JP4891135B2 - Evaluation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide technology for evaluating singing according to a singing part in a song with a plurality of singing parts. <P>SOLUTION: Singing voice of a singer is collected by microphones 151 and 152, and converted to singing voice data. A control section 11 of a karaoke device 1 detects a pitch from the singing voice data, and generates a singing pitch data. The control section 11 compares the singing pitch data with each of a plurality of reference pitch data by note, and selects one of the reference pitch data based on the comparison result. When the same data is continuously selected predetermined number of times or more, the control section 11 adopts the selected data as data for evaluation. The control section 11 compares the singing voice data with the data for evaluation, and evaluates the singing voice data according to the comparison result. <P>COPYRIGHT: (C)2009,JPO&amp;INPIT

Description

  The present invention relates to a technique for evaluating singing (or performance).

  In a karaoke apparatus, various methods for scoring the skill of a singer's singing have been proposed. For example, in Patent Document 1, when comparing the singing and the reference that is a model, if the timing of the singing and the timing of the reference are shifted, the singing voice data and the reference data are shifted in the time axis direction to cross-correlate. And a method of scoring each note at the position with the highest cross-correlation has been proposed. According to this method, even when a singer sings using the “single” or “rubbert” singing technique, the singing timing can be compared and scored in accordance with the reference timing.

By the way, the karaoke music includes music having a plurality of singing parts such as a duet music sung by two singers and a music sung by a plurality of singers. Patent Document 2 proposes a method for scoring a song in a song having a plurality of song parts.
JP 2005-107330 A JP 2004-109260 PR

However, in the method described in Patent Document 2, the singer can only score the part selected using the part selection switch. The operation of selecting a part to sing in this way is often complicated for the singer. The same applies to the performance of musical instruments.
The present invention has been made under the above-described background, and an object of the present invention is to provide a technique for performing evaluation corresponding to a part being sung (or played) in a song having a plurality of singing (or performance) parts. And

  In order to solve the above-described problems, an evaluation apparatus according to a preferred aspect of the present invention includes a reference data storage unit that stores a plurality of different reference data composed of a sequence of notes, and detects a pitch from the first audio signal. Comparing the pitch detected by the first pitch detection means and the plurality of reference data stored in the reference data storage means for each note, and based on the difference between the plurality of reference data. Selected when first reference means for selecting any one of the reference data for each note and when the same reference data is continuously selected by the first selection means for a predetermined number of times or more. In this case, the reference data used immediately before is used as the evaluation data. A first evaluation data adopting unit that continuously adopts the data as evaluation data, and the evaluation data adopted by the first evaluation data adopting unit and the first audio signal are compared and compared. A first evaluation unit that evaluates the first audio signal based on a result, and an output unit that outputs data indicating an evaluation result by the first evaluation unit.

  In the above aspect, the first selection unit compares the pitch detected by the first pitch detection unit with a plurality of reference data stored in the reference data storage unit for each note, and the difference is the largest. Small reference data may be selected for each note.

  In the above aspect, the second pitch detection means for detecting the pitch from the second audio signal, and the first evaluation data adoption means among the plurality of reference data stored in the reference data storage means. A second evaluation data adopting unit that employs any one of the unprocessed data as the evaluation data, the evaluation data employed by the second evaluation data employing unit, and the second audio signal. A second evaluation unit that compares and evaluates the second audio signal based on the comparison result, wherein the output unit evaluates the evaluation result by the first evaluation unit and the evaluation by the second evaluation unit. Data indicating at least one of the results may be output.

  In the above-described aspect, the second pitch detection means for detecting the pitch from the second audio signal, and a plurality of references stored in the reference data storage means for the pitch detected by the second pitch detection means. A second selection means for comparing data and each note, and selecting any one of the plurality of reference data for each note based on the difference; and a number of times determined in advance by the second selection means When the same reference data is selected in succession, the selected data is adopted as evaluation data. In other cases, the reference data adopted immediately before is continuously adopted as evaluation data. Second evaluation data adoption means, evaluation data adopted by the second evaluation data adoption means, and the first And a second evaluation means for evaluating the second audio signal based on the comparison result, and the output means includes the evaluation result by the first evaluation means and the second evaluation signal. Data indicating at least one of the evaluation results by the evaluation means may be output.

Further, in the above-described aspect, the apparatus includes an evaluation result selection unit that selects either one of the evaluation result by the first evaluation unit and the evaluation result by the second evaluation unit, and the output unit selects the evaluation result selection Data indicating the evaluation result selected by the means may be output.
In the above-described aspect, the evaluation result selection unit may select the evaluation result having the higher evaluation result among the evaluation result by the first evaluation unit and the evaluation result by the second evaluation unit.

In the above aspect, the output unit may calculate an average score between the evaluation result obtained by the first evaluation unit and the evaluation result obtained by the second evaluation unit, and output data indicating the calculation result. .
Further, in the above aspect, the evaluation result selection unit selects one of the evaluation result by the first evaluation unit and the evaluation result by the second evaluation unit for each note, and the output unit includes: Data indicating the evaluation result selected for each note by the evaluation result selecting means may be output in units of notes.
In the above-described aspect, the first power detection means for detecting the power of the audio signal from the first audio signal, and the second power detection for detecting the power of the audio signal from the second audio signal. And the evaluation result selection means compares the power detected by the first power detection means with the power detected by the second detection means, and evaluates the audio signal having the larger power. Results may be selected.

  ADVANTAGE OF THE INVENTION According to this invention, the evaluation corresponding to the part currently sung (or performed) can be performed in the music which has several singing (or performance) parts.

<A: Configuration>
FIG. 1 is a block diagram showing a hardware configuration of a karaoke apparatus 1 according to an embodiment of the present invention. In the figure, the control unit 11 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory), and reads and executes a computer program stored in the ROM or the storage unit 12. Thus, each part of the karaoke apparatus 1 is controlled via the bus BUS. The storage unit 12 is a storage unit for storing a computer program executed by the control unit 11 and data used at the time of execution, and is, for example, a hard disk device. The display unit 13 includes a liquid crystal panel and displays various images under the control of the control unit 11. The operation unit 14 outputs a signal corresponding to an operation by the user of the karaoke apparatus 1 to the control unit 11. Each of the microphones 151 and 152 is a sound collection unit that collects the voice of the singer and outputs an analog signal representing the collected voice signal. The audio processing unit 16 A / D converts analog signals output from the microphones 151 and 152 into digital data. In addition, the audio processing unit 16 performs D / A conversion on the digital data, converts the digital data into an analog signal, and outputs the analog signal to the speaker 17. The speaker 17 is a sound emitting unit that emits sound with an intensity corresponding to the sound signal that is converted from the digital data into an analog signal and output by the sound processing unit 16.

In this embodiment, the case where the microphones 151 and 152 and the speaker 17 are included in the karaoke apparatus 1 will be described. However, the audio processing unit 16 is provided with an input terminal and an output terminal, and the input is performed via an audio cable. An external microphone may be connected to the terminal. Similarly, an external speaker may be connected to the output terminal via an audio cable. In this embodiment, the case where the audio signal input from the microphones 151 and 152 to the audio processing unit 16 and the audio signal output from the audio processing unit 16 to the speaker 17 are analog audio signals will be described. Data may be input / output. In such a case, the audio processing unit 16 does not need to perform A / D conversion or D / A conversion. The same applies to the display unit 13, and an external output terminal may be provided to connect an external monitor.
In this embodiment, two microphones 151 and 152 are shown, but the number of microphones is not limited to two, and may be one, or may be three or more.

  As illustrated in FIG. 1, the storage unit 12 of the karaoke apparatus 1 includes a music data storage area 121, a background image data storage area 122, and an evaluation result data storage area 123. The music data storage area 121 stores music data representing accompaniment sounds and lyrics of music. The background image data storage area 122 stores background image data representing a moving image displayed as a background at the time of karaoke accompaniment. The evaluation result data storage area 123 stores evaluation result data indicating the evaluation result of the singing voice of the singer.

  Here, an example of the contents of the music data stored in the music data storage area 121 will be described. As shown in FIG. 2, the music data has a header and a plurality of tracks. Reference data in which reference data representing the content of a melody (pitch) to be sung by the user is described in the plurality of tracks. There are a data track, a performance track describing performance data representing the contents of karaoke performance sounds, and a lyrics track describing lyrics data representing the contents of lyrics. The music data has reference data tracks as many as the singing parts of the music, and reference data representing the contents of the pitch of each singing part is described in each reference data track. In the example shown in FIG. 2, the music data has two reference data tracks. In the case of a song with one singing part, the song data has one reference data track. In the case of a song with three singing parts, the song data has three reference data tracks. have. When the music data has a plurality of reference data tracks, different reference data are described in each reference data track. The header portion of the music data includes music number data for specifying the music, music title data indicating the music title, genre data indicating the genre, performance time data indicating the performance time of the music, etc., as shown in FIG. Yes. The above music data is described according to the MIDI format.

  Next, a specific example of reference data described in each reference data track will be described. Since FIG. 3 is a matrix of rows and columns, first the columns will be described. The delta time in the first column indicates the time interval between events, and is represented by the number of tempo clocks. When the delta time is “0”, it is executed simultaneously with the immediately preceding event. In the second column, the contents of messages of each event of the performance data are described. This message includes a control change message in addition to a note-on message (NoteOn) indicating a sounding event and a note-off message (NoteOff) indicating a mute event. In the example shown in FIG. 3, the control change message is not included.

The third column describes channel numbers. Here, for simplicity of explanation, the channel number of the reference data track is “1”. When music data includes a plurality of reference data, different numbers are used as channel numbers of the respective reference data.
In the fourth column, a note number (NoteNum) or a control number (CtrlNum) is described. Which is described depends on the content of the message. For example, in the case of a note-on message or a note-off message, a note number indicating a musical scale is described here, and in the case of a control change message, a control number indicating its type is described.
The fifth column describes specific values (data) of the MIDI message. For example, in the case of a note-on message, the velocity value indicating the sound intensity is described here, and in the case of a note-off message, the velocity value indicating the speed at which the sound is turned off is described. If there is, the value of the parameter corresponding to the control number is described.

Next, each row shown in FIG. 3 is a musical sound parameter indicating the attribute of each note of the melody to be sung, and is composed of a note-on event and a note-off event.
In the example shown in FIG. 3, the length of the delta time 480 is the length of a quarter note. In this case, it is indicated that the C4 sound is generated over the length of the quarter note by the event processing of the first row and the second row, and the G4 sound is changed to the quarter note by the event processing of the third row and the fourth row. It is shown to be pronounced over length. Then, it is shown that the F4 sound is pronounced over the length of the half note by the processing of the fifth and sixth lines.

  When the user performs a song designation operation, the designated song data is read from the song data storage area 121 based on the song number data and transferred to the RAM. The controller 11 sequentially reads and processes the music data in the RAM, so that the music performance progresses. At this time, the reference data is also read in synchronization with the progress of the music, and the control unit 11 generates reference pitch data according to the note and velocity of the reference data.

On the other hand, the voice of the singer input to the microphones 151 and 152 becomes an audio signal, which is output from the speaker 17 and input to the audio processing unit 16. After the voice processing unit 16 A / D converts these voice signals, the control unit 11 extracts the pitch of the singing voice and outputs it as singing pitch data. In this case, the process of extracting the pitch of the singing voice is performed every frame (approximately 10 ms).
In the following description, for convenience of explanation, digital data output from the microphone 151 and A / D-converted by the sound processing unit is referred to as “singing sound data SA1”, and is output from the microphone 152 and is processed by the sound processing unit. The A / D converted digital data is referred to as “singing voice data SA2”.

<B: Operation>
FIG. 4 is a flowchart showing a flow of processing performed by the control unit of the karaoke apparatus 1. The operation of this embodiment will be described below with reference to FIG. When the user performs a music designation operation using the operation unit 14, the music data of the designated music is transferred from the music data storage area 121 to the RAM, and the background image data corresponding to the designated music is displayed as the background image. The data is transferred from the data storage area 122 to the RAM. The control unit 11 performs karaoke accompaniment by sequentially reading music data events in the RAM, and executes lyrics and background image display processing by sequentially reading lyrics data and background image data in the RAM. (Step S1). Specifically, the control unit 11 outputs event data described in the performance track of the music data to the sound processing unit 16. The sound processing unit 16 generates a music sound signal based on the supplied event data. The generated music audio signal is output from the speaker 17. In addition, the control unit 11 outputs the background image data and the lyrics data of the lyrics track to the display unit 13. As a result, the karaoke accompaniment sound is emitted from the speaker 17, while the lyrics represented by the lyrics data and the background image represented by the background image data are displayed on the display unit 13.

  Next, the control unit 11 generates reference pitch data indicating the pitch to be sung from the note number and velocity of the note-on event (NoteOn) of the reference data read in accordance with the progress of the music (step S2). At this time, when the music data has a plurality of reference data tracks, the control unit 11 generates reference pitch data for each of the plurality of reference data described in the plurality of reference data tracks.

  The singer sings along with the accompaniment sound emitted from the speaker 17. When there is only one singer, the singer sings by holding the microphone 151 or the microphone 152. On the other hand, when there are two singers, each singer sings by holding the microphone 151 and the microphone 152, respectively. The singing voice of the singer is input from the microphones 151 and 152, and an analog signal is generated. The analog signal is A / D converted by the voice processing unit 16, and the control unit 11 extracts the pitch from the singing voice data SA1 and SA2 A / D converted by the voice processing unit 16, respectively. Output as SP2 (step S3). At that time, the singing pitch is extracted every 10 ms. As will be described later, in the comparison process in step S4, the reference pitch is compared for each note. For this purpose, the average value of the singing pitches extracted within the pronunciation time of the note to be compared is calculated and compared.

  Next, the control unit 11 performs an evaluation by comparing each of the singing voice data SA1 and the singing voice data SA2 with the reference data by executing the processes of steps S4 to S7 described later. At this time, when the music data has a plurality of reference data, the control unit 11 selects any one of the plurality of reference data and compares the selected reference data with the singing voice data for evaluation. Do. This evaluation process performs the same process for each of the singing voice data SA1 and the singing voice data SA2. Therefore, in the following description, the singing voice data SA1 will be described, and the detailed description of the evaluation process for the singing voice data SA2 will be omitted.

  First, the control unit 11 compares the reference pitch data generated in step S2 with the singing pitch data SP1 generated in step S3 in units of notes, and stores comparison result data indicating the comparison result in the RAM (step S4). In addition, when several reference data are contained in music data, the control part 11 compares singing pitch data SP1 with each of several reference pitch data for every note. In this embodiment, the control unit 11 compares the reference pitch data and the singing pitch data SP1 in units of notes, and when the difference between the two is equal to or greater than a predetermined threshold value, the control unit 11 determines “NG”. Otherwise, it is determined as “OK”, and comparison result data indicating the determination result (“OK” or “NG”) is stored in the RAM. In this embodiment, an area for storing comparison result data for three notes is secured in the RAM, and during the karaoke accompaniment, the RAM of the karaoke apparatus 1 shows the comparison results for the past three notes. Data is always stored.

Next, the control unit 11 refers to the comparison result data stored in the RAM and determines whether or not to change the reference data employed as the evaluation data (step S5). This determination method will be described below with reference to a specific operation example shown in FIG.
FIG. 5 shows an example of a change in pitch when a certain singer z sings music having two parts, part 1 and part 2. FIG. In the figure, the horizontal axis indicates time, and the vertical axis indicates pitch. The solid line pz indicates the pitch of the singer z singing voice, the chain line p1 indicates the pitch of part 1, and the solid line p2 indicates the pitch of part 2. In the example shown in FIG. 5, the singer z changes the singing part from part 2 to part 1 in the middle of the music.

  In the figure, evaluation result data D1, D2, D3,... Are data indicating the evaluation result for each note, the evaluation for part 1 is shown in the upper part, and the evaluation result for part 2 is shown in the lower part. Yes. In the evaluation result data D1, D2, D3,..., “O” indicates “OK” and “N” indicates NG. In this embodiment, as described above, in addition to the evaluation result for the latest note, the evaluation for the previous and second previous notes is also recorded as a history. In D1 in this example, for Part 1, all the current and past two notes are NG, for Part 2, the current note is OK, and for the past two notes, it is NG. It is shown.

  The control unit 11 evaluates the singing pitch for the note at the timing of note-off of the reference data of part 1 and part 2. As described above, this evaluation is recorded up to the evaluation of the previous two notes. The control unit 11 determines whether or not to change the reference data adopted as the evaluation data based on the correspondence between the evaluation result of part 1 and the evaluation result of part 2. In this embodiment, the control unit 11 compares the combination of the evaluation result of part 1 and the evaluation result of part 2 with a table stored in advance in the storage unit 12, and changes the reference data used as the evaluation data. It is determined whether or not to do.

  FIG. 6 is a diagram illustrating an example of the contents of a table referred to in this determination process. Each piece of data in the table shown in FIG. 6 shows a correspondence relationship between a combination of evaluation results of part 1 and part 2 and reference data employed as evaluation data. In this table, “1” indicates that the reference data of part 1 is employed, “2” indicates that the reference data of part 2 is employed, and “−” indicates the reference that is currently employed. Indicates that the data will continue to be adopted as evaluation data. In this embodiment, the control unit 11 collates combinations of evaluation result data of the past three notes with the table shown in FIG. 6, and determines whether or not to switch reference data to be adopted as evaluation data.

  In the example shown in FIG. 5, the evaluation for Part 1 is OK for the first time in the third note of Part 1, but the same reference data is still used as evaluation data in this pattern. In the fourth note of the next part 1, since the part 1 is OK twice and the part 2 is NG, referring to the item “NOO”-“ONN” in FIG. 6, it is “1”. Therefore, the evaluation data is changed to the reference data of part 1.

  Returning to the description of FIG. If the determination result of step S5 is affirmative (step S5; YES), the control unit 11 changes the evaluation data based on the contents of the table (step S6). On the other hand, in other cases (step S5; NO), the control unit 11 continues to adopt the reference data adopted immediately before as evaluation data. Note that since there is no reference data adopted immediately before the beginning of the music, in this embodiment, the control unit 11 determines predetermined reference data (for example, reference data of the main part) at the beginning of the music. Etc.) are adopted as evaluation data. Note that the method of using the evaluation data at the beginning of the music is not limited to this. For example, each part is evaluated for each note, and the reference data that has been previously evaluated as “OK” is used as the evaluation data. You may do it.

  Next, the control unit 11 compares the evaluation data changed in step S6 with the singing voice data SA1, evaluates the singing voice data SA1 based on the comparison result, and stores data indicating the evaluation result in the storage unit 12. It memorize | stores in the evaluation result data storage area 123 (step S7). Specifically, for example, the control unit 11 compares the reference data adopted as the evaluation data with the pitch of the singing voice data SA1 in units of frames, and evaluates the evaluation result to be higher as the difference is smaller. Also good. This evaluation process is not limited to the comparison of pitches, and the evaluation may be performed using at least one of each pitch, power, and spectrum. Since this evaluation process is the same as the process performed in the conventional karaoke apparatus, detailed description thereof is omitted here.

  Note that the control unit 11 performs the processing shown in steps S4 to S7 on the singing voice data SA2 as well as the above-described singing voice data SA1, performs reference data selection processing, and selects the selected reference data. And the singing voice data SA2 are compared.

  Next, the control unit 11 selects either one of the evaluation result for the singing voice data SA1 and the evaluation result for the singing voice data SA2, and outputs data representing an image indicating the selected evaluation result to the display unit 13 (step). S8). In this embodiment, the control unit 11 compares the evaluation result for the singing voice data SA1 with the evaluation result for the singing voice data SA2, and outputs image data indicating the evaluation result with the higher evaluation to the display unit 13. The display unit 13 displays an image indicating the evaluation result based on the data output from the control unit 11.

  FIG. 7 is a diagram illustrating an example of a screen displayed on the display unit 13. As shown in the figure, the display unit 13 displays the lyrics telop W1 and images G1, G2,. In the example illustrated in FIG. 5, the control unit 11 represents data representing an image in which a note whose difference between the evaluation data and the singing voice data is equal to or greater than a threshold and other notes are displayed in different colors. Is output to the display unit 13.

  By the way, in general, since the display screen of the karaoke apparatus 1 is often limited in size, when a plurality of singers sing at the same time, displaying the evaluation results of all the singers makes the screen cumbersome. In many cases. On the other hand, in this embodiment, since any one evaluation result is displayed, it can prevent that a screen becomes complicated.

  Returning to the description of FIG. When the karaoke accompaniment of the music has not ended (step S9; NO), the control unit 11 returns to the process of step S2 and performs a reference data selection process and an evaluation process for each note. Thus, the control part 11 repeatedly performs the process of step S2-step S8 until the karaoke accompaniment of the designated music is complete | finished (step S9; YES). That is, in this embodiment, during the reproduction of the karaoke accompaniment, a comparison process with reference data is performed for each note, and the reference data is adopted as evaluation data based on the comparison result.

  When the karaoke accompaniment of the designated music ends (step S9; YES), the control unit 11 aggregates the evaluation results for each note in step S7, and generates data indicating the evaluation results of the entire music (step S10). The control part 11 outputs the data which show the evaluation result of the whole music to the display part 13, and displays an evaluation result on the display part 13 (step S11).

  By the way, when evaluating karaoke singing about a music piece having a plurality of parts, a method of evaluating by considering that the part having the closest pitch is being sung can be considered. However, in this case, even if the pitch is shifted from the original part and accidentally close to the pitch of the other part, the evaluation result becomes high. On the other hand, in this embodiment, even if the pitch of the singer is temporarily deviated from the pitch of the part being sung, it is possible to evaluate the part being sung, thereby enabling more accurate evaluation. It can be performed.

  Moreover, in the conventional karaoke apparatus, it is necessary for the singer to select a singing part to be evaluated, and the singing part to be evaluated cannot be switched in the middle of the song. On the other hand, in this embodiment, even if a singer changes a singing part, the singing part used as evaluation object can be switched and more accurate evaluation can be performed. Moreover, in order to switch the singing part used as evaluation object, a singer needs only to change the part to sing, and does not need to perform complicated operation of selecting a singing part using an operation part.

Thus, according to this embodiment, there is no need for the singer to perform a complicated operation such as selecting a singing part for evaluation using the operation unit, and the singer only sings the part he / she has selected. Thus, an accurate evaluation can be performed.
Moreover, in this embodiment, even if it is a case where a singer changes a song part in the middle of a music, evaluation according to the changed part can be performed.

<C: Modification>
As mentioned above, although embodiment of this invention was described, this invention is not limited to embodiment mentioned above, It can implement with another various form. An example is shown below. In addition, you may combine each following aspect suitably.
(1) In the above-described embodiment, when a plurality of singers sing a song having a plurality of parts, the control unit 11 determines a predetermined microphone and part for a predetermined time at the beginning of the song. The reference data to be adopted as the evaluation data in each microphone may be specified according to the correspondence relationship. In this case, when the predetermined time has passed, the control unit 11 changes the reference data adopted as the evaluation data by performing the evaluation process shown in the above-described embodiment instead of the predetermined correspondence relationship. That's fine.
Conversely, only the beginning part of the music is subjected to the evaluation process shown in the above embodiment to associate the microphone with the part, and after a predetermined time, the evaluation is performed with the pair of the microphone and part determined at the beginning part. You may do it.

(2) In the above-described embodiment, the evaluation result of the singing voice data SA1 is compared with the evaluation result of the singing voice data SA2, and the evaluation result with the higher evaluation is displayed on the display unit 13 (FIG. 4). (See step S8). Instead of this, both evaluation results may be displayed on the display unit 13. Further, for example, the evaluation result of the voice having the higher evaluation result of the immediately preceding note may be displayed.
In addition, the power of the singing voice collected by the microphone 151 is detected, the power of the singing voice picked up by the microphone 152 is detected, and the power of the singing voice is evaluated by comparing both powers. Data indicating the result may be output to the display unit 13. In particular, since there is no previous evaluation result at the beginning of the music, it is not possible to select which singing voice the evaluation result is to be displayed. Therefore, the display timing can be prevented from being delayed by displaying the evaluation result for the singing voice having the higher power at the beginning of the music.

(3) In embodiment mentioned above, the comprehensive evaluation result of both singing voice data SA1 and singing voice data SA2 was displayed on the display part 13 (refer step S11 of FIG. 4). Instead of this, an average of the two evaluation results may be calculated, and the calculation result may be displayed on the display unit 13 as a comprehensive evaluation. Further, for example, one of the comprehensive evaluation results may be displayed on the display unit 13 such that the evaluation result having the higher total score is displayed.

(4) In the above-described embodiment, the singer is notified by displaying the evaluation result on the display unit 13, but the notification mode is not limited thereto, and may be notified by, for example, outputting a voice message. Moreover, the form which transmits the information which shows an evaluation result to a singer's mail terminal in an email format may be sufficient. In addition, information indicating the evaluation result may be output to the storage terminal and stored. In this case, the singer can refer to the information by reading the information from the storage medium using a computer. it can. The evaluation result may be printed out on a predetermined sheet. In short, any information may be output as long as it can output a message or information to the singer by any means.
Further, the evaluation result may be notified in real time during the singing, or the evaluation result may be notified after the singing is finished, and the timing for notifying the evaluation result is arbitrary.

(5) In the above-described embodiment, reference data that is not adopted as the evaluation data for the singing voice data SA1 among the plurality of reference data included in the music data is adopted as the evaluation data for the singing voice data SA2. It may be.

(6) In the above-described embodiment, the reference data to be compared with the singing voice may be, for example, data representing a guide melody of music, or data representing a singing voice serving as an example of music. In other words, any data may be used as long as it represents the sound that serves as an example of the music.

(7) In the above-described embodiment, the singing voice of the singer is evaluated in real time, but it is not always necessary to analyze in real time. For example, the singing voice data is stored in advance in the storage unit and stored. Singing voice data may be evaluated. Further, for example, the karaoke apparatus 1 is provided with a communication unit for performing data transmission via a communication network such as the Internet, so that the singing voice data is received via the communication network, and the received singing voice data is evaluated. It may be.

(8) In the above-described embodiment, the pitch of the voice is used as the voice feature and the reference pitch data and the singing pitch data are compared. However, the voice feature is not limited to the pitch. The sound volume (power) and spectrum may be used as the sound features. For example, in the case of using voice power, the control unit 11 detects the volume from the reference data and generates the reference volume data, and detects the volume from the singing voice data, as in the above-described embodiment. Volume data is generated, and reference volume data and singing volume data are compared in a predetermined unit. Also, the comparison process may be performed using a plurality of features, such as using the pitch and power of the speech as the features of the speech.

(9) In the above-described embodiment, the control unit 11 generates the reference pitch data from the reference data included in the music data stored in the music data storage area 121. Instead, this is an example. Reference pitch data representing the pitch of the sound may be stored in the storage unit 12 in advance.

(10) In the above-described embodiment, the control unit 11 compares the reference pitch data and the singing voice pitch data for each note. However, the unit to be compared is not limited to a note unit, for example, a frame unit. For example, the comparison may be performed for each measure, and the comparison may be performed for each predetermined unit. If reference data is selected in comparison with each frame (about 10 ms), even if the part to be sung in the middle of a note is changed in the case of a melody composed of notes with a long pronunciation time, it will be changed accordingly Evaluation can be performed. In this case, it is effective that the condition for switching the reference data for evaluation in steps S5 and S6 is, for example, at a time point when the selection is continuously performed 10 times or more. That is, the number of times is set larger than the case of comparing for each note.

(11) In the above-described embodiment, the singing voice of the singer and the reference data are compared. However, instead of the singing voice of the singer, the performance sound of the instrument by the performer and the reference data may be compared. As described above, the “voice” referred to in the present embodiment includes various sounds such as a voice generated by a person and a performance sound of a musical instrument.

(12) In the above-described embodiment, it is determined whether or not to change the evaluation data by comparing the singing pitch data with a plurality of reference data for each note and comparing the comparison result with a predetermined table. . Instead, for example, when “OK” continues for a predetermined number of times in any reference, the reference data may be adopted as the evaluation data. For example, referring to the evaluation results of three buffered notes, the reference data having the larger number of “OK” may be adopted as the evaluation data. In short, the singing pitch data may be compared with a plurality of reference data for each note, and it may be determined whether or not to change the evaluation data based on a predetermined number of notes.

  Further, as another example of the evaluation data changing process, for example, singing pitch data is compared with a plurality of reference pitch data for each note, and any one of the plurality of reference data is recorded as a note based on the difference. When the same reference data is selected continuously for a predetermined number of times or more, the selected reference data is used as the evaluation data. The adopted reference data may be continuously adopted as evaluation data. This specific example will be described below.

  FIG. 8 is a diagram illustrating an example of the pitch of the singing voice of the singer x. The evaluation process of the singing voice of the singer x will be described with reference to FIG. In the example shown in FIG. 8, a song G having two types of singing parts (hereinafter “singing part a” and “singing part b”) is converted into two singers (hereinafter “singer x” and “singer y”). An example of the operation when singing will be described. First, it is assumed that singer x sings singing part a, while singer y sings singing part b. In FIG. 8, the horizontal axis indicates time and the vertical axis indicates pitch. In the figure, the solid line pa indicates the pitch of the singing part a, and the solid line pb indicates the pitch of the singing part b. A chain line px indicates the pitch of the singing voice of the singer x.

  The control unit 11 uses the singing pitch data of the singer x (hereinafter “singing pitch data SPx”), the reference pitch data of the singing part a (hereinafter “reference pitch data SPa”), and the reference pitch data of the singing part b (hereinafter “song pitch data SPa”). The reference pitch data SPb ") is compared with each note, and either the reference pitch data SPa or the reference pitch data SPb is selected for each note according to the comparison result. In the example shown in FIG. 8, “SPa”, “SPa”, “SPa”, “SPb”, “SPa”.

  When the same reference data is selected three or more times continuously, the control unit 11 employs the selected reference data as evaluation data. In this operation example, the reference data of the singing part a (hereinafter referred to as “reference data Ra”) is adopted as the evaluation data based on the selection result of the first three notes. Thereafter, in the fourth note, the reference pitch data SPb is selected, but since the reference pitch data SPb is not continuously selected three times or more, the control unit 11 continues the reference data of the singing part a. To use as evaluation data.

  Next, it is assumed that the singer x switches the singing from the singing part a to the singing part b. For example, a scene in which the singer y cannot understand the pitch of the singing part b and the singing is interrupted, and the singer x sings the singing part b instead of the singer y is assumed. The control unit 11 compares the singing pitch data SPx of the singer x with the reference pitch data SPa and the reference pitch data SPb for each note, and selects one for each note according to the comparison result. Therefore, in this operation example, the control unit 11 continuously selects the reference pitch data SPb from when the singing part is switched. The controller 11 switches the evaluation data for the singing voice x from the reference data Ra to the reference data Rb when the reference pitch data SPb is selected three or more times continuously.

(13) In the above-described embodiment, the karaoke apparatus 1 performs all the processes according to the present embodiment. On the other hand, two or more devices connected by a communication network may share the functions according to the above-described embodiment, and a system including the plurality of devices may realize the karaoke device 1 of the same embodiment. . For example, a computer device including a microphone, a speaker, a display device, an operation unit, and the like may be configured as a system in which a server device that executes reference data selection processing, evaluation processing, and the like is connected via a communication network. In this case, for example, the computer apparatus converts the sound collected by the microphone into singing voice data and transmits it to the server apparatus, and the server apparatus evaluates the received singing voice data and indicates the evaluation result. And the generated image data may be transmitted to a computer device.

(14) The program executed by the control unit 11 of the karaoke apparatus 1 in the above-described embodiment is recorded on a recording medium such as a magnetic tape, a magnetic disk, a flexible disk, an optical recording medium, a magneto-optical recording medium, RAM, or ROM. Can be provided in state. It is also possible to download to the karaoke apparatus 1 via a network such as the Internet.

(15) In the embodiment described above, the singing evaluation result is displayed on the display unit 13, but in addition to this, the skill of the singing singing technique is evaluated, and the evaluation result is also displayed on the display unit 13. You may do it. In this case, for example, the evaluation result to be displayed may be thinned out, such as displaying only the better evaluation. As another example of the evaluation result thinning-out process, for example, when the detection frequency of the singing technique is larger than a predetermined value, the detection result may be thinned out. Specifically, for example, the number of icons of the singing technique to be displayed may be adjusted so that the number to be displayed is within a predetermined range. FIG. 9 shows an example of the screen. In the screen illustrated in FIG. 9, icons I1, I2,... Indicating the singing technique used by the singer are displayed. The control unit 11 detects a singing technique such as “vibrato” or “fist” by analyzing the singing voice of the singer. Specifically, for example, the control unit 11 analyzes a pattern of temporal change in pitch detected from the singing voice data, and the pitch continuously fluctuates within a predetermined range above and below the central frequency. May be detected as a section in which the singing technique of “Vibrato” is used. For example, the control unit 11 may detect a section in which the short-term singing pitch is increased as a section in which the “fist” singing technique is used. The control unit 11 displays icons I1, I2,... Indicating the detection results on the display unit 13.

It is a block diagram which shows an example of the hardware constitutions of a karaoke apparatus. It is a figure which shows the structure of music data. It is a figure which shows the content of the reference data track contained in music data. It is a flowchart which shows the flow of the process which the control part of a karaoke apparatus performs. It is a figure which shows an example of the pitch of a singer's song voice. It is a figure which shows an example of the content of the table referred in the selection process of the data for evaluation. It is a figure which shows an example of the screen displayed on a display part. It is a figure which shows an example of the pitch of a singer's song voice. It is a figure which shows an example of the screen displayed on a display part.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Karaoke apparatus, 11 ... Control part, 12 ... Memory | storage part, 13 ... Display part, 14 ... Operation part, 16 ... Audio | voice processing part, 17 ... Speaker, 121 ... Music data storage area, 122 ... Background image data storage area, 151, 152 ... Microphones.

Claims (9)

Reference data storage means for storing a plurality of different reference data each composed of a row of notes;
First pitch detection means for detecting a pitch from the first audio signal;
The pitch detected by the first pitch detection means is compared with a plurality of reference data stored in the reference data storage means for each note, and any one of the plurality of reference data is calculated based on the difference. A first selection means for selecting for each note;
When the same reference data is selected continuously by the first selection means for a predetermined number of times or more, the selected reference data is adopted as the evaluation data. A first evaluation data adoption means for continuously adopting the adopted reference data as evaluation data;
First evaluation means for comparing evaluation data adopted by the first evaluation data adoption means and the first audio signal, and evaluating the first audio signal based on a comparison result;
And an output unit that outputs data indicating an evaluation result by the first evaluation unit. The first selection means compares the pitch detected by the first pitch detection means with a plurality of reference data stored in the reference data storage means for each note, and takes the reference data having the smallest difference as a note. The evaluation device according to claim 1, wherein the evaluation device is selected every time. Second pitch detecting means for detecting the pitch from the second audio signal;
Second evaluation data adoption means that employs any one of the plurality of reference data stored in the reference data storage means that is not adopted by the first evaluation data adoption means as evaluation data. When,
Second evaluation means for comparing the evaluation data adopted by the second evaluation data adoption means with the second audio signal and evaluating the second audio signal based on the comparison result. And
The evaluation apparatus according to claim 1, wherein the output unit outputs data indicating at least one of an evaluation result by the first evaluation unit and an evaluation result by the second evaluation unit. . Second pitch detecting means for detecting the pitch from the second audio signal;
The pitch detected by the second pitch detection means is compared for each note with a plurality of reference data stored in the reference data storage means, and any one of the plurality of reference data is calculated based on the difference. A second selection means for selecting for each note;
When the same reference data is selected continuously by the second selection means for a predetermined number of times or more, the selected data is adopted as the evaluation data, while in other cases, it is adopted immediately before Second evaluation data adoption means for continuously adopting the reference data as evaluation data;
Second evaluation means for comparing the evaluation data adopted by the second evaluation data adoption means with the second audio signal and evaluating the second audio signal based on the comparison result. ,
The evaluation apparatus according to claim 1, wherein the output unit outputs data indicating at least one of an evaluation result by the first evaluation unit and an evaluation result by the second evaluation unit. . An evaluation result selection means for selecting either the evaluation result by the first evaluation means or the evaluation result by the second evaluation means;
The evaluation apparatus according to claim 3 or 4, wherein the output means outputs data indicating an evaluation result selected by the evaluation result selection means. The evaluation apparatus according to claim 5, wherein the evaluation result selection unit selects a higher evaluation result from the evaluation result obtained by the first evaluation unit and the evaluation result obtained by the second evaluation unit. The output means calculates an average score between an evaluation result by the first evaluation means and an evaluation result by the second evaluation means, and outputs data indicating the calculation result. The evaluation apparatus as described in. The evaluation result selection means selects one of the evaluation result by the first evaluation means and the evaluation result by the second evaluation means for each note,
The evaluation apparatus according to claim 5, wherein the output unit outputs data indicating an evaluation result selected for each note by the evaluation result selection unit in units of notes. First power detection means for detecting the power of the audio signal from the first audio signal;
Second power detection means for detecting the power of the audio signal from the second audio signal,
The evaluation result selection means compares the power detected by the first power detection means with the power detected by the second detection means, and selects the evaluation result of the audio signal having the higher power. The evaluation apparatus according to claim 5.

Images