JP4244537B2 - Music information registration system and music information search system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a music information registration system and a music information search system for registering and searching for a desired music piece from a large number of music information.
[0002]
[Prior art]
Various search methods have been proposed in the past to reduce the time and effort required to find a desired song, such as a song that the user likes or wants to listen to, among the enormous amount of music that is overflowing in the world. There are the following methods.
[0003]
In the first method, for example, a table having “song”, “performer”, “composer”, “lyricist”, etc. as columns is created in a relational database, and information on the music is registered in advance for each music. deep. At the time of search, a desired music is searched by logically combining, for example, an arbitrary character string, a numerical value, etc. with respect to a column arbitrarily selected by the user from the above columns.
[0004]
In the second method, for example, as described in Japanese Patent Application Laid-Open No. 9-212180, a reference melody (melody) for searching is used as MIDI (Musical Instrument Digital Interface) data in a database or the like for each song. Register in advance. In the search, the user plays a part of the melody of the music, and searches the desired music by matching the played audio signal with the audio signal of the reference melody registered in the database.
[0005]
[Problems to be solved by the invention]
However, the search method as described above has the following problems.
[0006]
According to the first method, when the user remembers the title information such as “title”, “performer”, “composer”, etc. of the music to be searched for, although the search can be performed, If you have forgotten the information and cannot remember it, it will be very difficult to find the desired song from the enormous number of songs.
[0007]
In the second method, since it is necessary to play the melody of the music to be searched, the accuracy of the search is easily influenced by the user's ability to express music, and it is difficult to say that it is widely intended for general users. is there. Furthermore, the processing load when collating the voice data of the search condition with the voice data of the reference melody increases, so that the load increases, and a dedicated device or the like for inputting the played voice data is also required. come.
[0008]
Further, even if the first and second methods are combined, the processing becomes complicated due to an increase in the amount of processing at the time of search, and it is difficult to say that it is practical.
[0009]
The present invention has been made in view of such problems, provides a musical information registration system accurately registered without being influenced by the musical expression capability of the user, to search as possible out and music information retrieval system For the purpose.
[0010]
[Means for Solving the Problems]
In order to achieve the above object, the present invention provides a music information registration system and a music information search system described in the following (1) to (4).
(1) For a part or all of the music, an input unit that inputs the length of each sound constituting the music as the length of the push button time;
A conversion unit that converts the pushbutton time input by the input unit into registration data including binary data indicating a long sound and a short sound that are relative lengths;
A music information registration system comprising: a registration unit for registering the registration data converted by the conversion unit in a music database as information on the music.
(2) For a part or all of the desired music, an input unit for inputting the length of each sound constituting the music as the length of the push button;
A conversion unit that converts the push button time input by the input unit into a search key for searching for music composed of binary data indicating a long sound and a short sound that are relative lengths;
A music information search system comprising: a search unit that searches the binary data related to the music that matches the search key converted by the conversion unit from a music database.
(3) For a part or all of the music, an input unit for inputting the length of each sound constituting the music as the length of the push button time, and inputting the silent portion constituting the music as the time when the music is not pushed. ,
A conversion unit that converts the push button time input by the input unit and the non-push button time into registration data composed of ternary data indicating a relative length of long sound, short sound, and silence; ,
A music information registration system comprising: a registration unit for registering the registration data converted by the conversion unit in a music database as information on the music.
(4) For a part or all of the desired music, input the length of each sound constituting the music as the length of the push button time and the silent part constituting the music as the time when the push button is not pushed And
The push button time inputted by the input unit and the time when the push button is not pushed are converted into a search key for music search composed of ternary data indicating a relative length of long sound, short sound, and silence. A conversion unit;
A music information search system, comprising: a search unit that searches the music database for the ternary data related to the music that matches the search key converted by the conversion unit.
[0012]
Here, “music” refers to all music pieces such as vocal music, instrumental music, orchestral music.
[0013]
According to the present invention, the music melody is registered in the database as simple binary data or ternary data, and the search is also performed by simple binary data or ternary data at the time of search. The search can be performed accurately regardless of the music expression ability, and the processing amount during the search can be reduced.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
[0015]
(System configuration)
FIG. 1 is a diagram showing an outline of a music information search system according to this embodiment. This music information search system includes an input unit 1, a conversion unit 2, a registration unit 3, a database 4, and a search unit 5.
[0016]
The input unit 1 inputs a part or all of the music piece by simply indicating the transition of the melody in the time axis direction with a long sound and a short sound. The conversion unit 2 converts the instruction input from the input unit 1 into a registration key composed of binary data of long sound and short sound, or a search key composed of binary data of long sound and short sound. At the time of music information registration, the registration unit 3 registers the registration data converted by the conversion unit 2 in the database 4 as information on the music. At the time of music information search, the search unit 5 searches the database 4 for music that matches the search key converted by the conversion unit 2.
[0017]
Note that music information can be registered in a database as ternary data of long sound, short sound, and silence and can be searched. In this case, the input unit 1 inputs a part or all of the music piece by simply indicating the transition of the melody in the time axis direction with a long sound, a short sound, and silence. The conversion unit 2 converts the instruction input by the input unit 1 into a registration key composed of ternary data of long sound, short sound, and silence, or a search key composed of ternary data of long sound, short sound, and silence. . At the time of music information registration, the registration unit 3 registers the registration data converted by the conversion unit 2 in the database 4 as information on the music. At the time of music information search, the search unit 5 searches the database 4 for music that matches the search key converted by the conversion unit 2.
[0018]
The music information retrieval system can be realized by dedicated hardware having the above components, but all or part of the above components are configured as software and operate on a computer device such as a personal computer. This can also be realized. In FIG. 1, the input unit 1 and the conversion unit 2 are illustrated as being shared between music information registration and music information search. However, the input unit 1 and the conversion unit 2 are prepared for music information registration and music information search, respectively. May be.
[0019]
In the first place, music is a combination of sound intensity, length, pitch, timbre, chord, etc., selected and combined in a certain way. . Stave notation is one of the most widely used methods in the world for visually expressing the sound of music. The staff notation is a method of expressing music by writing notes, rests, speed symbols, time signatures, performance symbols, etc. on the staff. The note represents the pitch of the sound by the position in the staff with five parallel lines, and the length of the sound by the shape. There are various forms to represent the length of this sound, such as all notes, half notes, quarter notes, eighth notes, 16th notes, 32nd notes, 64th notes, etc. Is expressed. There is also a dotted note with a small dot on the note, indicating that it is 1.5 times longer than the original note. Furthermore, there are rests such as all rests, 2-minute rests, 4-minute rests, 8-minute rests, and the like as codes representing portions that do not produce sound.
[0020]
In this music information retrieval system, the features of sound length are simplified and input into binary or ternary values like Morse code so that ordinary users can easily express the melody of music. Like to do.
[0021]
FIG. 2 shows an example in which the music represented by the staff in the figure is converted into binary data of long sound and short sound, and the long sound is represented by “−” and the short sound is represented by “·”.
[0022]
In FIG. 2A, a half note is expressed as a long sound “-”, and a quarter note is expressed as a short sound “.”. In FIG. 2B, a quarter note is expressed as a long sound “-”, and an eighth note. Is expressed as a short sound “・”. Since ordinary people can easily recognize the length of a sound as a relative length, figures (a) and (b) are different in the notation of a staff notation, but are simplified to binary data of long and short sounds. Expressed in the same pattern.
[0023]
FIG. 2C shows an example in which three types of notes are combined. For ordinary people, it is difficult to recognize and express a half note and a dotted quarter note separately. However, as shown in the figure, both are simplified and expressed as a long sound “-” so that they can be easily recognized and expressed. can do.
[0024]
FIG. 2D shows an example including a rest. The rest part tends to be recognized and expressed differently depending on the person. As shown in the figure, some people recognize / represent the quarter note just before the quarter rest as a short “·” (binary data (1)). Some people will recognize / represent as a long sound "-" (binary data (2)). In the example shown in FIG. 2E, the dotted quarter note, the quarter note and the eighth note are recognized / represented as a short “·”, and only the dotted half note is expressed as a long sound “−”. If there is a person who recognizes / represents (binary data (1)), only the eighth note is recognized / represented as a short note “·”, and the dotted quarter note, quarter note, and dotted half note are long notes. Some people will recognize / represent “−” (binary data (2)). In these cases, at the time of music information registration, either one of registration data consisting of binary data (1) or (2) may be registered in the database 4, but both of them may be registered. It may be registered in the database 4.
[0025]
In FIG. 3, the music represented by the staff in the figure is converted into ternary data of long sound, short sound, and silence, and the long sound is represented by “−”, the short sound is represented by “•”, and the silence is represented by “!”. An example is shown. In this example, some people recognize / represent a half note as a long sound “-”, a quarter note as a short sound “•”, and a quarter note as a silence “!” (Ternary data (1)). A person who does not recognize / represent a note, recognizes / represents a half note as a long sound “-”, and a quarter note as a short sound “•” (binary data (1)), or 4 immediately before a quarter rest Some people will recognize / represent half-notes and half-notes as a long note “-”, and other quarter notes as a short-tone “•” (binary data (2)).
[0026]
In such a case, at the time of music information registration, any of registration data consisting of ternary data (1) or binary data (1) or (2) may be registered in the database 4. However, all of these may be registered in the database 4.
[0027]
When registering the registration data in the database 4, all the phrases of the music may be registered, or one or more representative phrases may be picked up and registered.
[0028]
Next, the processing operation of the music information search system according to this embodiment will be described.
[0029]
(Binary input-conversion processing)
FIG. 4 is a flowchart showing an example of a processing operation related to input-conversion of binary music information, and FIG. 7 illustrates a melody input screen 10 of music as a GUI (Graphical User Interface) of the input unit 1. .
[0030]
When the melody of music is indicated by long and short sounds and converted to registration data consisting of binary data of long and short sounds, or a search key for search consisting of binary data of long and short sounds, Then, the threshold of the long sound / short sound is set by the slider 13 on the melody input screen 10 (Step 01). The threshold set here is a threshold value (th1) for expressing a long sound / short sound by pressing and releasing the mouse button.
[0031]
Next, the mouse cursor 12 is placed on the text field 11 of the music melody input screen 10, and the mouse button is pressed (Step 02) or released (Step 03) along the music melody to be registered.
[0032]
The conversion part 2 calculates | requires the time when the button of the mouse was pushed by following Formula.
[0033]
[Expression 1]
The time when the button is pressed = the time when the button is released−the time when the button is pressed, and the obtained time is compared with the threshold value (th1) set in Step 01 to determine whether the sound is short or long (Step 04).
[0034]
If the time during which the mouse button is pressed is below the threshold (th1), the conversion unit 2 determines that the sound is short and generates short sound data (Step 05), and vice versa. It is determined that the sound is long and long sound data is generated (Step 06).
[0035]
The above steps 02 to 06 are repeated until the input of the music melody is completed (Step 07). When the enter button 14 is pressed, the input of the music melody is completed. At this time, the generated long sound and short sound binary data is registration data at the time of music information registration, and is a search key for search at the time of music information search.
[0036]
(Ternary input-conversion processing)
As described above, the processing operation related to the registration has shown an example in which the melody of the music is input-converted by the binary of long sound and short sound, but FIG. 5 shows the melody of the music by the three values of long sound, short sound, and silence. FIG. 8 is a flowchart illustrating an example of input-conversion, and FIG. 8 illustrates a melody melody input screen 20 which is a GUI (Graphical User Interface) of the input unit 1 used for the input-conversion. The silent part is used to express a rest part or a break part of a phrase when expressed as a note.
[0037]
When registering the melody of music in the database 4 as registration data consisting of ternary data, first, the threshold of the long sound / short sound is adjusted by the slider 23 on the melody input screen 20 of the music (Step 11), The silent threshold is adjusted by the slider 24 (Step 12). These thresholds serve as a threshold (th1) for expressing a long sound / short sound by pressing and releasing a mouse button and a threshold (th2) for expressing silence.
[0038]
Next, the mouse cursor 22 is placed on the text field 21 of the music melody input screen 20 and the mouse button is pressed (Step 14) or released along the melody of the music to be registered (Step 17).
[0039]
The conversion part 2 calculates | requires the time when the button of a mouse | mouth is not pressed by following Formula.
[0040]
[Expression 2]
The time when the button is not pressed = the time when the button is pressed−the time when the button is released, and by comparing the obtained time with the threshold value (th2) set in Step12, it is determined whether or not there is silence (Step14). If it is determined, silence data is generated (Step 15).
[0041]
Further, the conversion unit 2 obtains the time during which the mouse button is pressed by the following equation.
[0042]
[Equation 3]
Time when the button is pressed = Time when the button is released-Time when the button is pressed And the obtained time is compared with the threshold value (th1) set in Step 11 (Step 17), and the time when the mouse button is pressed is the threshold value (th1) If it is less than (), it is determined that the sound is short and data of short sound is generated (Step 18), and if the reverse is true, data of long sound is determined and generated (Step 19).
[0043]
The above steps 13 to 19 are repeated until the input of the music melody is completed (Step 20). When the enter button 25 is pressed, the input of the music melody is completed. At this time, the generated ternary data of long sound, short sound, and silence is registration data when registering music information, and is a search key for searching when searching for music information.
[0044]
(registration process)
Next, the operation of processing for registering registration data obtained by input-conversion as described above in the database 4 will be described.
[0045]
FIG. 9B is a diagram illustrating a configuration example of the database 4. In this example, the “ID”, “title”, and “melody” columns are “ID for identifying the song”, “title of the song”, and “registration data representing the long and short notes of the song as binary data”, respectively. The music is registered in the table called “music”. In this example, registration data composed of binary data of long sound and short sound obtained by binary-input-conversion processing is represented by “−” for long sound and “·” for short sound.
[0046]
For example, in a relational database in which the registration unit 3 and the database 4 support SQL, data registration uses an SQL statement “insert”. In the figure, the data values in the "id", "title", and "melody" columns are "2", "△△△△", and "················ The SQL statement when registering the data "...-" is as follows.
[0047]
[Expression 4]
[SQL sentence 1]
insert into music values ('2', '△△△△', '...----------------)
Data can be registered in the database 4 by executing this SQL statement in the registration unit 3.
[0048]
The method for registering registration data consisting of binary data of long tones and short tones has been described above, but when registering registration data consisting of ternary data of long tones, short tones and silences to the database, “ Data for registration consisting of ternary data of long sound, short sound, and silence obtained by ternary input-conversion processing may be registered as data values in the “melody” column.
(Search process)
Next, an operation of searching for information on music pieces registered in the database 4 as described above will be described.
[0049]
FIG. 6 is a flowchart showing an example of the processing operation related to the search.
[0050]
First, in accordance with the procedure shown in the binary input-conversion process, the melody of the music is input and converted to obtain a search key for search consisting of binary data of long and short sounds (Step 31).
[0051]
Next, the search unit 5 compares the search key for search with registration data consisting of binary data of long and short sounds registered in the database 4 (Step 32). Step 33), the matched music information is read from the database 4 and presented to the user (Step 34).
[0052]
The above steps 32 to 34 are performed for all the music pieces registered in the database 4 (Step 35).
[0053]
In the example shown in FIG. 9, a search key for search consisting of binary data of long and short tones shown in FIG. 9 (a), and a long sound of a song whose “id” is “4” shown in FIG. 9 (b). This indicates that a part of registration data composed of binary data of short tones coincides. In this case, information on a song having “id” “4” is obtained as a search result.
[0054]
For example, when the search unit 5 and the database 4 are relational databases corresponding to SQL, an SQL statement “select” is used to perform a search similar to the processing described above. A search key “·············· −” that is composed of binary data of long tones and short tones shown in FIG. The SQL sentence when searching for music from the table “music” shown is as follows.
[0055]
[Equation 5]
[SQL sentence 2]
select * from music where melody like '% ...
By executing this SQL sentence in the search unit 5, information on the music whose “id” is “4” is obtained as a search result from the database 4.
[0056]
The method for searching from the database in which registration data consisting of binary data of long tones and short tones is registered using the search key for search consisting of binary data of long tones and short tones as a condition has been described. It is also possible to search from a database in which registration data consisting of sound and silence ternary data is registered, using a search key for search consisting of ternary data of long sound, short sound and silence as a condition.
As described above, the melody of music is registered in the database as registration data consisting of simple binary data or ternary data, and the search consisting of simple binary data or ternary data at the time of search is also performed. Since the search is performed by the key, the search can be performed accurately without being influenced by the music expression ability of the user, and the processing amount at the time of the search can be reduced.
[0057]
Although the embodiment of the present invention has been described in detail above, the present invention is not limited to the present embodiment, and various improvements and modifications may be made without departing from the spirit of the present invention.
[0058]
For example, in the present embodiment, a button is prepared as an input device, and the melody of the music is expressed as long sound, short sound binary data or long sound, short sound, silent ternary data from the time the button is pressed or released. However, binary data and ternary data can also be obtained by preparing a plurality of buttons and assigning each button to a long sound, a short sound, and silence. Furthermore, voice data such as a nasal song input from a microphone or the like may be converted into binary data or ternary data by signal processing.
[0059]
Further, in the present embodiment, registration data composed of binary data of long and short sounds as the melody of the music, or a text in the text format of “−” as the long sound as the search key for search and “·” as the short sound. Although expressed as a chord, the long sound is represented by “0”, the short sound is represented by “1”, the long sound is represented by “A”, and the short sound is represented by “B”. Or in binary format.
[0060]
In this embodiment, the melody feature of the music is expressed by binary data of long sound and short sound, or ternary data of long sound, short sound, and silence, but the rest length is set in two stages. It can be expressed with four-value data of long sound, short sound, long silence, short silence, or the sound length can be set to 3 levels and expressed with ternary data of long sound, medium sound, short sound, etc. It may be expressed in a range that is easy for the user to instruct.
[0061]
【The invention's effect】
According to the present invention, the melody of music is registered in the database as registration data consisting of simple binary data (or ternary data), and the search consisting of simple binary data (or ternary data) is also performed during the search. Since the search is performed using the key, the search can be performed accurately regardless of the music expression ability of the user. In addition, the melody input unit can be realized with a simple configuration.
[0062]
In addition, since it can be realized by partial matching search between character strings of search keys at the time of search, the processing amount at the time of search can be reduced. Further, since a search algorithm of the same system as a search system using a conventional relational database can be used, the conventional search system can be easily diverted.
[0063]
Therefore, it can be easily used for personal use such as creating a music database by an individual.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of a music information search system in the present embodiment.
FIG. 2 is an image diagram showing an example in which the melodic feature of music is expressed in binary in the music information retrieval system shown in FIG. 1;
FIG. 3 is an image diagram showing an example in which the melody feature of music is expressed by ternary values in the music information retrieval system shown in FIG. 1;
4 is a music information search system shown in FIG. 1, in which a melodic feature of music is input, registration data consisting of binary data of long tones and short sounds, or a search consisting of binary data of long tones and short tones. It is a flowchart which shows the example of a procedure of the process converted into a key.
5 is a music information retrieval system shown in FIG. 1, in which the melodic features of music are input, registration data consisting of ternary data of long sound, short sound, and silence, or ternary values of long sound, short sound, and silence. It is a flowchart which shows the example of a procedure of the process converted into the search key which consists of data.
6 is a flowchart showing a procedure example of a process of searching for a desired music from a database in the music information search system shown in FIG.
7 is an image diagram showing an example of a GUI of an input unit in the music information search system shown in FIG. 1, and is an example of a screen used when expressing the melodic feature of music in binary.
8 is an image diagram showing an example of a GUI of an input unit in the music information search system shown in FIG. 1, and is an example of a screen used when expressing the melody characteristics of music in three values.
FIG. 9 is an image diagram showing an example of a database in the music information search system shown in FIG. 1;
[Explanation of symbols]
1 .... Input unit 2 .... Conversion unit 3 .... Registration unit 4 .... Database 5 .... Search unit 10 .... Melodies melody input screen (binary)
11 .... Text field 12 .... Mouse cursor 13 .... Slider 14 .... Enter button 20 .... Melody input screen (3 values)
21 .... Text field 22 .... Mouse cursor 23, 24 .... Slider 25 .... OK button

Claims (4)

For some or all of the easy songs, an input unit for inputting the length of each sound constituting the music piece, as a length of the push button time,
A conversion unit that converts the pushbutton time input by the input unit into registration data including binary data indicating a long sound and a short sound that are relative lengths ;
Wherein said registration data converted by the conversion unit, the music information registration system characterized by comprising a registration unit for registering the music database as information of the song. An input unit for inputting the length of each sound constituting the music as the length of the push button time for part or all of the desired music;
A conversion unit that converts the push button time input by the input unit into a search key for searching for music composed of binary data indicating a long sound and a short sound that are relative lengths ;
A music information search system comprising: a search unit that searches the binary data related to the music that matches the search key converted by the conversion unit from a music database .   For a part or all of the music, an input unit for inputting the length of each sound constituting the music as the length of the push button time, and inputting the silent part constituting the music as a time when the push button is not pushed,
  A conversion unit that converts the push button time input by the input unit and the non-push button time into registration data composed of ternary data indicating a relative length of long sound, short sound, and silence; ,
  A music information registration system comprising: a registration unit that registers the registration data converted by the conversion unit in a music database as information of the music.   For a part or all of the desired music, an input unit for inputting the length of each sound constituting the music as the length of the push button time, and inputting the silent part constituting the music as a time when the push button is not pushed,
  The push button time inputted by the input unit and the time when the push button is not pushed are converted into a search key for music search composed of ternary data indicating a relative length of long sound, short sound, and silence. A conversion unit;
  A music information search system, comprising: a search unit that searches the music database for the ternary data related to the music that matches the search key converted by the conversion unit.

Images