JP6554826B2 - Music data retrieval apparatus and music data retrieval program - Google Patents

Description

  The present invention obtains performance data representing the performance of a musical instrument, and retrieves music data representing music data (for example, accompaniment style data, music data, etc.) representing music that is compatible with the performance represented by the performance data. The present invention relates to an apparatus and a music data search program.

  2. Description of the Related Art Conventionally, as described in Patent Document 1 below, for example, an electronic musical instrument that stores a plurality of music data (accompaniment style data, music data, etc.) is known. This electronic musical instrument includes a keyboard device. When the user performs using the keyboard device, music data having a rhythm pattern that is the same as or similar to the rhythm pattern of the performance is searched, and the name of the searched music data is displayed.

  For example, as described in Patent Document 2 below, a music search device (computer program) for searching for music (song title) is known. According to this music search apparatus, when the user inputs a melody (single tone series), the music including the input melody is searched from the database, and the name of the searched music is displayed.

International Publication No. 2012/074070 Japanese Patent No. 5552968

  In the electronic musical instrument described in Patent Document 1, a good search result is obtained when a user plays a rhythm part (drum part). Since this electronic musical instrument does not assume that the user plays a performance part such as a melody part, a chord part, or a bass part, for example, the user plays a melody with the right hand in order to search for ballad-like music data. However, when an arpeggio is played with the left hand, music data of a completely different genre may be searched. Therefore, the user needs to play the rhythm part (drum part) in order to obtain a good search result.

  Moreover, in the music search apparatus described in Patent Document 2, the length of each sound (or sounding timing) cannot be specified. That is, if the music includes a portion in which the pitch (pitch) change between the sounds of the melody input by the user is the same (or similar), the lengths of the sounds to be compared (that is, the rhythm of the melody) ) Are completely different, it is determined that the music matches (or resembles) the input melody. Therefore, a music completely different from the music intended by the user may be searched.

  The present invention has been made to address the above-described problems, and an object of the present invention is to provide a music data search device and a music data search program that have improved search accuracy (accuracy) as compared with the prior art. In addition, in the description of each constituent element of the present invention below, in order to facilitate understanding of the present invention, reference numerals of corresponding portions of the embodiment are described in parentheses, but each constituent element of the present invention is The present invention should not be construed as being limited to the configurations of the corresponding portions indicated by the reference numerals of the embodiments.

In order to achieve the above object, the present invention is characterized in that performance data acquisition means (S11) for acquiring performance data including pitch information indicating the pitch of each musical tone and timing information indicating the sounding timing of each musical tone; A music data storage means (14) for storing a plurality of music data each composed of a plurality of performance part data representing performances of a plurality of performance parts, and a performance part included in the performance represented by the performance data; A performance part determination means (S12) , which is a performance part determination means for determining the performance part based on the pitch of each musical tone and the distribution of the sounding timing, and a performance part for each stored music data. The similarity calculation means (S) for calculating the similarity between the performance of the performance part corresponding to the determined performance part and the performance of the determined performance part 4) and output means (S15) for generating and outputting search result data representing a search result of music data similar to the performance data based on the calculated similarity, and a music data search device (10) That is.

In this case, the performance part determining means divides the performance represented by the acquired performance data into a plurality of sections, and determines the performance part based on a ratio of a section where a single sound is generated or a section where a chord is generated. Good.

Further, in this case, the similarity calculation means extracts a sequence of musical sounds of a predetermined performance part from the performance data, and any one of the pitch information and the timing information representing the extracted musical sound sequences. The similarity may be calculated using one or both.

Further, in this case, the music data search apparatus includes a selection unit that selects one performance part among the determined performance parts, and the similarity calculation unit corresponds to the performance part selected by the selection unit. By selecting music data, one of the performance parts constituting the selected music data, the performance part being the same as or corresponding to the performance part selected by the selection means, and the selection means The similarity with the performance of the selected performance part may be calculated. Further, in this case, the similarity calculation means extracts a musical tone sequence of the selected performance part from the performance data, and the pitch information and the timing information representing the extracted musical tone sequence are extracted. The similarity may be calculated using either one or both.

  Generally, when a user tries to search for accompaniment style data, the user often thinks about a characteristic part (performance part) of the music first and plays that part. And the characteristic part changes with music. That is, the performance part performed by the user may be a base part or a melody part. In some cases, the user plays a plurality of performance parts with both hands. Therefore, in the music data retrieval apparatus according to the present invention, the performance part included in the performance represented by the performance data is determined based on the performance data. Next, the similarity between the performance of the determined performance part and the performance of the predetermined performance part corresponding to the determined performance part, which is the performance part of each music data, is calculated, and search is performed according to the similarity. The result is displayed. According to the present invention, the user does not need to perform a special performance in order to obtain a good search result. That is, when playing a part of a piece of music to search for music data, the user can play any performance part without being aware of the search result. Further, as described above, the similarity between the performance of the determined performance part and the performance of the predetermined performance part corresponding to the determined performance part is calculated. Thereby, search accuracy (accuracy) can be improved. For example, if the similarities between performance parts representing the characteristics of music are compared, the search accuracy can be further improved.

The present invention is not limited to implementation as a music data search device, but can also be implemented as a computer program (music data search program) or a music data search method applied to a computer provided in the music data search device.

It is the schematic showing the outline of the performance apparatus which concerns on one Embodiment of this invention. It is a flowchart of an accompaniment style data search program. It is a table | surface which shows the determination conditions used when determining the performance part contained in the performance which performance data represent, the extracted key-on data, and the performance part to be compared. It is the graph which showed the performance data in the piano roll format.

  An electronic musical instrument 10 according to an embodiment of the present invention will be described. First, an outline of the electronic musical instrument 10 will be described. The electronic musical instrument 10 has an automatic accompaniment function. That is, the electronic musical instrument 10 includes a plurality of accompaniment style data. The accompaniment style data represents an accompaniment (phrase) having a predetermined length (for example, a length of two bars). The accompaniment style data is composed of a plurality of performance part data representing performances of a plurality of performance parts. The accompaniment style data includes at least performance parts data of chord parts and bass parts. Each performance part data is composed of a plurality of note data (MIDI data). Each note data includes key-on data indicating the start of sound generation or key-off data indicating the stop of sound generation. The key-on data includes pitch data (key number) representing the pitch of the sound at which sound generation is started. The key-off data includes pitch data (key number) representing the pitch of the sound whose sound generation is stopped. Each note data includes timing data representing the timing of starting sound generation or the timing of stopping sound generation. When the user inputs chords (chords) using the keyboard device while playing the accompaniment pattern (presses a plurality of keys), the pitch of the performance part of the accompaniment style data is changed according to the chords.

  Further, the electronic musical instrument 10 has a search function for accompaniment style data. That is, when a user plays a part of a song (for example, a characteristic part of a song) using the keyboard device, the name of accompaniment style data that is compatible with the performance is displayed. In the following description, the operation mode of the electronic musical instrument 10 when searching for accompaniment style data is referred to as an accompaniment style data search mode. The operation mode of the electronic musical instrument 10 when performing using the keyboard device is referred to as a performance mode.

  Next, a specific configuration of the electronic musical instrument 10 will be described. As shown in FIG. 1, the electronic musical instrument 10 includes an input operator 11, a computer unit 12, a display 13, a storage device 14, an external interface circuit 15, a sound source circuit 16, and a sound system 17, which are bus BUS. Connected through.

  The input operator 11 includes a setting operator that is used for selecting a timbre when the operation mode of the electronic musical instrument 10 is switched. The setting operator includes a switch corresponding to an on / off operation, a volume or rotary encoder corresponding to a rotation operation, a volume or linear encoder corresponding to a slide operation, a mouse, a touch panel, and the like. Further, the input operator 11 includes a performance operator that is used when playing music. The performance operator includes, for example, a keyboard device. The keyboard device is composed of a plurality of keys. Each key is assigned a different key number. For example, the keyboard device is composed of 88 keys (“A0” to “C8”). “21”, “22”, “23”,..., “108” are assigned as key numbers in order from the lowest tone (“A0”) key to the highest tone (“C8”) key. . When the user operates the input operator 11, operation information representing the operation content is supplied to the computer unit 12 described later via the bus BUS. For example, when a key is pressed, key-on data including the key number of the pressed key is supplied to the computer unit 12. When the key is released, key-off data including the key number of the released key is supplied to the computer unit 12.

  The computer unit 12 includes a CPU 12a, a ROM 12b, a RAM 12c, and a timer 12d connected to the bus BUS. For example, when a key of the keyboard device is pressed, the CPU 12a supplies the key number acquired from the keyboard device to the tone generator circuit 16 to start sounding a musical tone corresponding to the key number. Further, for example, when the key of the keyboard device is released, the CPU 12a supplies the key number acquired from the keyboard device to the tone generator circuit 16, and stops the tone corresponding to the key number. Further, for example, the CPU 12a reads and executes an accompaniment style data search program, which will be described later in detail, from the ROM 12b, and searches for accompaniment style data that is compatible with the music played by the user. In addition to the accompaniment style data search program and other programs, the ROM 12b stores initial setting parameters, graphic data for generating display data representing an image displayed on the display 13, character data, and the like. Various data are temporarily stored in the RAM 12c when various programs are executed. The timer 12d supplies an interrupt signal to the CPU 12a at a time interval according to an instruction from the CPU 12a. For example, the timer 12d is used when calculating the timing for generating a metronome click sound. For example, the timer 12d is used when reproducing accompaniment based on accompaniment style data. Specifically, each time the CPU 12a receives an interrupt signal from the timer 12d, the CPU 12a increments a counter that represents the playback progress position of the accompaniment (each piece of performance part data constituting the accompaniment style data). Further, the CPU 12a detects the input code based on the key number of the key being pressed. Then, the CPU 12a selects note data having timing data corresponding to the count value of the counter from the note data constituting the accompaniment style data. The CPU 12a corrects the selected note data (specifically, the pitch of each performance part) according to the detected chord. The CPU 12a supplies key-on data or key-off data constituting the corrected note data to the tone generator circuit 16.

  The display 13 is configured by a liquid crystal display (LCD). The computer unit 12 generates display data representing contents to be displayed using graphic data, character data, and the like, and supplies the display data to the display unit 13. The display device 13 displays an image based on the display data supplied from the computer unit 12. For example, a search result of accompaniment style data is displayed.

  The storage device 14 includes a large-capacity nonvolatile recording medium such as an HDD, a CD, and a DVD, and a drive unit corresponding to each recording medium. The storage device 14 stores a plurality of accompaniment style data, music data, and the like. These accompaniment style data and music data may be stored in the storage device 14 in advance, or may be acquired from an external device via an external interface circuit 15 described later.

  The external interface circuit 15 includes a connection terminal that allows the electronic musical instrument 10 to be connected to an external device such as another electronic musical instrument or a personal computer. The electronic musical instrument 10 can be connected to a communication network such as a LAN (Local Area Network) or the Internet via the external interface circuit 15.

  Connected to the tone generator circuit 16 is a waveform memory WM that stores timbre waveform data representing the acoustic waveforms of a plurality of instrument sounds such as piano, organ, violin, trumpet and the like. The tone generator circuit 16 reads the timbre waveform data designated by the CPU 12a from the waveform memory WM, and uses the read timbre waveform data to generate acoustic waveform data representing the acoustic waveform of the musical tone corresponding to the key-on data (or key-off data). Generated and supplied to the sound system 17. Note that an effector circuit for adding various effects such as a chorus effect and a reverberation effect to the performance sound is also included in the tone generator circuit 16.

  The sound system 17 includes a D / A converter that converts the acoustic waveform data supplied from the sound source circuit 16 into an analog audio signal, an amplifier that amplifies the converted analog audio signal, and the amplified analog audio signal as an acoustic signal. It has a pair of left and right speakers that convert and emit sound.

  Next, a search function for accompaniment style data will be described. When the user switches the operation mode of the electronic musical instrument 10 to the accompaniment style data search mode using the input operator 11, the CPU 12a reads the accompaniment style data search program shown in FIG. 2 from the ROM 12c and executes it. In step S10, the CPU 12a starts accompaniment style data search processing. Next, CPU12a makes a user play a music in step S11. The user first sets the time signature and tempo of the music to be played using the input operator 11. The CPU 12a operates the metronome according to the set tempo. Then, the user instructs to start recording performance data using the input operator 11, and performs using the keyboard device. The user operates the keyboard device with one hand or both hands, and plays any one or more performance parts of the melody part, chord part, and bass part. The CPU 12a supplies the key-on data and key-off data acquired from the keyboard device to the tone generator circuit 16, and records the acquired key-on data and key-off data as performance data in association with their acquisition timing (stored in the RAM 12c or ROM 12b). ) When the user finishes the performance, the CPU 12a stops recording the performance data and calculates the timing corresponding to each bar line. At this time, the CPU 12a may insert bar line data indicating that the calculated timing is a bar line into the performance data.

  Next, in step S12, the CPU 12a determines which performance part (combination of performance parts) the performance part played by the user is based on the recorded performance data and the determination table shown in FIG. To do.

  First, the CPU 12a divides the recorded performance data into a plurality of sections (see FIG. 4). For example, the length Δt of each section is 10 milliseconds. The CPU 12a divides each section into a section where only one musical sound is started (that is, a section where a single sound is generated), and a section where a plurality of musical sounds are started to be generated (that is, a section where chords are generated). Classify into: In FIG. 4, the recorded performance data is displayed in a so-called piano roll format.

  As shown in FIG. 3, for example, the ratio of a single sound section is 90% or more, and the average key number of a musical sound (key-on data) is a middle sound range (for example, “60” to “83”) or a high sound range (for example, , “84” to “108”), the CPU 12a determines that the melody part has been played.

  When the proportion of a single tone section is 90% or more and the average key number of musical tones (key-on data) is included in the low sound range (for example, “59” or less), the CPU 12a determines that the bass part has been played.

  The proportion of a single tone segment is 90% or less, the proportion of a chord segment composed of three or more musical tones is low (for example, 30% or less), and the adjacent musical pitches of the musical tones constituting the chord are When the average of the key number differences is small (for example, “3” or less), the CPU 12a determines that the chord part has been played with one hand. In addition, when the proportion of a single tone segment is 90% or less, the proportion of a chord segment composed of three or more musical sounds is high (for example, greater than 30%), and each chord can be played with one hand, The CPU 12a determines that the chord has been played with one hand. For example, when the difference between the key numbers of the lowest tone and the highest tone among the tones constituting the chord is not more than “12”, the CPU 12a “can play the chord with one hand”. Is determined.

  The proportion of single notes is 90% or less, the proportion of chords composed of three or more musical notes is high (for example, greater than 60%), and the key numbers of the musical notes that make up the chords cannot be played with one hand When the chord is biased toward the low range, the CPU 12a determines that the chord part has been played with both hands. For example, when the difference between the key numbers of the musical tone having the lowest pitch and the musical tone having the highest pitch among musical tones constituting the chord is equal to or greater than a predetermined value (for example, “15”), the CPU 12a The key numbers of the musical sounds to be played are distributed in a range that cannot be played with one hand. If the key number difference is “12”, it can be considered that performance (key pressing) can be performed with one hand. When the key number difference exceeds “12”, it is difficult to perform with one hand. In particular, when the key number difference exceeds “17”, it is almost impossible to perform with one hand. Impossible. Therefore, it is appropriate to set the key number difference to any one of “13” to “17”. Further, the CPU 12a determines that “the chord is biased toward the low frequency range” when the average of the key numbers of the musical tones is “60 (middle center)” or less, for example.

  The proportion of single notes is 90% or less, the proportion of chords composed of three or more musical notes is high (for example, greater than 60%), and the key numbers of the musical notes that make up the chords cannot be played with one hand When the chord is biased toward the high range, the CPU 12a determines that the chord part has been played with both hands. For example, the CPU 12a determines that “the chord is biased toward the high frequency range” when the average of the key numbers of the musical sounds is greater than “72”.

  The proportion of a single note is 90% or less, the proportion of a chord composed of three or more musical sounds is low (for example, 30% or less), and the difference between the key numbers of the highest and lowest chords is predetermined. CPU12a determines that the bass part is played with the left hand and the melody part is played with the right hand.

  The proportion of a single tone segment is 90% or less, and the proportion of a chord segment composed of three or more musical tones is medium (for example, greater than 30% and less than 60%). When the average of the key numbers is smaller than the average of the key numbers of the musical tones in the chord section, the CPU 12a determines that the bass part is played with the left hand and the chord part is played with the right hand.

  The proportion of a single tone segment is 90% or less, and the proportion of a chord segment composed of three or more musical tones is medium (for example, greater than 30% and less than 60%). When the average of the key numbers is equal to or greater than the average of the key numbers of the musical tones in the chord section, the CPU 12a determines that the chord part is played with the left hand and the melody part is played with the right hand.

  Next, in step S13, the CPU 12a analyzes the recorded performance data, for example, in the same procedure as described in JP-A-6-295179, and selects one of the key-on data included in each section. Alternatively, “LB”, “LT”, “RB”, and “RT” are added as identifiers to a plurality of data (see FIG. 4). Note that “LB” is an identifier added to the key-on data of the lowest tone among the musical tones played with the left hand. “LT” is an identifier added to the key-on data of the highest tone among the musical sounds played with the left hand. “RB” is an identifier added to the key-on data of the lowest tone among the musical tones played with the right hand. “RT” is an identifier added to the key-on data of the highest tone among the musical tones played with the right hand.

  Specifically, the CPU 12a analyzes the recorded performance data for each measure and adds each identifier to the key-on data. First, the CPU 12a selects a measure to be processed. Thereafter, the CPU 12a selects and analyzes the processing target sections one by one in order from the first section to the last section of the processing target measure. The CPU 12a refers to the identifier added to the key-on data of the section before the processing target section (one previous section or several previous sections), and adds the identifier to the key-on data of the processing target section. . That is, in consideration of the movement of the user's hand, key-on data relating to a key pressed by the user with the right hand and the left hand in the processing target section is detected. This makes it possible to extract a note string (a series of key-on data) that constitutes a bass, melody, or chord. Note that there is no previous section in the first section. Therefore, for example, the key-on data of the highest note in the measure to be processed is temporarily generated as the key-on data of the previous section, and “RT” is added to the generated key-on data. In addition, the key-on data of the lowest sound in the measure to be processed is temporarily generated as key-on data of the previous section, and “LB” is added to the generated key-on data. Further, based on the key numbers of the two generated key-on data, key-on data to which “RB” and “LT” are added is temporarily generated. For example, the two key numbers are interpolated (interpolated) using the key numbers of the two generated key-on data. Then, key-on data having a key number obtained by the interpolation calculation is temporarily generated, and “RB” and “LT” are added to the generated key-on data.

  For example, when the section to be processed is a single-tone section, the CPU 12a obtains key-on data with “RT” or “LB” among the key-on data of the previous section (or several previous sections). select. Next, the CPU 12a selects, from the selected key-on data, key-on data whose key number is close to the key number of the key-on data in the section to be processed. Then, the CPU 12a adds the same identifier as the identifier attached to the selected key-on data to the key-on data of the section to be processed.

  Further, for example, when the section to be processed is a chord section, if the pitch difference between the highest sound and the lowest sound is 1 octave or more, it is highly likely that the user has played with both hands in this section. Therefore, the CPU 12a adds “RT” to the key-on data of the highest sound and adds “LB” to the key-on data of the lowest sound. When other key-on data exists between the highest sound and the lowest sound, the CPU 12a selects the key-on data having the smallest key number among the other key-on data. If the key number of the selected key-on data is closer to the key number of the highest note than the key number of the lowest note, the CPU 12a adds “RB” to the selected key-on data. In this case, the CPU 12a adds “LT” to the lowest sound.

  Further, for example, when the section to be processed is a chord section, if the pitch difference between the highest sound and the lowest sound is smaller than one octave, the user is highly likely to perform with one hand in this section. In this case, the CPU 12a calculates a base reliability BR and a melody reliability MR described below.

The base reliability BR is calculated as follows. The CPU 12a selects a section including the key-on data with “RB” and “LB”, which is the section nearest to the section to be processed. Then, the key number KN _{RB of the} key-on data with “RB” and the key number KN _{LB of the} key-on data with “LB”, the lowest key number KN _CB and the highest key number KN _CT of the section to be processed. It applies to the following arithmetic expression (1). This calculation result is the base reliability BR.
BR = (KN _RB −KN _CB ) + (KN _CT −KN _CB ) −2 × (KN _CB −KN _LB ) (1)

The melody reliability MR is calculated as follows. The CPU 12a selects a section including the key-on data with “RT” and “LT”, which is the section nearest to the section to be processed. Then, the key number KN _{RT of the} key-on data with “RT” and the key number KN _{LT of the} key-on data with “LT”, and the lowest key number KN _CB and the highest key number KN _CT of the section to be processed. It applies to the following arithmetic expression (2). This calculation result is the melody reliability MR.
MR = (KN _CT −KN _LT ) + (KN _CT −KN _CB ) −2 × (KN _RT −KN _CT ) (2)

  When the base reliability BR is equal to or lower than “0” and the melody reliability MR is higher than “0”, the user is highly likely to perform with the right hand in the processing target section. Therefore, the CPU 12a adds “RT” to the highest key-on data of the section to be processed, and adds “RB” to the lowest-key key-on data. On the other hand, when the base reliability BR is larger than “0” and the melody reliability MR is “0” or less, the user is highly likely to perform with the left hand in the processing target section. Therefore, the CPU 12a adds “LT” to the key-on data of the highest sound in the section to be processed and adds “LB” to the key-on data of the lowest sound.

  Next, in step S14, the CPU 12a extracts a predetermined key-on data series from the recorded performance data in accordance with the determination result in step S12 described above. Then, the degree of similarity between the extracted key-on data series and the key-on data series constituting predetermined performance part data among the accompaniment style data is calculated. The similarity calculation procedure will be described later.

  For example, if it is determined in step S12 that the melody part has been played, the CPU 12a extracts key-on data with “RT” from the recorded performance data. In addition, for example, in the above-described step S12, when it is determined that the chord part is played with one hand, and when it is determined that the chord part is played with both hands, and the chord is biased toward the low range, The CPU 12a extracts key-on data with “RT”. In these cases, the CPU 12a calculates the similarity between the extracted key-on data series and the key-on data series of the chord part of each accompaniment style data.

  For example, when it is determined in step S12 that the bass part has been played, the CPU 12a extracts key-on data with “LB” from the recorded performance data. Also, for example, if it is determined in step S12 that the chord part is played with both hands and the chord is biased in the high range, the bass part is played with the left hand and the melody part is played with the right hand. Also when it determines, CPU12a extracts the key-on data to which "LB" was attached | subjected. Further, even when none of the determination conditions is satisfied, the CPU 12a extracts key-on data with “LB”. In these cases, the CPU 12a calculates the similarity between the extracted key-on data series and the key-on data series of the base part of each accompaniment style data.

  For example, if it is determined in step S12 that the chord part is played with the left hand and the melody part is played with the right hand, the CPU 12a extracts key-on data with “LB”. In this case, the CPU 12a calculates the similarity between the extracted key-on data series and the key-on data series of the chord part of each accompaniment style data.

  The similarity is calculated as follows. First, the CPU 12a uses a method similar to the method described in Patent Document 1 to determine the rhythm similarity between the extracted key-on data sequence and the comparison-target key-on data sequence at a timing corresponding to each key-on data. Calculate using the data. Then, the CPU 12a uses a method similar to the method described in Patent Document 2 above, and similarities between the key number change mode in the extracted key-on data sequence and the key number change mode in the key-on data sequence to be compared are used. Calculate the degree.

  Next, in step S15, the CPU 12a displays a list in which names of the accompaniment style data are arranged in descending order of the calculated similarity on the display unit 13 as a search result. And CPU12a complete | finishes an accompaniment style data search process in step S16.

  As described above, in the electronic musical instrument 10, it is determined which performance part has been played based on the key number of each musical tone constituting the performance data and the distribution of sound generation timing. Next, a sequence of key-on data corresponding to the determination result is extracted from the performance data. Then, the similarity between the extracted key-on data series and a predetermined performance part of each accompaniment style data is calculated, and the names of the accompaniment style data are displayed in descending order of the similarity. Thus, according to the electronic musical instrument 10, the user does not need to perform a special performance in order to obtain a favorable search result. That is, when playing a part of a piece of music to search for accompaniment style data, the user can play any performance part without being aware of the search result. Also, according to the table shown in FIG. 3, a performance part key-on data series suitable for calculating the similarity is extracted from the performance data, and the performance part to be compared is selected, so that the search accuracy (accuracy) is improved. Can be improved. For example, the performance of the bass part often includes a performance pattern specific to the music genre (ballad, rock, jazz, bossa nova, etc.). Therefore, by extracting a base part key-on data sequence from the performance data, the search accuracy (accuracy) can be improved.

  Furthermore, in carrying out the present invention, the present invention is not limited to the above embodiment, and various modifications can be made without departing from the object of the present invention.

  In the above-described embodiment, accompaniment style data that is compatible with performance data performed by the user using the keyboard device is retrieved. Instead, performance data recorded in advance may be read from the ROM 12b, the storage device 14 or the like, and accompaniment style data that is compatible with the read performance data may be searched.

  Further, instead of the accompaniment style data, music data including a plurality of performance parts may be searched. The accompaniment style data and music data (music data) may be audio data. That is, these music data may be composed of a plurality of sample values obtained by sampling performances (musical sounds) of a plurality of performance parts at a predetermined sampling period. In this case, the pitch and tone generation timing of each musical tone of each performance part may be extracted based on the audio data using a conventional method (for example, the method described in Japanese Patent Application Laid-Open No. 2009-217260).

DESCRIPTION OF SYMBOLS 10 ... Electronic musical instrument, 11 ... Input operation element, 12 ... Computer part, 13 ... Display, 14 ... Memory | storage device, 15 ... External interface circuit, 16 ... Sound source circuit , 17 ... Sound system, BR ... Base reliability, BUS ... Bus, MR ... Melody reliability, WM ... Waveform memory

Claims (9)

Performance data acquisition means for acquiring performance data including pitch information indicating the pitch of each musical sound and timing information indicating the sound generation timing of each musical sound;
Music data storage means for storing a plurality of music data each composed of a plurality of performance part data respectively representing performances of a plurality of performance parts;
Performance part determination means for determining a performance part included in the performance represented by the performance data, wherein the performance part determination means determines the performance part based on a pitch and a sounding timing distribution of each musical sound;
A similarity calculation means for calculating a similarity between the performance part corresponding to the determined performance part and the performance of the determined performance part, which is the performance part of each stored music data;
A music data search device comprising: output means for generating and outputting search result data representing a search result of music data similar to the performance data based on the calculated similarity. The music data search device according to claim 1,
The performance part determination means divides a performance represented by the acquired performance data into a plurality of sections, and determines the performance part based on a ratio of a section where a single sound is generated or a section where a chord is generated. Data retrieval device. In the music data search device according to claim 1 or 2,
The similarity calculation means extracts a musical tone sequence of a predetermined performance part from the performance data, and uses either one or both of the pitch information and the timing information representing the extracted musical tone sequence. And a music data search device for calculating the similarity.   The music data search device according to any one of claims 1 to 3,
  Selecting means for selecting one of the determined performance parts;
  The similarity calculation means selects music data corresponding to the performance part selected by the selection means, and is one performance part of the performance parts constituting the selected music data, and the selection means A music data retrieval device for calculating a similarity between a performance of a performance part that is the same as or corresponding to a selected performance part and a performance of a performance part selected by the selection means.   The music data search device according to claim 4, wherein
  The similarity calculation means extracts a musical tone sequence of the selected performance part from the performance data, and either or both of the pitch information and the timing information representing the extracted musical tone sequence A music data search device that calculates the similarity using Computer
Performance data acquisition means for acquiring performance data including pitch information indicating the pitch of each musical sound and timing information indicating the sound generation timing of each musical sound;
Music data storage means for storing a plurality of music data each composed of a plurality of performance part data respectively representing performances of a plurality of performance parts;
Performance part determination means for determining a performance part included in the performance represented by the performance data, wherein the performance part determination means determines the performance part based on a pitch and a sounding timing distribution of each musical sound;
A similarity calculation means for calculating a similarity between the performance part corresponding to the determined performance part and the performance of the determined performance part, which is the performance part of each stored music data;
A computer program that functions as a music data search device, comprising: output means for generating and outputting search result data representing a search result of music data similar to the performance data based on the calculated similarity.   The computer program according to claim 6,
  The computer,
  Selecting means for selecting one of the determined performance parts;
  The similarity calculation means selects music data corresponding to the performance part selected by the selection means, and is one performance part of the performance parts constituting the selected music data, and the selection means A computer program functioning as a music data search device for calculating a similarity between a performance of a performance part that is the same as or corresponding to a selected performance part and a performance of a performance part selected by the selection means.   A performance data acquisition step for acquiring performance data including pitch information representing the pitch of each musical sound and timing information representing the sound generation timing of each musical sound;
  A performance part determining step for determining a performance part included in the performance represented by the performance data, the performance part determining step for determining the performance part based on a pitch of each musical sound and a distribution of sound generation timing;
  Of the plurality of music data stored in the predetermined music data storage means, the determination among the performance parts of the plurality of music data each composed of a plurality of performance part data respectively representing performances of the plurality of performance parts A similarity calculation step for calculating a similarity between the performance of the performance part corresponding to the performance part and the performance of the determined performance part;
  An output step of generating and outputting search result data representing a search result of music data similar to the performance data based on the calculated similarity.   The music data search method according to claim 8, wherein
  A selection step of selecting one performance part of the determined performance parts;
  The similarity calculation step selects music data corresponding to the performance part selected in the selection step, and is one performance part of the performance parts constituting the selected music data, wherein the selection step A method for retrieving music data, comprising: calculating a similarity between the performance of a performance part that is the same as or corresponding to the performance part selected in step 1 and the performance of the performance part selected in the selection step.

Images