JP6074835B2 - Music practice support device - Google Patents

Description

  The present invention relates to a music practice support device.

  Some karaoke apparatuses have a function to support singing practice. For example, Patent Document 1 stores a point where both signals are shifted by comparing a singing signal obtained when a singer sings along with an accompaniment sound reproduced in karaoke and a sample vocal signal, An apparatus for reproducing the stored location has been proposed. The singer can practice the weak part by singing the reproduced part.

JP 2001-34278 A

In the technique described in Patent Document 1, it is effective for improving the music because it is possible to repeatedly practice the place where the pitch shift has occurred in the music once sung, but not necessarily improving the singing of other music. I can not connect it. The technique described in Patent Document 1 is applied to practice support for playing musical instruments as described above.
The present invention has been made in view of the above-described background, and an object thereof is to improve the singing ability (or performance ability) that can be applied to other musical compositions.

In order to solve the above-described problem, the present invention provides a sound data acquisition unit that acquires sound data representing a sound, a feature specification unit that specifies characteristics of sound data acquired by the sound data acquisition unit, and the feature specification Based on the characteristics specified by the unit, the evaluation unit that evaluates the sound data for each predetermined evaluation item, and one or more evaluation items that satisfy the predetermined evaluation result by the evaluation unit are specified. An evaluation item specifying unit; and a generation unit that generates music data for supporting the practice of singing or playing the sound data in a generation mode according to the evaluation item specified by the evaluation item specifying unit ; The evaluation item specifying unit acquires instruction history data indicating a generation history of music data in the generation unit, and determines and determines a weighting coefficient for each evaluation item according to the acquired instruction history data. Providing music training support apparatus which is characterized in that said specified using weighting factors. In the present embodiment, the sound data includes singing voice.

  In a preferred aspect of the present invention, the generation unit is specified by the evaluation item specifying unit from a storage unit that stores a pitch template representing a temporal change in pitch and a rhythm template representing a timing sequence for each evaluation item. A rhythm template and a pitch template corresponding to the evaluation item are read from the storage unit, and a note row according to a note generation timing row indicated by the read rhythm template, wherein the pitch of each note is the read pitch. You may generate | occur | produce the music data showing the music comprised by the row | line | column of the note which is a pitch according to the time change of the pitch shown by a template.

Further, the present invention provides a sound data acquisition unit that acquires sound data representing a sound, a feature specification unit that specifies a feature of the sound data acquired by the sound data acquisition unit, and a feature specified by the feature specification unit. An evaluation unit for evaluating the sound data for each predetermined evaluation item, an evaluation item specifying unit for specifying one or a plurality of the evaluation items for which the evaluation result by the evaluation unit satisfies a predetermined condition, A generation unit that generates song data for supporting the practice of singing or playing the sound data in a generation mode according to the evaluation item specified by the evaluation item specifying unit , From a storage unit that stores a pitch template that represents a temporal change and a rhythm template that represents a timing sequence for each evaluation item, a resource corresponding to the evaluation item specified by the evaluation item specifying unit. A template of notes and a pitch template read from the storage unit, and a sequence of notes according to a sequence of sound generation timings of notes indicated by the read rhythm template, wherein the pitch of each note is equal to the pitch indicated by the read pitch template. A piece of music data representing a piece of music composed of a series of notes having a pitch according to a temporal change is generated, and at the time of the reading, a plurality of rhythm templates and pitch templates having different degrees of difficulty are provided for each evaluation item. A rhythm template and a pitch template corresponding to the evaluation item specified by the evaluation item specifying unit, the rhythm template and the pitch template having a difficulty corresponding to the evaluation result of the evaluation unit are read from the storage unit A music practice support device characterized by That.

Further, the present invention provides a sound data acquisition unit that acquires sound data representing a sound, a feature specification unit that specifies a feature of the sound data acquired by the sound data acquisition unit, and a feature specified by the feature specification unit. An evaluation unit for evaluating the sound data for each predetermined evaluation item, an evaluation item specifying unit for specifying one or a plurality of the evaluation items for which the evaluation result by the evaluation unit satisfies a predetermined condition, A song selection unit that selects song data corresponding to the evaluation item specified by the evaluation item specifying unit from among the plurality of song data stored in the storage unit in association with the evaluation item, and the evaluation item The specifying unit acquires instruction history data indicating a selection history of music data in the song selection unit, determines a weighting coefficient for each evaluation item according to the acquired instruction history data, and determines the weighting determined Providing music training support apparatus which is characterized in that said identified using the coefficients.

Moreover, in a preferable aspect of the present invention, the generation unit acquires instruction history data indicating a generation history of music data in the generation unit, and stores a plurality of pitch templates for different evaluation ranges for each evaluation item. A pitch template corresponding to the evaluation item specified by the evaluation item specifying unit and having a sound range corresponding to the acquired instruction history data may be read from the storage unit.
In a preferred embodiment of the present invention, a singing voice synthesizing unit that synthesizes a typical singing voice and a sound represented by the music data and reproduces the synthesized sound may be provided.

  According to the present invention, it is possible to improve the singing ability (or performance ability) that can be applied to other musical compositions.

System configuration diagram of an embodiment of the present invention The figure for demonstrating an example of the evaluation content by the singing pitch evaluation engine 12 The figure for demonstrating an example of the evaluation content by the pitch and pitch jump evaluation engine 13 The figure for demonstrating an example of the evaluation content by the timing evaluation engine 15 The figure for demonstrating an example of the evaluation content by the rhythm evaluation engine 16 The figure for demonstrating an example of the content of evaluation by the singing technique evaluation engine 17 The figure for demonstrating an example of the evaluation content by the expressive power evaluation engine 18 The figure for demonstrating an example of the processing content of the singing instruction selection engine 14 The figure which shows an example of the calculation method of a priority The block diagram which shows an example of a structure of the singing instruction composition engine 30 Diagram showing an example of a template A diagram showing an example of the contents of rhythm data Figure showing an example of melody data A diagram showing an example of the contents of rhythm data Figure showing an example of melody data The figure which shows an example of the hardware constitutions of a karaoke apparatus Flow chart showing the processing flow of this embodiment Flow chart showing the processing flow of this embodiment System configuration diagram in a modification of the present invention

<Embodiment>
<Configuration>
FIG. 1 is a block diagram showing an example of the configuration of a singing practice support system 100 according to an embodiment of the present invention. The singing practice support system 100 includes a singing instruction engine 10, a storage unit 20, a singing instruction composition engine 30, a network 40, a server 50, a sequencer 60, a microphone 80, and a speaker 81. The singing instruction composition engine 30 and the server 50 are connected by a network 40. The singing instruction engine 10 compares the singing voice of the singer collected by the microphone 80 with the singing pitch to be sung by the singer, and performs evaluation according to the comparison result. The storage unit 20 is a storage unit for storing various data, and is, for example, an HDD (Hard Disk Drive) or a nonvolatile memory. The server 50 has an instruction history data storage area 51 for storing instruction history data representing instruction history of singing so far.

  The singing instruction composition engine 30 dynamically generates data (hereinafter referred to as “singing instruction music data”) indicating the music for singing practice (hereinafter referred to as “singing instruction music”) according to the evaluation result of the singing instruction engine 10. In this embodiment, the singing instruction engine 10 evaluates the specified singer's pitch with respect to three items of pitch / pitch jumping, rhythm, and expressiveness, and the singing instruction composing engine 30 is suitable for practice of each item. Generate teaching music data. The song teaching music data is, for example, data in the MIDI (Musical Instrument Digital Interface) format. The singing instruction composition engine 30 outputs the generated singing instruction composition data to the sequencer 60. The network 40 is a communication network such as a LAN or the Internet. The server 50 has an instruction history data storage area 51 for accumulating a generation history of song instruction song data by the song instruction composition engine 30. The server 50 is connected to the singing instruction / composition engine 30 via the network 40. The sequencer 60 reproduces the singing instruction song data supplied from the singing instruction composition engine 30.

  The storage unit 20 includes a music information data storage area 21, a composition template data storage area 22, and an accompaniment data storage area 23. The music information data storage area 21 stores a plurality of music data records composed of a plurality of items such as “song number”, “song name”, “singer name”, “genre”, “accompaniment data”, and “GM data”. ing. “Song number” is a number for uniquely identifying a music piece, and is composed of, for example, a 4-digit parent number and a 2-digit branch number. “Song name” represents the name of each music piece. “Singer name” represents the name of each singer. “Genre” represents the genre of music to which each musical piece belongs among a plurality of genres classified according to a predetermined classification standard. “Accompaniment data” represents accompaniment data of each music piece. “GM data” represents guide melody data (hereinafter referred to as “GM data”), which is data indicating the melody of the vocal part of each musical piece, that is, data specifying the content of the constituent sound to be sung. The GM data is used as a reference for comparison when the singing pitch evaluation engine 12 evaluates the skill of singing by a singer. The GM data is described in the MIDI format, for example. The accompaniment data and the GM data are not limited to being stored in the music information data storage area 21, and an address indicating the storage location of the accompaniment data and GM data is stored in the music information data storage area. There may be.

  The composition template data storage area 22 stores a plurality of rhythm and melody templates used when the singing instruction composition engine 30 generates singing instruction data. In the composition template data storage area 22, a rhythm template representing a sequence of note sounding timings is stored for each evaluation item, and a pitch template representing a temporal change in pitch is stored for each evaluation item. . In the following description, for convenience of explanation, when there is no need to distinguish between a rhythm template and a pitch template, these will be referred to as “templates”.

  In the accompaniment data storage area 23, the accompaniment data representing the accompaniment sound used when the singing instruction composition engine 30 generates the singing instruction composition data is associated with the template stored in the template data storage area 22 for composition. It is remembered. The accompaniment data and the template may be associated one by one, or may be associated by a plurality of pairs.

  Next, the configuration of the singing instruction engine 10 will be described. The singing pitch specifying engine 11 acquires sound data representing the waveform of the sound collected by the microphone 80, and specifies the characteristics of the sound indicated by the acquired sound data. The singing pitch identification engine 11 is an example of a sound data acquisition unit according to the present invention and an example of a feature identification unit according to the present invention. The singing pitch identification engine 11 analyzes data representing the singer's voice collected by the microphone 80 (hereinafter referred to as “singing voice data”), and identifies the pitch of the sound indicated by the singing voice data. The singing pitch identification engine 11 outputs data representing the identified pitch (hereinafter referred to as “pitch data”) to the singing pitch evaluation engine 12, the timing evaluation engine 15, and the singing technique evaluation engine 17. An example of the pitch data is shown by a solid line P1 in FIG. 2, and the pitch at each time is shown along the time axis as shown.

  The singing pitch evaluation engine 12 and the pitch / pitch jump evaluation engine 13 evaluate the sound indicated by the singing voice data with respect to the evaluation items of “pitch” and “pitch jump” based on the pitch specified by the singing pitch specification engine 11. To do. Pitch jumping means that the pitch changes greatly between two adjacent notes. In this embodiment, first, the singing pitch evaluation engine 12 compares the singing pitch specified by the singing pitch specifying engine 11 with the pitch indicated by the GM data stored in the music information data storage area 21 for each note. The pitch is evaluated, and an evaluation value representing the evaluation result is recorded in a predetermined storage area of the storage unit 20.

  Here, an example of evaluation contents by the singing pitch evaluation engine 12 will be described with reference to FIG. The solid line P1 represents the singing pitch specified by the singing pitch specifying engine 11, and the note N1, note N2, note N3,... Represent the pitch represented by the GM data (that is, the singing pitch that the singer should sing). The singing pitch evaluation engine 12 compares the singing pitch specified by the singing pitch specifying engine 11 with the pitch of the GM data, and performs evaluation according to the difference between the two. The evaluation value representing the evaluation result is, for example, 100 points (that is, no deduction points) if the pitch difference between the two is within a predetermined allowable range in a certain note, and the pitch difference between the two is within the above range. If the period of the portion that does not fit in the GM data is half of the note length in this note, it may be 50 points.

  In the example illustrated in FIG. 2, the singing pitch evaluation engine 12 performs evaluation so that the higher the difference between the singing pitch specified by the singing pitch specifying engine 11 and the pitch of the GM data, the lower the evaluation. Specifically, the note N1 is evaluated as “90 points” because the difference is small, the note N2 is evaluated as “20 points” because the difference is large, and the difference is not so large for the note N3. An evaluation of “70 points” has been made.

  The pitch / pitch jump evaluation engine 13 records the pitch score result of the singing pitch evaluation engine 12 for each note and the score result regarding the pitch jump from one note to the next note as the evaluation value of the pitch jump. The pitch / pitch jump evaluation engine 13 totals the results after the performance of the karaoke music piece, and uses the score score and the evaluation value at the point where the pitch jump score is equal to or less than the threshold value to the singing instruction selection engine 14 as the evaluation value of the pitch jump evaluation item. Notice.

  FIG. 3 is a diagram for explaining an example of evaluation contents by the pitch / pitch jump evaluation engine 13. In the example shown in FIG. 3, the pitch / pitch jump evaluation engine 13 counts the scoring results in units of notes for each scale (see (a) of FIG. 3), and sings a musical scale whose sum is less than the threshold value. Notify the selection engine 14. In the example shown in FIG. 3A, it is shown that the note of scale A # 3 appears twice, the note of scale D4 appears three times, and the note of scale D # 4 appears four times in the music. Yes. In the “Total” column, an average value of scoring results for each scale is recorded as an evaluation value. In addition, the pitch / pitch jump evaluation engine 13 counts the scoring results of two consecutive notes for each pitch jump amount (see FIG. 3B), and singing guidance selection is selected for pitch jumps where the count result is equal to or less than a threshold value. The engine 14 is notified. In the example shown in FIG. 3 (b), the music piece includes continuous notes that jump from the scale A # 3 to the scale D # 4, for example, notes having a relationship of “note N1 to note N2” shown in FIG. This shows a case in which there are two sets of notes, and there are three sets of consecutive notes jumping from the scale D # 4 to the scale D4, for example, the notes having the relationship of “note N2 to note N3” shown in FIG. . In the “Total” column, an average value of scoring results for each jump amount is recorded as an evaluation value.

  The timing evaluation engine 15 and the rhythm evaluation engine 16 convert the sound indicated by the singing voice data based on the pitch specified by the singing pitch specifying engine 11 into “table beat”, “back beat”, “triplet”, “ Evaluation is performed for each evaluation item related to the rhythm, such as “sixteenth note”. In this embodiment, first, the timing evaluation engine 15 compares the singing pitch specified by the singing pitch specifying engine 11 with the beat timing indicated by the GM data (that is, the beat timing that the singer should sing) for each note. Evaluate the timing and score.

  FIG. 4 is a diagram illustrating an example of evaluation contents by the timing evaluation engine 15. In FIG. 4, the horizontal axis indicates time, and the vertical axis indicates pitch. In the example shown in FIG. 4, since the singing timing of the note N3 seems to be running (the timing at which the solid line P1 reaches the scale of the note N3 is slightly earlier than the timing of the note N3), the rhythm score of the note N3 ( Evaluation value) is low.

  The rhythm evaluation engine 16 classifies the scoring results from the timing evaluation engine 15 into evaluation items such as table beats, back beats, triplets, and sixteenth notes in a measure, and records them as evaluation values. At this time, when the target note corresponds to both a table beat and a sixteenth note, the rhythm evaluation engine 16 stores the note in both. The rhythm evaluation engine 16 totals the results after the music performance, and notifies the singing instruction selection engine 14 of the evaluation items whose evaluation values are equal to or less than the threshold value.

  FIG. 5 is a diagram for explaining an example of evaluation contents by the rhythm evaluation engine 16. As shown in FIG. 5, the rhythm evaluation engine 16 uses the scoring result in units of notes as rhythm types (evaluation items) such as “table beat”, “back beat”, “triplet”, and “16th note”. Aggregate every time. In the example shown in FIG. 5, note N1, note N2, and note N4 are recorded as table beats, and note N3 is recorded as a back beat.

  The singing technique evaluation engine 17 and the expressive power evaluation engine 18 evaluate the sound indicated by the singing voice data for each singing technique based on the pitch specified by the singing pitch specifying engine 11. In this embodiment, the singing technique evaluation engine 17 specifies a singing technique by changing the singing pitch, evaluates the singing technique for each note, and records an evaluation value representing an evaluation result. In this embodiment, the singing technique evaluation engine 17 identifies the notes in which the singing techniques “vibrato”, “shrimp”, “fist”, and “fall” are used. For “Vibrato”, the singing technique evaluation engine 17 analyzes the pattern of temporal change of the singing pitch, and identifies the section where the pitch continuously fluctuates within a predetermined range above and below the central pitch. Then, a note including the specified section is a note in which “vibrato” is used. For “Vibrato”, the singing technique evaluation engine 17 evaluates and scores the uniformity and depth of pitch fluctuation and the duration.

  For “shakuri”, the singing technique evaluation engine 17 analyzed the pattern of temporal change of the singing pitch, and identified and identified the section where the pitch continuously changed from the low pitch to the target pitch at the beginning of the phrase. A note including a section is assumed to be a note in which “shakuri” is used. For “shakuri”, the singing technique evaluation engine 17 evaluates and scores the pitch fluctuation range and the fluctuation time.

  For “Fall”, the singing technique evaluation engine 17 analyzes the pattern of temporal change of the singing pitch, and the interval in which the pitch continuously changes from a long tone that extends long at a constant pitch to a low pitch at the end of the phrase. And a note including the specified section is defined as a note in which “fall” is used. For “fall”, the singing technique evaluation engine 17 evaluates and scores the pitch fluctuation range and the fluctuation time.

  For “fist”, the singing technique evaluation engine 17 analyzes the temporal change pattern of the singing pitch, identifies the section where the pitch goes up and down in a short time in the phrase, and “notes the fist” including the identified section. "Is used. For “fist”, the singing technique evaluation engine 17 evaluates and scores the pitch fluctuation range and the fluctuation time.

  FIG. 6 is a diagram for explaining an example of processing contents of the singing technique evaluation engine 17. In the example shown in FIG. 6, scoring is performed with note N1 and vibrato is specified with note N3, and the singing technique is scored.

  The expressiveness evaluation engine 18 classifies the evaluation value of the singing technique acquired by the singing technique evaluation engine 17 for each note into which each singing technique is to be inserted, and records it as an evaluation value of expressiveness. The expressiveness evaluation engine 18 counts the results after playing the music, and notifies the singing instruction selection engine 14 of the singing technique whose evaluation value is equal to or less than the threshold value.

  FIG. 7 is a diagram illustrating an example of processing contents of the expressive power evaluation engine 18. In the example shown in FIG. 7, note N1 is crumpled and note N3 is classified and recorded as vibrato. The singing pitch evaluation engine 12, the pitch / pitch jump evaluation engine 13, the timing evaluation engine 15, the rhythm evaluation engine 16, the singing technique evaluation engine 17, and the expressive power evaluation engine 18 are examples of the evaluation unit according to the present invention.

  The singing instruction selection engine 14 specifies one or more evaluation items that satisfy the predetermined evaluation results of the pitch / pitch jump evaluation engine 13, the rhythm evaluation engine 16, and the expressive power evaluation engine 18. The singing instruction selection engine 14 is an example of an evaluation item specifying unit according to the present invention. In this embodiment, the singing instruction selection engine 14 performs predetermined weighting on the notification results of the pitch / pitch jump evaluation engine 13, the rhythm evaluation engine 16, and the expressiveness evaluation engine 18, and determines the priority according to the weighting results. The result of selection of one or more singing instructions is notified to the singing instruction composition engine 30. In this embodiment, the singing instruction selection engine 14 acquires instruction history data from the server 50 and dynamically changes the weighting coefficient according to the acquisition result.

  FIG. 8 is a diagram for explaining an example of processing contents of the singing instruction selection engine 14, and FIG. 9 is a diagram showing an example of a priority calculation method. As shown in FIG. 8, priorities are assigned to a plurality of items. In this embodiment, as shown in FIG. 9, the singing instruction selection engine 14 sets the evaluation value for each evaluation item notified by the pitch / pitch jump evaluation engine 13, the rhythm evaluation engine 16, and the expressive power evaluation engine 18 from 100. The subtracted value is multiplied by the weighting coefficient for each evaluation item, and the priority order is determined in descending order of the multiplication result value. For example, the singing instruction selection engine 14 may determine priorities in ascending order of evaluation value for each evaluation item without using a weighting coefficient. Any mode may be used as long as the singing instruction selection engine 14 determines the priority according to the evaluation value for each evaluation item. In the example shown in FIG. 9, the evaluation item “A # 3 to D # 4 pitch jump” has the highest priority, the “back beat” evaluation item is second, and “D # 4 pitch”. The priority of the evaluation item is the third.

  The singing instruction selection engine 14 selects one or more evaluation items in descending order of priority for a plurality of items, and notifies the singing instruction composition engine 30 of the selection result. Note that the number of evaluation items to be selected in descending order of priority can be appropriately changed according to the design or the like. In this embodiment, two high priority items are notified to the singing guidance composition engine 30. In the example shown in FIG. 9, two items of “pitch jump” and “back beat” are notified to the singing guidance composition engine 30.

  The singing instruction composition engine 30 reads out the rhythm template and pitch template corresponding to the notification of the singing instruction selection engine 14 from the storage unit 20 to create melody data, and generates singing instruction data by combining with the corresponding accompaniment.

  FIG. 10 is a block diagram showing an example of the configuration of the singing instruction / composition engine 30. In the figure, a template reading unit 31 reads a rhythm template and a pitch template corresponding to the evaluation item specified by the singing instruction selection engine 14 from the composition template data storage area 22. The template reading unit 31 is an example of a reading unit according to the present invention. The template reading unit 31 reads a template of a genre corresponding to the GM data referenced by the singing pitch evaluation engine 12. At this time, when the singing technique evaluation engine 14 specifies the evaluation item of the singing technique, the template reading unit 31 uses a long note in the case of vibrato, a template including a decorative note in the case of a fist, etc. Select and read a template that is easy to put in a singing technique. That is, as for the evaluation items of the singing technique, each singing technique is associated with a template including a note that makes it easy to practice the singing technique.

  When there are a plurality of rhythm templates corresponding to the evaluation items specified by the singing instruction selection engine 14, the template reading unit 31 selects one of the corresponding rhythm templates according to a predetermined algorithm, Read the selected rhythm template. More specifically, for example, the template reading unit 31 may select any one of a plurality of corresponding rhythm templates according to a predetermined random function, and, for example, assign an identifier to the rhythm template in advance. In addition, the identifiers may be selected in ascending or descending order. Further, the rhythm template selected and read by the template reading unit 31 is not limited to 1, and may be plural. The same applies to the pitch template. When there are a plurality of pitch templates corresponding to the evaluation items specified by the singing instruction selection engine 14, the template reading unit 31 is determined in advance from the corresponding plurality of pitch templates. Select one according to the algorithm, and read the selected template. Further, the number of pitch templates selected and read by the template reading unit 31 is not limited to 1, and may be plural.

  The melody rhythm generation unit 32 and the melody pitch generation unit 33 generate singing instruction song data using the rhythm template and pitch template read by the template reading unit 31. More specifically, the melody rhythm generation unit 32 and the melody pitch generation unit 33 are a sequence of notes in accordance with the sequence of note sounding timings indicated by the rhythm template read by the template reading unit 31, Singing instruction music data representing music composed of a sequence of notes in which the pitch of each note is a pitch according to a temporal change of the pitch indicated by the pitch template read by the template reading unit 31 is generated. The melody rhythm generation unit 32 and the melody pitch generation unit 33 are examples of a music data generation unit according to the present invention. In this embodiment, the melody rhythm generating unit 32 generates rhythm data representing a sequence of note sounding timings according to the rhythm template read by the template reading unit 31. The melody pitch generating unit 33 generates melody data (music data) composed of a sequence of notes according to the pitch template read by the template reading unit 31.

  FIG. 11 is a diagram showing an example of the contents of the rhythm template and the pitch template. As described above, the rhythm template is data representing a sequence of note sounding timings, and the pitch template is data representing a temporal change in pitch. In this embodiment, as shown in FIG. 11, a plurality of rhythm templates and pitch templates are stored for each genre of music (pops, enka, jazz, samba, etc.). Note that the rhythm template and the pitch template are not limited to each genre, and may be stored for each tempo. Furthermore, for each of a plurality of genres, as a rhythm template, a plurality of rhythm templates are stored for each item such as evaluation items “table beat”, “back beat”, “triplet”, and “16th note”. Has been. The “table beat” rhythm template RT11,... Is data representing a sequence of note sounding timings (sounding start timing, sounding end timing (or sounding time length)) according to the table beat rhythm. The “back beat” rhythm template RT21,... Is data representing a sequence of note sounding timings according to the back beat rhythm. The “triplet” rhythm template RT31,... Is a plurality of rhythm templates corresponding to “triplet”, and is data representing a pronunciation timing sequence including a triplet. The “16th note” rhythm template RT41,... Is data representing a sequence of pronunciation timings including the “16th note” 16th note.

  Further, as shown in FIG. 11, for each of a plurality of genres, pitch templates PT51,... Whose evaluation item is “pitch jump” and pitch templates PT61,. The pitch template PT51 of “pitch jump” is data representing a temporal change in pitch including a portion where the pitch difference (pitch difference) between adjacent quarter notes is large (greater than a predetermined threshold). . The “decorative note” pitch template PT61,... Is data representing a temporal change in pitch including the ornamental note. In this embodiment, as shown in FIG. 11, the pitch templates PT51,..., PT61,.

  For template reading, the template reading unit 31 takes into consideration the genre of the song sung (pops, enka, jazz, samba, etc.), tempo, and past teaching history, in addition to the notification result of the singing guidance selection engine 14. Randomly select from those that match the conditions.

  Here, the processing contents of the melody rhythm generation unit 32 and the melody pitch generation unit 33 will be described with reference to FIGS. (A) of FIG. 12 is a figure which shows an example of the content of the rhythm data produced | generated by the melody rhythm production | generation part 32. FIG. The melody rhythm generation unit 32 generates rhythm data representing a sequence of notes indicated by the rhythm template read by the template reading unit 31. FIG. 12A is a diagram illustrating an example of the content of rhythm data generated when the rhythm template RT21 (see FIG. 11) corresponding to the evaluation item “back beat” is read. In the example shown in FIG. 12A, rhythm data representing notes N101, N102, N103, N104, N105,... Is generated.

  Next, the melody pitch generating unit 33 specifies the pitch of each note constituting the rhythm data generated by the melody rhythm generating unit 32 according to the pitch template read by the template reading unit 31. Here, the operation when the template reading unit 31 reads the rhythm template RT21 and the pitch template PT51 illustrated in FIG. 11 will be described. FIG. 12B is a diagram schematically showing a temporal change in pitch represented by the pitch template PT51. In FIG. 12B, the horizontal axis indicates time, and the vertical axis indicates pitch. As shown in the figure, the pitch template PT51 represents a change in pitch over time in units of quarter notes.

  The melody pitch generating unit 33 specifies, for each note indicated by the rhythm template, a pitch corresponding to the time axis direction in the pitch template as the pitch of each note. FIG. 12C is a diagram schematically showing each note when the pitch of each note included in the rhythm template RT21 is specified according to the pitch template PT51. In FIG. 12C, the horizontal axis indicates time, and the vertical axis indicates pitch. In the example shown in FIG. 12C, the pitch p1 is specified as the pitch of the notes N101 and N104, and the pitch p2 is specified as the pitch of the notes N102, N103, and N105.

  As shown in FIG. 12C, the melody pitch generating unit 33 specifies the pitch at the time in the time axis direction corresponding to each note in the pitch template as the pitch of each note indicated by the rhythm template. At this time, there may be a plurality of pitches corresponding to the notes indicated by the rhythm template. Specifically, for example, the note indicated by the rhythm template is a half note, and the pitch of two quarter note notes of the pitch template corresponding to the note of the half note is different. In such a case, the melody pitch generating unit 33 may specify the pitch that appears first in the time axis direction as the pitch of the note. Note that the method for specifying the pitch of each note by the melody pitch generating unit 33 is not limited to the above-described mode, and may be another mode. For example, the melody pitch generation unit 33 generates a string of notes in which the sound generation timing and the pitch are specified by associating the notes indicated by the rhythm template with the notes indicated by the pitch template on a one-to-one basis. It may be. In short, the melody pitch generation unit 33 is a sequence of notes according to the sequence of note generation timings indicated by the rhythm template, and the pitch of each note follows the temporal change of the pitch indicated by the pitch template. Anything can be used as long as it generates a sequence of notes having a pitch.

  FIG. 13 is a diagram showing the melody data generated by the rhythm template RT21 and the pitch template PT51, and shows the note sequence schematically shown in FIG. 12C as a score. That is, FIG. 13 shows melody data in which a pitch pattern including “pitch jump from A # 3 to D # 4” is added to rhythm data including “back beat” generated by the melody rhythm generation unit 32. Has been generated.

  FIG. 14 and FIG. 15 are diagrams for explaining the contents of processing when melody data is generated using rhythm template RT11 and pitch template PT61 shown in FIG. FIG. 14A illustrates a case where a rhythm template RT11 including “table beat” is selected and melody rhythm generation unit 32 generates rhythm data.

  FIG. 14B schematically shows a temporal change in pitch represented by the pitch template PT61, and FIG. 14C shows the pitch of each note included in the rhythm template RT11. FIG. 6 is a diagram schematically showing each note when specified according to a pitch template PT61. In FIGS. 14B and 14C, the horizontal axis indicates time, and the vertical axis indicates pitch. The process in which the melody pitch generation unit 33 generates melody data using the pitch template PT61 is the same as the process described with reference to FIGS. In the example shown in FIG. 14C, the pitch p1 is specified as the pitch of the notes N201 and N204, and the pitch p2 is specified as the pitch of the notes N202, N203, and N205.

  FIG. 15 is a diagram showing the melody data generated by the rhythm template RT11 and the pitch template PT61, in which the sequence of notes schematically shown in FIG. 14C is shown as a score. In the example shown in FIG. 15, a pitch pattern including a decoration note (note N203) into which “fist” is to be inserted is added to the rhythm data including “table beat” generated by the melody rhythm generator 32. . The melody data is generated as described above.

  By the way, when a plurality of rhythm templates are read by the template reading unit 31, the melody rhythm generation unit 32 may generate rhythm data by using a plurality of rhythm templates in order, for example. Specifically, for example, when three rhythm templates RT11, RT12, RT13 are read by the template reading unit 31, the melody rhythm generating unit 32 is a combination of the three rhythm templates RT11, RT12, RT13. May be generated as rhythm data. The same applies to the pitch template. When a plurality of pitch templates are read out by the template reading unit 31, the melody pitch generating unit 33 may generate the melody data using the plurality of pitch templates in order. When the number of read rhythm templates and the number of pitch templates are different, the melody rhythm generation unit 32 or the melody pitch generation unit 33 repeatedly uses the smaller template for multiple times of the rhythm template. The length of the length and the time length of the pitch template may be aligned.

  The accompaniment reading unit 34 reads accompaniment data corresponding to the template read by the template reading unit 31 from the accompaniment data storage area 23. At this time, when there are a plurality of accompaniment data corresponding to the template read by the template reading unit 31, the accompaniment reading unit 34 selects and reads any accompaniment data according to a predetermined random function, for example. Alternatively, for example, the number of times accompaniment data is read may be stored in the storage unit 20 as a history, and accompaniment data with a small number of times read may be selected and read. The mode of selection of accompaniment data can be changed as appropriate according to the design and the like. The melody data generated by the melody pitch generating unit 33 and the accompaniment data read by the accompaniment reading unit 34 are combined by the adder 35 and output to the sequencer 60. The sequencer 60 reproduces the singing instruction music data supplied from the singing instruction composition engine 30 and emits the singing instruction music from the speaker 81.

  FIG. 16 is a block diagram illustrating an example of a hardware configuration of the karaoke apparatus 1 having the singing instruction function according to the present embodiment. The karaoke apparatus 1 is an example of a music practice support apparatus according to the present invention. The karaoke apparatus 1 includes a CPU 2, a ROM 3, a RAM 4, a display unit 5, an operation unit 6, a sound processing unit 7, a microphone 80, a speaker 81, and a storage unit 20, and these units are connected via a bus 9. . The CPU 2 reads out the computer program stored in the ROM 3 or the storage unit 20 and loads it into the RAM 4 and executes it, thereby realizing the singing instruction engine 10, the singing instruction composition engine 30, and the sequencer 60 shown in FIG. The

  The operation unit 6 includes various operators and outputs an operation signal representing the operation content of the singer to the CPU 2. The display unit 5 includes a liquid crystal panel, for example, and displays various images such as lyrics telop and background video according to each karaoke piece under the control of the CPU 2. The microphone 80 outputs an analog audio signal representing the collected audio to the audio processing unit 7. The audio processing unit 7 has an A / D (Analog / Digital) converter, converts an analog audio signal output from the microphone 80 into digital audio data, and outputs the digital audio data, and the CPU 2 acquires this. The audio processing unit 7 includes a D / A (Digital / Analog) converter, converts digital audio data received from the CPU 2 into an analog audio signal, and outputs the analog audio signal to the speaker 81. The speaker 81 emits a sound based on the analog audio signal received from the audio processing unit 7. In FIG. 16, the voice processing unit 7 is illustrated. However, in FIG. 1, the voice processing unit 7 is not illustrated in order to prevent the drawing from being complicated.

  In this embodiment, the case where the microphone 80 and the speaker 81 are included in the karaoke apparatus 1 will be described. However, the audio processing unit 7 is provided with an input terminal and an output terminal, and the input terminal is connected to the input terminal via an audio cable. An external microphone may be connected, and similarly, an external speaker may be connected to the output terminal via an audio cable. In this embodiment, the audio signal output from the microphone 80 to the speaker 81 is an analog audio signal. However, digital audio data may be input / output. In such a case, it is not necessary to perform A / D conversion or D / A conversion in the voice processing unit 7. The same applies to the operation unit 6 and the display unit 5, and an external output terminal may be provided and a monitor may be connected to the outside.

  The lyric data storage area 24 stores lyric data indicating the lyrics of each song and background video data representing the background video displayed on the background of the lyrics. The lyrics indicated by the lyrics data are displayed on the display unit 5 as lyrics telop as the music progresses during karaoke singing. Further, the background video represented by the background video data is displayed on the display unit 5 as the background of the lyrics telop as the music progresses during karaoke singing.

<Operation>
FIG. 17A and FIG. 17B are flowcharts showing the flow of processing of this embodiment. When a song is requested by the singer via the operation unit 6 (step S101), the CPU 2 searches for the requested song from the storage unit 20. Specifically, the CPU 2 searches the music information data storage area 21 and the lyrics data storage area 24 using the music number or the music title of the selected music as a key and searches for data related to the music, and the search result data Is read into the RAM.

  Next, the CPU 2 reproduces karaoke music based on the accompaniment data, video data, and GM data stored in the RAM (step S102). Specifically, the CPU 2 causes the speaker 81 to emit sound based on the accompaniment data and GM data, and causes the display unit 5 to display a video based on the video data.

  When singing is started by the singer (step S103), the CPU 2 obtains singing voice data obtained by converting the singing voice of the singer collected by the microphone 80 into digital data by the voice processing unit 7, The data is stored in a predetermined area of the storage unit 20. CPU2 analyzes singing voice data (step S104), and specifies the pitch of singing voice data.

  The CPU 2 performs a pitch / pitch jump evaluation process (step S105), a timing evaluation process (step S106), and an expressive power evaluation process (step S107) according to the analysis result of the singing voice data. That is, the CPU 2 performs processing of the pitch / pitch jump evaluation engine 13, the rhythm evaluation engine 16, and the expressiveness evaluation engine 18 described above.

  When the singer finishes singing (step S108) and the reproduction of the karaoke music ends (step S109), the CPU 2 determines the singer's according to the difference between the pitch specified from the singing voice data and the pitch indicated by the GM data. The singing voice is evaluated, and the evaluation result is displayed on the display unit 5 (step S110).

  Next, the CPU 2 performs processing of the singing instruction selection engine 14 described above. That is, the singing instruction selection engine 14 acquires instruction history data from the server 50 (step S111), and dynamically changes (adjusts) the singing instruction selection coefficient (weighting coefficient) according to the acquired instruction history data (step S112). ). Specifically, the singing instruction selection engine 14 acquires instruction history data representing the instruction history of the singer from the server 50 on the network 40, changes the weighting coefficient according to the acquisition result, and the instruction contents are changed. Adjust dynamically to avoid bias. More specifically, for example, the singing instruction selection engine 14 refers to the instruction history data and totals the number of times the singing instruction song data has been generated so far for each evaluation item, and the total result is large (that is, singing). The value of the weighting coefficient may be decreased as the number of times that the instruction music data is generated). By doing in this way, the priority of the evaluation item which has not practiced so far can be made high. Note that the manner of changing the weighting coefficient is not limited to that described above. For example, the singing instruction selection engine 14 may dynamically change the value of the weighting coefficient for each evaluation item according to a predetermined random function. In addition, the weighting coefficient for each evaluation item may be set in advance without dynamically changing the weighting coefficient value.

  Next, the singing instruction selection engine 14 acquires evaluation results from the pitch / pitch jump evaluation engine 13, the rhythm evaluation engine 16, and the expressiveness evaluation engine 18 (step S113), and the priority of the singing instruction contents (each evaluation item). Is calculated (step S114). The singing instruction selection engine 14 selects the singing instruction content (an evaluation item for generating singing instruction song data) according to the calculated priority (step S115).

  Next, the CPU 2 performs the process of the template reading unit 31 described above. That is, the template reading unit 31 acquires the genre and tempo information of the reproduced music (step S116), and acquires information indicating the singing instruction selection result (selected evaluation item) from the singing instruction selection engine 14 ( Step S117). The template reading unit 31 selects a rhythm template and a pitch template according to the acquisition results of step S116 and step S117 (step S119).

  When the rhythm template and the pitch template are selected, the CPU 2 performs processing of the melody rhythm generation unit 32, the melody pitch generation unit 33, and the accompaniment reading unit 34. That is, the melody pitch generating unit 33 acquires the rhythm template read by the template reading unit 31 (step S120). In addition, the melody pitch generating unit 33 acquires the pitch template read by the template reading unit 31 (step S121). Further, the accompaniment reading unit 34 acquires accompaniment data corresponding to the template read by the template reading unit 31 (step S122).

  The melody rhythm generation unit 32 generates rhythm data according to the acquired rhythm template (step S123). The melody pitch generation unit 33 generates melody data by adding a pitch according to the acquired pitch template to the rhythm data generated by the melody rhythm generation unit 32 (step S124). The adder 35 of the singing instruction composition engine 30 combines the melody data generated by the melody pitch generating unit 33 and the accompaniment data read by the accompaniment reading unit 34 (step S125) to generate singing instruction song data. (Step S126).

  When the generation process of the singing instruction music data is completed, the CPU 2 performs a process of causing the singer to select whether or not to perform the singing instruction (step S127). In this process, for example, the CPU 2 may cause the display unit 5 to display a screen for selecting whether or not to perform singing instruction. The singer selects whether or not to perform the singing instruction by operating the operation unit 6. When it is selected that the singing instruction is not performed by the singer (step S127; NO), the CPU 2 ends the process as it is. On the other hand, when it is selected in step S127 that singing instruction is performed (step S127; YES), the CPU 2 starts singing instruction (step S128). In other words, the sequencer 60 reproduces the singing instruction music data generated by the singing instruction composition engine 30.

  The singer practices singing with pronunciation such as “La La La ...” in accordance with the singing instructional music to be reproduced. When the reproduction of the singing instruction music is finished, the CPU 2 analyzes the singing voice data (step S129), ends the singing instruction (step S130), and displays the singing evaluation result on the display unit 5 (step S131).

  As described above, in this embodiment, the CPU 2 compares the pitch of the singing voice of the singer with the pitch of the GM data, and evaluates the singing voice of the singer with respect to a plurality of evaluation items according to the comparison result. The CPU 2 classifies and summarizes the evaluation results for each evaluation item such as a pitch or pitch jump, and determines an evaluation item that satisfies a predetermined evaluation value as an item that the singer is not good at. CPU2 dynamically produces | generates the song guidance music which can practice intensively the evaluation item determined to be weak, and provides it to a singer. This makes it possible to provide singers with singing instructional songs that are not limited to the standard singing practice song, but with various melody and rhythm variations. Moreover, since CPU2 produces | generates song guidance music data using the template provided for every genre, it can perform the guidance which does not make a singer tired by practicing peculiar rhythm and tempo by a genre, and the same practice.

<Modification>
The above embodiment can be modified as follows. In addition, you may implement the following modifications suitably combining. Moreover, you may implement combining the said embodiment and the following modifications.

<Modification 1>
In the above-described embodiment, the singing pitch specifying engine 11 specifies the pitch of the voice collected by the microphone 80, but the voice feature specified by the feature specifying unit according to the present invention is not limited to the pitch of the voice. The characteristics of speech include, for example, speech volume, frequency analysis results obtained using FFT (Fast Fourier Transform), power fluctuation of specific overtone, ratio of power of specific overtone and fundamental, power of overtone component And the ratio of the fundamental power, the S / N ratio, and the loudness (a value obtained by correcting the volume according to the frequency characteristics of the hearing. Also referred to as “A characteristic sound pressure level” or “sound level”), etc. Any device may be used as long as it represents the characteristics of the above. Any configuration may be used as long as the voice is evaluated using the specified feature. There are cases where the accuracy of the evaluation process can be improved by performing the evaluation using features other than the pitch of the voice.

<Modification 2>
In the above-described embodiment, the CPU 2 of the karaoke apparatus 1 performs “pitch”, “pitch jump”, “table beat”, “back beat”, “triplet”, “16th note”, “vibrato”, “shrimp”. Evaluation items such as “”, “fist”, and “fall” were evaluated, but the evaluation items are not limited to those described above. For example, evaluation items such as tremolo, and false set may be evaluated. In addition, the “evaluation item” according to the present invention includes a specific scale such as “scale D # 4” and “scale D4”, or a specific scale such as “scale A # 3 to scale D # 4”. It also includes musical scale jumping to the musical scale. “Musical scale D # 4” and “musical scale D4” may be handled as different evaluation items, respectively, or may be handled as a group of evaluation items “musical scale”.

<Modification 3>
In the above-described embodiment, the singing instruction selection engine 14 presets a weighting coefficient for each evaluation item, multiplies the evaluation value for each evaluation item by the weighting coefficient, and specifies two evaluation items according to the multiplication result. The specific aspect of the evaluation item is not limited to this. For example, the singing instruction selection engine 14 may multiply the evaluation value for each evaluation item by a weighting coefficient for each evaluation item, and specify the evaluation item having the smallest multiplication result. Further, as another example, for example, the singing instruction selection engine 14 specifies the evaluation value for each evaluation item in advance, such as specifying three scales with the lowest evaluation value in the order of the evaluation value from the bottom. Evaluation items that are less than or equal to the threshold value may be specified. In short, the singing instruction selection engine 14 may be any engine as long as it selects one or a plurality of evaluation items whose evaluation values satisfy a predetermined condition. Also in this aspect, since the singer's poor evaluation items are selected in the same manner as in the above-described embodiment, the singer can practice singing on items that he is not good at.

<Modification 4>
In the above-described embodiment, the template reading unit 31 reads the rhythm template and the pitch template corresponding to the evaluation item specified by the singing instruction selection engine 14 from the composition template data storage area 22, but is not limited thereto. One of the rhythm template and the pitch template may be read out. For example, when the template corresponding to the evaluation item specified by the singing instruction selection engine 14 is only the rhythm template, the template reading unit 31 may read only the rhythm template. In this case, the singing instruction music composition engine 30 may generate the singing instruction music data using the rhythm template read out by the template reading unit 31 and the predetermined pitch template, and only the rhythm template is read out. In this case, the method for determining the pitch of each note is arbitrary. In short, the singing instruction music composition engine 30 generates singing instruction music data representing music composed of a sequence of notes according to the sequence of pronunciation timings of notes indicated by the rhythm template read by the template reading unit 31. Any thing can be used. For example, when generating a practice song for a drum, the singing instruction / composition engine 30 generates melody data in which only the pronunciation timing of each note is specified without specifying the pitch of each note constituting the song. May be. According to this aspect, melody data can be generated without using a pitch template.

  As another example, for example, when the template corresponding to the evaluation item specified by the singing instruction selection engine 14 is only the pitch template, the template reading unit 31 may read only the pitch template. In this case, the singing instruction composition engine 30 may generate the singing instruction song data using the pitch template read by the template reading unit 31 and a predetermined rhythm template, and only the pitch template is read out. In this case, the method for determining the sound generation timing of each note is arbitrary. In short, the singing instruction / composition engine 30 generates melody data representing music composed of a sequence of notes according to a temporal change in pitch indicated by the pitch template read by the template reading unit 31. Anything is acceptable. In the above-described embodiment, the rhythm template and the pitch template are stored in the composition template data storage area 22. However, the rhythm template is not stored, and only the pitch template is stored. It is good. Moreover, it is good also as a structure which is not set as the structure which memorize | stores a pitch template, but memorize | stores only the rhythm template in the template data storage area 22 for composition.

  Further, as another example, a template representing a note sequence indicating a sound generation timing and a pitch (that is, a template in which a rhythm template and a pitch template are integrated) is not used as a configuration in which a pitch template and a rhythm template are used separately. It is good also as a structure to use. Also in this case, as in the above-described embodiment, the template is stored for each evaluation item in the composition template data storage area 22, and the template reading unit 31 is evaluated by the singing instruction selection engine 14. What is necessary is just to read the template corresponding to an item. The pitch template according to the present invention only needs to be data representing at least a change with time of the pitch, and the pitch template according to the present invention also includes note string data indicating both the sounding timing and the pitch.

  In the above-described embodiment, the pitch template representing the pitch per unit time length corresponding to the quarter note is used, but the pitch template is not limited to this. For example, a pitch template that represents a temporal change in pitch with an eighth note as a unit time length may be used, and any pitch template that represents a temporal change in pitch may be used. .

<Modification 5>
In the above-described embodiment, the singing instruction composition engine 30 generates singing instruction tune data obtained by synthesizing the melody data generated by the melody pitch generation unit 33 and the accompaniment data, and outputs the singing instruction tune data to the sequencer 60. Instead of synthesizing the data, only the melody data generated by the melody pitch generating unit 33 may be output. Also in this aspect, the singer can practice singing according to the melody data to be reproduced, as in the above-described embodiment.

  In the above-described embodiment, the accompaniment data is stored in association with the template. However, the present invention is not limited to this. For example, accompaniment data may be stored for each genre. In this case, the accompaniment reading unit 34 may read and output the accompaniment data corresponding to the genre type of the GM data referred to by the singing pitch evaluation engine 12. According to this aspect, singing instruction music data can be generated using the accompaniment data suitable for the genre of music.

<Modification 6>
In the above-described embodiment, the rhythm template and the pitch template may be stored in association with each other, and the template reading unit 31 may read the rhythm template and the pitch template corresponding to each other. According to this aspect, since rhythm templates and pitch templates corresponding to each other are used, music can be generated more naturally.
In the above-described embodiment, a template may be prepared for each difficulty level. In this case, the rhythm template and the pitch template are stored in the composition template data storage area 22 for each evaluation item and each difficulty level. In this case, the template reading unit 31 selects a difficulty level according to the evaluation value for each evaluation item specified by the singing instruction selection engine 14, and generates a rhythm template and a pitch template corresponding to the selected difficulty level as template data for composition. Read from the storage area 22. According to this aspect, the singing instruction song data can be generated using the template corresponding to the skill of the singer's singing.

  Moreover, in the above-mentioned embodiment, the vocal range which can be sung may be analyzed according to the singer's previous instruction history, and singing instruction may be performed to gradually expand the vocal range by gradually eliminating poor pitches. Specifically, for example, pitch templates having different sound ranges to be used are stored in the composition template data storage area 22, and the template reading unit 31 follows the instruction history data stored in the instruction history data storage area 51. It is good also as a structure which reads a pitch template so that a sound range may spread gradually, whenever it practices. According to this aspect, it is possible to perform singing guidance that broadens the voice range by gradually eliminating the poor pitch of the singer.

<Modification 7>
In the above-described embodiment, the CPU 2 reproduces the generated singing instruction song data, but not limited thereto, the singing instruction song data may be output to an externally connected storage device. You may make it output song guidance music data by transmitting to the server apparatus connected via the communication network. According to this aspect, the singing instruction song data generated by the karaoke apparatus 1 can be reproduced by another apparatus.

  In the above-described embodiment, the singing instruction composition engine 30 generates the singing instruction data every time. However, the singing instruction data storage area for storing the singing instruction data is provided in the storage unit 20, and the singing instruction is provided. It is good also as a structure which reads and reproduce | regenerates the song guidance music data corresponding to the evaluation item selected by the selection engine 14 from the said storage area. According to this aspect, it is not necessary to generate singing instruction music data each time, and the processing load of the singing instruction composition engine 30 can be reduced.

<Modification 8>
In the above-described embodiment, in addition to the rhythm template and the pitch template, the lyrics template representing the lyrics is stored in the composition template data storage area 22, and the template reading unit 31 reads the rhythm template and the pitch template. In addition, the lyrics template may also be read and the lyrics represented by the read lyrics template may be displayed on the display unit 5 or the like. In this case, the lyrics template may be stored in association with the rhythm template, and the lyrics template corresponding to the rhythm template read by the template reading unit 31 may be used. According to this aspect, it is possible to present a song teaching song with lyrics to the singer.

  Moreover, in the above-mentioned embodiment, it is set as the structure which provides the song synthesis engine 70 (it shows with a dashed line in FIG. 1) which produces | generates a synthetic song sound, and may reproduce | regenerate the song synthesis sound used as a model at the time of song guidance. That is, the singing voice synthesis engine 70 may synthesize the singing voice as an example and the sound represented by the singing guidance song data, and output the synthesized sound to the sequencer 60 for reproduction. In this case, the singing voice synthesis engine 70 may generate a singing voice synthesis sound using a predetermined phoneme string such as “La La La ...”, or, for example, using the above-described lyrics template. A singing synthesized sound corresponding to the lyrics represented by the read lyrics template may be generated. According to this aspect, it is possible to present a singing synthesized sound that serves as a model during singing guidance to the singer.

<Modification 9>
In the above-described embodiment, the CPU 2 may perform the singing evaluation again after performing the singing instruction. More specifically, after finishing the process of step S131 of FIG. 17, it returns to the process of step S110, and performs a song evaluation process (step S110) and a song guidance music production | generation process (step S111 thru | or step S126) again. May be. According to this aspect, the singing evaluation can be performed again on the singing voice after the singing instruction.

  In the above-described embodiment, the CPU 2 of the karaoke apparatus 1 has a configuration in which the singer selects whether or not to perform the singing instruction after generating the singing instruction tune data, but the processing order is limited to this. It is good also as composition which CPU2 generates song guidance music data only when it is made to choose a singer to perform singing guidance and it is chosen when singing guidance is performed by a singer. According to this aspect, when the singer does not desire singing instruction, the singing instruction song data is not generated, so that the processing load on the CPU 2 can be reduced.

  In the above-described embodiment, the CPU 2 performs the scoring process in real time while the singing is performed. However, the CPU 2 stores the singing voice data in a predetermined storage area. In addition, scoring may be performed after the singing ends. According to this aspect, since the scoring process is not performed during the reproduction of the karaoke accompaniment, the processing load on the CPU 2 during the reproduction of the karaoke accompaniment is reduced.

<Modification 10>
In the above-described embodiment, the CPU 2 generates singing instruction music data for the singer to practice singing. However, the CPU 2 generates performance instruction music data for the player who performs the musical instrument to perform the performance practice. May be. The “speech” referred to in the present embodiment includes various sounds such as a sound generated by a person and a performance sound of a musical instrument.

<Modification 11>
In the above-described embodiment, two or more devices connected by a communication network share the functions related to the karaoke device 1 of the above-described embodiment, and a system including the plurality of devices uses the karaoke device 1 of the same embodiment. It may be realized. For example, a singing instruction engine 10, a storage unit 20, a singing instruction composition engine 30, and the like shown in FIG. 1 and a second computer apparatus including a sequencer 60, a speaker, and the like are communication networks. It may be configured as a connected system. In this case, for example, the first computer device generates singing instruction song data and transmits it to the second computer device, and the second computer device reproduces the received singing instruction song data and releases it from the speaker. It is good also as a structure to sound.

<Modification 12>
In the above-described embodiment, the singing instruction engine 10, the singing instruction composition engine 30, and the sequencer 60 shown in FIG. 1 are realized by the CPU 2 reading and executing the computer program stored in the ROM 3 or the storage unit 20, The configuration is not limited to the CPU, and may be realized by another control unit such as an MPU (Micro-Processing Unit) or a DSP (Digital Signal Processor). In addition, the singing instruction engine 10, the singing instruction composition engine 30, and the sequencer 60 are not limited to being implemented as software by a control unit such as a CPU, and the karaoke apparatus 1 corresponds to each of the functional elements shown in FIG. You may have exclusive hardware (circuit).

<Modification 13>
In addition to the music practice support device, the present invention can be understood as a method for realizing these and a program for causing a computer to realize a music practice support function. Such a program may be provided in the form of a recording medium such as an optical disk storing the program, or may be provided in the form of being downloaded to a computer via the Internet or the like and installed and used. According to this aspect, a service according to the above-described embodiment can be provided by a home PC (Personal Computer), a mobile terminal, or the like (including a smartphone).

<Modification 14>
In the above-described embodiment, the singing instruction composition engine 30 generates the singing instruction data according to the evaluation item selected by the singing instruction selection engine 14, but is not limited to generating the singing instruction data, but is stored in advance. The configuration may be such that the singing instruction song data is read out. This aspect will be described with reference to FIG. FIG. 18 is a block diagram illustrating an example of a system configuration in this aspect, and corresponds to the block diagram illustrated in FIG. 1 in the above-described embodiment. The system shown in FIG. 18 differs from the system shown in FIG. 1 in that it has a song teaching song data storage area 25 instead of the composition template data storage area 22 and the accompaniment data storage area 23, and the song It is the point which has replaced with the instruction | indication composition engine 30, and has the singing instruction | indication music selection engine 30A, About another component, it is the same as that of FIG. 1 in embodiment mentioned above. Therefore, in the following description, the same components as those in the above-described embodiment are denoted by the same reference numerals, and the description thereof is omitted as appropriate.

  In the song guidance song data storage area 25, song guidance song data is stored in advance in association with evaluation items. The singing instruction song data is the same as the singing instruction song data in the above-described embodiment. In this aspect, the singing instruction selection engine 14 selects the evaluation item with the highest priority, and the singing instruction selection engine 30A is stored in association with the evaluation item selected by the singing instruction selection engine 14. The song data is read from the singing instruction song data storage area 25 and supplied to the sequencer 60 for reproduction. In this aspect, the singing guidance song data is stored in association with the evaluation item, but instead of the singing guidance song data, melody data composed of a string of notes is stored in association with the evaluation item. The singing instruction music selection engine 30A may read out and reproduce the melody data corresponding to the evaluation item selected by the singing instruction selection engine 14. According to this aspect, it is not necessary to generate singing instruction song data or melody data.

<Modification 15>
In the above-described embodiment, the singing instruction music composition engine 30 generates the singing instruction music data using the template corresponding to the evaluation item selected by the singing instruction selection engine 14, but the generation mode of the singing instruction music data is as follows. Not exclusively. For example, the singing instruction / composition engine 30 generates melody data from the probability distribution of notes learned by performing different learning processes on the score for each evaluation item, analyzes the generated melody data, and matches the corresponding accompaniment sounds. You may produce | generate song guidance music data. In this case, the singing instruction / composition engine 30 randomly generates the first note according to the random function, and generates the second and subsequent notes one or a plurality of previous notes, and for each evaluation item. It is good also as a structure produced | generated according to a predetermined probability distribution. Here, the singing instruction / composition engine 30 may add pitches to the generated musical note sequence in accordance with the learned probability distribution. The music genre may be limited when learning the probability distribution of the notes, or the genre-like singing instruction song data may be generated by adding musical rules when generating the notes according to the probability distribution. In short, the singing instruction / composition engine 30 may be anything as long as it generates music data in a generation mode according to the specified evaluation item.

DESCRIPTION OF SYMBOLS 1 ... Karaoke apparatus, 10 ... Singing instruction engine, 11 ... Singing pitch specific engine, 12 ... Singing pitch evaluation engine, 13 ... Pitch / pitch evaluation engine, 14 ... Singing instruction selection engine, 15 ... Timing evaluation engine, 16 ... Rhythm Evaluation engine, 17 ... Singing technique evaluation engine, 18 ... Expressive power evaluation engine, 20 ... Storage unit, 21 ... Music information data storage area, 22 ... Composition template data storage area, 23 ... Accompaniment data storage area, 30 ... Singing instruction Composition engine, 40 ... network, 50 ... server, 51 ... instruction history data storage area, 60 ... sequencer, 70 ... singing synthesis engine, 80 ... microphone, 81 ... speaker, 100 ... singing practice support system.

Claims (6)

A sound data acquisition unit for acquiring sound data representing sound;
A feature specifying unit for specifying the characteristics of the sound data acquired by the sound data acquiring unit;
An evaluation unit that evaluates the sound data for each predetermined evaluation item based on the features specified by the feature specifying unit;
An evaluation item specifying unit that specifies one or a plurality of the evaluation items that satisfy a predetermined evaluation result by the evaluation unit;
A generating unit that generates music data for supporting the practice of singing or playing the sound data in a generation mode according to the evaluation item specified by the evaluation item specifying unit ;
The evaluation item specifying unit acquires instruction history data indicating a generation history of music data in the generation unit, determines a weighting coefficient for each evaluation item according to the acquired instruction history data, and uses the determined weighting coefficient A music practice support device characterized by performing identification . The generator generates a rhythm template corresponding to an evaluation item specified by the evaluation item specifying unit from a storage unit that stores a pitch template indicating a temporal change in pitch and a rhythm template indicating a timing sequence for each evaluation item. And a pitch template read from the storage unit, and a note sequence in accordance with a sequence of note sounding timings indicated by the read rhythm template, wherein the pitch of each note is the time of the pitch indicated by the read pitch template. The music practice support device according to claim 1, wherein music data representing music composed of a series of notes having a pitch according to a change is generated. A sound data acquisition unit for acquiring sound data representing sound;
A feature specifying unit for specifying the characteristics of the sound data acquired by the sound data acquiring unit;
An evaluation unit that evaluates the sound data for each predetermined evaluation item based on the features specified by the feature specifying unit;
An evaluation item specifying unit that specifies one or a plurality of the evaluation items that satisfy a predetermined evaluation result by the evaluation unit;
A generating unit that generates music data for supporting practice of singing or playing the sound data in a generation mode according to the evaluation item specified by the evaluation item specifying unit;
Comprising
The generator generates a rhythm template corresponding to an evaluation item specified by the evaluation item specifying unit from a storage unit that stores a pitch template indicating a temporal change in pitch and a rhythm template indicating a timing sequence for each evaluation item. And a pitch template read from the storage unit, and a note sequence in accordance with a sequence of note sounding timings indicated by the read rhythm template, wherein the pitch of each note is the time of the pitch indicated by the read pitch template. To generate music data representing music composed of a series of notes that are pitches according to various changes,
At the time of the reading, a rhythm template and a pitch template corresponding to the evaluation item specified by the evaluation item specifying unit from a storage unit storing the plurality of rhythm templates and pitch templates having different difficulty levels for each evaluation item in a, the said evaluation of the evaluation you wherein the reading difficulty of the rhythm template and the pitch template corresponding to the result easier song practice support device. A sound data acquisition unit for acquiring sound data representing sound;
A feature specifying unit for specifying the characteristics of the sound data acquired by the sound data acquiring unit;
An evaluation unit that evaluates the sound data for each predetermined evaluation item based on the features specified by the feature specifying unit;
An evaluation item specifying unit that specifies one or a plurality of the evaluation items that satisfy a predetermined evaluation result by the evaluation unit;
A music selection unit that selects music data corresponding to the evaluation item specified by the evaluation item specifying unit from among a plurality of music data stored in the storage unit in association with the evaluation item ;
The evaluation item specifying unit acquires instruction history data indicating a selection history of music data in the song selection unit, determines a weighting coefficient for each evaluation item according to the acquired instruction history data, and uses the determined weighting coefficient The music practice support apparatus characterized by performing said specification . The generation unit acquires instruction history data indicating a generation history of music data in the generation unit, and stores a plurality of pitch templates having different sound ranges to be used for each evaluation item from the evaluation item specifying unit. A pitch template corresponding to the evaluation item specified by the above-described pitch template corresponding to the acquired instruction history data is read out.
The music practice support device according to claim 2 or 3, characterized in that. Synthesizes the sound representing the singing voice to be typical range of the music data, the music according to any one of claims 1 to 5, characterized by comprising the singing synthesis section for reproducing the synthesized sound Practice support device.

Images