storage unit 32 stores the computer program and a plurality of performance data acquired from the musical instrument 2c as a performance database. The computer program stored in the storage unit 32 includes reference data for each piece of music indicating the performance contents that serve as a model for the performer. The communication unit 33 performs data communication with the musical instrument 2 via the network 4 under the control of the control unit 31.

  FIG. 3 is a block diagram illustrating a functional configuration of the analysis apparatus 3. The performance data acquisition unit 301 is a function realized by the control unit 31 and the communication unit 33, and acquires performance data generated based on the performance by the performer via the network 4. The reference data storage unit 302 is a function realized by the storage unit 32 and stores the above-described reference data. The comparison unit 303 is a function realized by the control unit 31. The comparison unit 303 compares the performance data acquired by the performance data acquisition unit 301 with the reference data stored in the reference data storage unit 302. Identify mistakes in performance. The statistical processing unit 304 calculates the performance pattern of the performance at the wrong place specified by the comparison unit 303 by statistical processing. The output unit 305 outputs information on the performance pattern calculated by the statistical processing unit 304 to, for example, a communication terminal (not shown) or a musical instrument 2 (in the case of an electronic musical instrument) operated by the performer. The performance analysis apparatus according to the present invention includes a performance data acquisition unit 301, a reference data storage unit 302, a comparison unit 303, and a statistical processing unit 304, which are configured as shown by a solid line in FIG. The output unit 305 having the configuration shown is not necessarily an essential configuration and may or may not be provided.

[2. Operation]
Next, the operation of this embodiment will be described with reference to the flowchart shown in FIG. The following is an example when the musical instrument 2 is an electronic musical instrument. The performer operates the operation unit (not shown) of the musical instrument 2 to select a song to be performed, and then performs the performance using the musical instrument 2. The performance data generated based on this performance and the music identification information for identifying the selected music are transmitted from the musical instrument 2 to the analysis device 3. The control unit 31 of the analysis device 3 acquires performance data and song identification information through the communication unit 33 (performance data acquisition unit 301) (step S1).

  The control unit 31 (comparison unit 303) of the analysis device 3 reads the reference data of the music corresponding to the music identification information from the storage unit 32 (reference data storage unit 302), and compares the read reference data with the acquired performance data. Then, the wrong place of the performance by the performer is specified (step S2). Specifically, the control unit 31 first selects either the performance data or the reference data so that the performance data and the reference data can be associated with each other even when the performance data and the reference data are shifted in time. Time normalization (DTW: Dynamic Time Warping) is performed to expand or contract one of the time axes, and both time axes are aligned. As a method for performing this DTW, there is DP (Dynamic Programming) matching, which may be used. Then, the control unit 31 compares the pitch represented by the reference data with the pitch of the performance sound represented by the performance data, and identifies the difference between the two. For example, the difference between the two voices is specified, such as how much the pitch of one voice is higher or lower than the pitch of the other voice. Further, when the musical instrument 2 is a wind instrument, if the difference between the pitch represented by the reference data and the pitch of the performance sound represented by the performance data is a specific value, a difference in harmonics (a higher harmonic will be unexpectedly generated). Therefore, the control unit 31 also identifies such a difference. In addition, the control unit 31 compares the sound generation timing represented by the reference data with the timing at which the pitch of the performance data is switched (that is, the sound generation timing of the performance sound) to identify the difference between the two. For example, the difference between the two is specified, such as how early or late the sounding timing of one sound is compared to the sounding timing of the other sound. Further, the control unit 31 may specify the difference between the reference data and the performance data, for example, with respect to a pitch fluctuation curve representing a state in which the tone color, length, or pitch of the performance sound changes in time series. The control unit 31 identifies such differences for all of the corresponding portions of the reference data and the performance data, and places where the difference exceeds a predetermined threshold as incorrect portions where the performance by the performer is wrong. Identify. Further, when recording with a microphone using a piezo element for performance using a wind instrument, noise due to a key is generated. Since the generation period of this noise is preferably as short as possible, the control unit 31 stores reference data and performance data. In comparison, a period in which noise is generated is specified as a difference between the two, and a part where the period exceeds a threshold is specified as an erroneous part in which the performance by the performer is wrong. In other words, when playing a wind instrument, it is necessary to move several fingers at the same time quickly. However, if there are variations in the movement of the fingers, sounds different from the reference data may be mixed or the rhythm of the performance may be twisted. So detect such mistakes.

  And the control part 31 judges whether the process of step S1, S2 was performed about the required number of performance data (step S3). This required number may be, for example, the number of times determined in advance as the number of times that the same performer has performed the same song, and may be, for example, one or more. The processes of steps S1 and S2 are repeated until the required number of performance data is reached (step S3; NO).

  When the processing of steps S1 and S2 is performed on the required number of performance data (step S3; YES), the control unit 31 (statistical processing unit 304) calculates the performance pattern of the performance at the specified error location by statistical processing. (Step S4). This statistical processing will be described with reference to FIGS.

  FIG. 5A illustrates an example of a performance pattern when a performance portion of a song is played with a violin. The performance number is a number assigned to each performance sound from the start to the end of the song. The line is the number assigned to the string used when playing the performance sound. The position is a number corresponding to the position of the hand when performing the performance sound. The finger is a number assigned to the finger used when playing the performance sound. As shown in the upper example of FIG. 5A, the combination of lines, positions, and fingers when performing a performance including three performance sounds that are continuous in time corresponds to the performance pattern in performance numbers x to x + 2. To do. At this time, if the performance of performance number x + 2 is an incorrect place, the performance patterns in performance numbers x to x + 2 that are earlier in time are specified as the performance pattern of the incorrect place. On the other hand, in the example in the lower part of FIG. 5A, if the performance number y + 2 is an incorrect place, the performance patterns in the previous performance numbers y to y + 2 are specified as the incorrect place performance patterns. The

  FIG. 6 is a conceptual diagram illustrating statistical processing. The horizontal axis of the graph of FIG. 6 is all performance patterns that appear in a certain musical piece, and the vertical axis is the number of times that the performance pattern is determined to be an error. For example, performance pattern 1 is a performance pattern at performance numbers x to x + 2 in FIG. 5A, performance pattern 2 is a performance pattern at performance numbers y to y + 2 in FIG. If there is a performance error in these performance patterns, the number of errors on the vertical axis in FIG. 6 is incremented. In this way, when the number of errors is counted for every performance pattern, it can be said that a performance pattern in which the number of errors more than a predetermined threshold has occurred is a performance pattern that is easily mistaken for the performer.

  For example, when the threshold value in FIG. 6 is set to 3 and the example in FIG. 5A is considered, the performance pattern in the performance numbers x to x + 2 in the upper part of FIG. However, since the number of performance pattern mistakes in performance numbers y to y + 2 in the lower part of FIG. And when comparing the upper and lower tiers, the only difference in the performance pattern is D → A or D → G. It can be estimated that there is (step S5).

  FIG. 5B illustrates a combination of fingers simultaneously pressed by a wind instrument. In this example, the finger movement of 5 → 4 and the finger movement of 8 → 7 in the wrong place are 2 The movement of the fingers of the book occurs simultaneously. In such a case, since the number of mistakes is equal to or greater than the threshold, it can be estimated that the performance of movement of two fingers is the cause of the mistake (step S5).

  FIG. 5C shows another example of a combination of fingers simultaneously pressed by a wind instrument. Comparing the upper and lower tiers in this example, the common point is that the performance pattern related to finger movement is that the number of mistakes is greater than or equal to a threshold value in performance involving finger movement of 8 → 7. For this reason, it can be estimated that the performance with the finger movement of 8 → 7 is the cause of the mistake (step S5).

  When the control unit 31 (statistical processing unit 304) calculates the performance pattern of the performance at the wrong place by statistical processing as described above, it outputs information on the performance pattern (step S6). Specifically, the control unit 31 outputs a performance pattern (that is, a combination of at least one of a line, a position, and a finger) corresponding to the error location, or the cause of the performance error estimated as described above. Output processing is performed, such as outputting or prompting a practice by presenting a tune with frequent performance equivalent to the cause.

  As described above, according to the present embodiment, it is possible to inform a performer of a performance pattern when an incorrect performance is performed.

[3. Modified example]
The above-described embodiment may be modified as follows.
[3-1] In the embodiment, a performance pattern extends from a performance identified as an error location to a performance temporally prior to the performance (performance in a period that is backed by three performance sounds including the performance location). However, the rule for specifying the performance pattern is not limited to the example of the embodiment.
For example, the control unit 31 (statistical processing unit 304) may specify a performance pattern from a performance specified as an error location to a performance later in time than the performance. For example, when an error occurs in the performance number x, the performance pattern in the performance numbers x to x + 2 may be specified as the performance pattern in the wrong place, and the performance pattern in the performance numbers x-1 to x + 1 is specified. You may specify as a performance pattern of a mistake part.

  The control unit 31 (statistical processing unit 304) may determine the time range of the performance pattern based on the performance period or the number of sounds produced by the performance (number of performance sounds). For example, in the embodiment, the number of performance sounds is three, but this number is arbitrary as long as it is two or more. Further, a performance period may be used instead of the number of performance sounds. For example, when an error occurs in the performance number x, the performance pattern for 5 seconds before the performance number x may be specified as the performance pattern of the error location.

  Further, the control unit 31 (statistical processing unit 304) may change the performance period for determining the time range of the performance pattern or the number of pronunciations by the performance according to a predetermined condition. As predetermined conditions at this time, for example, the player's attributes such as the player's performance level and practice amount may be used, and the attributes of the song to be played such as the difficulty level and tempo of the song may be used. It may be used. In addition, the performer himself / herself may designate the performance period for determining the time range of the performance pattern or the number of pronunciations by the performance.

[3-2]
In the embodiment described above, the above analysis is performed based on performance data corresponding to one performer. However, the above analysis may be performed on a plurality of performers as a whole. In this case, as a result of the analysis by the analysis device 3, it is possible to present a performance pattern and its cause that are likely to cause a performance error in the entire performer group.

[3-3]
Although FIG. 6 shows an example in which the number of mistakes is statistically processed for each performance pattern, the number of errors may be statistically processed for each performance pattern for each type of error. For example, in the case of a performance by a wind instrument, the control unit 31 is determined by a simple pitch (fingering) error as described above, a pitch error due to a so-called lead error, or noise when recording using a piezo microphone. Statistical processing may be performed by incrementing the number of performance mistakes for each type of mistake, such as mistakes.

[3-4]
In the above-described embodiment, the analysis system 1 in which the musical instrument 2 and the analysis device 3 are connected via a communication network realizes all the functions according to the present embodiment. On the other hand, three or more devices connected by a communication network may share the above functions, and a system including the plurality of devices may realize a system equivalent to the present embodiment. For example, a plurality of devices connected to the network 4 include the performance data acquisition unit 301, the reference data storage unit 302, the comparison unit 303, and the statistical processing unit 304 illustrated in FIG. 3, and these plurality of devices cooperate. The same function as the analysis device 3 according to the embodiment may be realized. In this case, the performance analysis device according to the present invention is realized by the plurality of devices.
Alternatively, one device may realize all the functions according to the present embodiment.

[3-5]
The present invention may be implemented as a performance analysis method performed by the analysis device 3 or a program executed by the analysis device 3 in addition to a performance analysis device such as the analysis device 3, for example.
The performance analysis method according to the present invention is a performance data acquisition step for acquiring performance data generated based on a performance by a performer, the performance data acquired in the performance data acquisition step, and an example of the performance. Comparing with reference data, a comparison step for specifying a wrong place of the performance by the performer, and a statistical processing step for calculating a performance pattern of the performance at the wrong place specified in the comparison step by statistical processing are provided. It is characterized by.
The programs executed by the control unit 31 of the analysis apparatus 3 are magnetic tape, magnetic disk, flexible disk, optical recording medium, magneto-optical recording medium, CD (Compact Disk) -ROM, DVD (Digital Versatile Disk), RAM. Or the like stored in a recording medium. It is also possible to download to the analysis device 3 via a network such as the Internet.

DESCRIPTION OF SYMBOLS 1 ... Analysis system, 2a, 2b, 2c ... Musical instrument, 3 ... Analysis apparatus, 4 ... Network, 31 ... Control part, 32 ... Storage part, 33 ... Communication part, 301 ... Performance data acquisition part, 302 ... Reference data storage part , 303: comparison unit, 304 ... statistical processing unit, 305 ... output unit

Claims (8)

Performance data acquisition means for acquiring performance data generated based on a performance by a performer;
Comparison means for comparing the performance data acquired by the performance data acquisition means with reference data serving as an example of the performance, and identifying a wrong place of the performance by the performer,
A performance analysis apparatus comprising: statistical processing means for calculating a performance pattern of a performance at an erroneous location specified by the comparison means by statistical processing. The performance analysis apparatus according to claim 1, wherein the statistical processing means identifies a combination of fingers used for a performance identified as an error location as the performance pattern. The performance analysis apparatus according to claim 1 or 2, wherein the statistical processing means identifies a combination of strings used for a performance identified as an error location as the performance pattern. The performance analysis apparatus according to claim 1 or 2, wherein the statistical processing means identifies a combination of positions in a performance identified as an error location as the performance pattern. The said statistical processing means specifies the said performance pattern over the performance before the said performance from the performance specified as the mistake location, The any one of Claims 1-4 characterized by the above-mentioned. Performance analysis device. The said statistical processing means specifies the said performance pattern over the performance after the performance from the performance specified as the mistake location. The one of the Claims 1-5 characterized by the above-mentioned. Performance analysis device. The performance analysis apparatus according to claim 5 or 6, wherein the statistical processing means determines a time range of the performance pattern based on a performance period or the number of pronunciations by performance. The performance analysis apparatus according to claim 7, wherein the statistical processing unit changes the performance period for determining a time range of the performance pattern or the number of pronunciations by the performance according to a predetermined condition.

Images