JP2013190564A - Voice evaluation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suitably evaluate singing in a singing method placing emphasis on inflection and rhythmic sense.SOLUTION: When playback of a karaoke musical piece ends, a control unit 10 performs evaluation processing of user's singing voice data stored in a user's singing voice data storage area 25. In sections other than sections of specific singing, the control unit 10 performs evaluation in accordance with a difference in pitch between user's singing voice data and GM data. Meanwhile, in the section of specific singing, the control unit 10 performs evaluation in accordance with a difference in variation between displacement points of pitch of the user's singing voice data and displacement points indicated by scoring data.

Description

  The present invention relates to a voice evaluation apparatus.

  Some karaoke apparatuses have a function of scoring the skill of singing by a singer. For example, Patent Document 1 discloses a method of scoring an audio signal input from a microphone in a plurality of modes based on different scoring standards.

JP 10-26992 A

By the way, there is a song that uses a singing method called rap. In laps, not the pitch value but emphasis and rhythmic feeling are emphasized, so scoring may be difficult with a conventional method such as the technique described in Patent Document 1.
This invention is made | formed in view of the above-mentioned background, and it aims at performing the evaluation of the singing in the singing method where emphasis and a feeling of rhythm are emphasized suitably.

  In order to solve the above-described problems, the present invention provides an audio acquisition unit that acquires audio data indicating a waveform of an audio, an exemplary sound acquisition unit that acquires exemplary sound data indicating characteristics of an exemplary sound, and the audio A feature specifying unit that specifies a feature of the sound indicated by the sound data acquired by the acquiring unit, a change amount of the feature specified by the feature specifying unit, and a change amount of the feature of the sound indicated by the exemplary sound data. A difference specifying unit that specifies a difference between each change amount, and an evaluation unit that evaluates a sound indicated by the audio data based on the difference specified by the difference specifying unit and outputs an evaluation result. A voice evaluation device is provided.

  In a preferred aspect of the present invention, when a point where a slope of a graph representing a temporal change in a sound feature changes more than a predetermined threshold is set as a displacement point, the displacement point of the feature specified by the feature specifying unit is specified. A displacement point specifying unit, a displacement point related to the feature indicated by the model sound data, and a displacement point of the audio data specified by the displacement point specifying unit appearing within a predetermined time difference from the time of the displacement point. An association unit for associating, wherein the difference identifying unit is configured to identify a feature change amount at a displacement point related to the audio data and a displacement point related to the exemplary sound data associated by the association unit in identifying the difference. You may specify the difference with the variation | change_quantity of a characteristic.

  Further, in a further preferred aspect of the present invention, there is provided a time displacement specifying unit that specifies a difference between the time of the displacement point related to the audio data and the time of the displacement point related to the exemplary sound data, and the evaluation unit The sound indicated by the audio data may be evaluated based on the difference specified by the specifying unit and the difference specified by the time displacement specifying unit, and the evaluation result may be output.

  Further, in a further preferred aspect of the present invention, the evaluation unit outputs the same evaluation result when the difference specified by the difference specifying unit is within a predetermined threshold, while in the other cases May output an evaluation result indicating a lower evaluation as the identified difference is larger.

  In a further preferred aspect of the present invention, the audio data includes a section data acquisition unit that acquires section data representing a section in which singing is performed in a predetermined singing mode, and the evaluation unit is acquired by the section data acquisition unit. In the section indicated by the section data, the evaluation based on the difference specified by the difference specifying unit is performed or the conventional singing evaluation based on the pitch is performed, while the section other than the section indicated by the section data is performed. The evaluation may be performed based on the difference between the feature specified by the feature specifying unit and the sound feature indicated by the model sound data.

  Further, in another preferred aspect of the present invention, the audio data is analyzed according to a predetermined algorithm, and a section specifying unit for specifying a section for performing singing according to a predetermined singing mode according to the analysis result is provided. In the section specified by the section specifying section, the evaluation section performs the evaluation based on the difference specified by the difference specifying section or performs the conventional singing evaluation mainly on the pitch while the specified section is specified. In a section other than the section, evaluation based on the difference between the feature specified by the feature specifying unit and the sound feature indicated by the model sound data may be performed.

  Further, the present invention provides a voice acquisition unit that acquires voice data indicating a waveform of a voice, a feature specification unit that specifies a feature of a sound indicated by the voice data acquired by the voice acquisition unit, and the feature specification unit. A peak value specifying unit for specifying a plurality of peak values appearing in a graph representing a temporal change of the specified feature, a change amount specifying unit for specifying a change amount of the peak value specified by the peak value specifying unit, and the change There is provided an audio evaluation apparatus comprising: an evaluation unit that evaluates a sound indicated by the audio data based on a change amount specified by an amount specifying unit and outputs an evaluation result.

Further, the present invention is indicated by an audio acquisition unit that acquires audio data representing an audio waveform, a beat data acquisition unit that acquires beat data indicating the beat of the music, and audio data acquired by the audio acquisition unit. A feature specifying unit for specifying a feature of the sound, and a displacement point specifying unit for specifying a location where a slope of a graph representing a temporal change of the feature specified by the feature specifying unit changes by a predetermined threshold or more as a displacement point; Evaluation that evaluates the sound indicated by the audio data based on the time difference between the displacement point specified by the displacement point specifying unit and the time indicated by the beat data acquired by the beat data acquisition unit, and outputs an evaluation result A speech evaluation apparatus characterized by comprising:
Further, in this aspect, a difference specifying unit for specifying a difference between the change amount of the feature specified by the feature specifying unit and the change amount of the sound feature indicated by the model sound data representing the feature of the model sound. And the evaluation unit performs an evaluation based on a time difference between the displacement point and the time indicated by the beat data, and evaluates a sound indicated by the audio data based on a difference specified by the difference specifying unit. Also good.

  According to the present invention, it is possible to suitably perform singing evaluation in a singing method in which inflection and rhythmic feeling are important.

System configuration diagram of an embodiment of the present invention Block diagram showing hardware configuration of karaoke equipment Schematic diagram showing the contents of the accompaniment data storage area Diagram showing an example of the contents of scoring data The figure which shows an example of the contents of the displacement point data Block diagram showing an example of the functional configuration of a karaoke device Block diagram showing an example of the functional configuration of the scoring unit The figure for demonstrating the content of the process which a pitch comparison part performs The figure for demonstrating the content of the process which a pitch comparison part performs Flow chart showing the flow of processing performed by the control unit Flow chart showing the flow of scoring process performed by the control unit The figure for explaining the contents of the pitch comparison processing The figure for explaining the contents of the pitch comparison processing The figure for explaining the contents of the pitch comparison processing The figure for explaining the contents of the score calculation processing The figure for explaining the contents of the pitch comparison processing Figure showing a histogram of displacement points The figure for explaining the contents of the pitch comparison processing The figure which shows an example of the change of the pitch of singing voice

<Embodiment>
<Configuration>
FIG. 1 is a diagram showing the configuration of a system according to an embodiment of the present invention. This system includes a karaoke device 100, a server device 200, and a network NW. The karaoke device 100 is a device that reproduces karaoke music according to a request from the user and evaluates the singing by the user about the karaoke music to be played. The network NW is a LAN (Local Area Network) or the Internet, and is a communication network in which data communication is performed between the karaoke apparatus 100 and the server apparatus 200. The server device 200 stores various data such as content data related to karaoke music in a storage means such as an HDD (Hard Disk Drive) provided inside or outside the server device 200, and via the network NW according to a request from the karaoke device 100 The content data is supplied to the karaoke apparatus 100. Here, the content refers to a combination of audio and video related to karaoke music. That is, the content data is composed of so-called accompaniment data composed of accompaniment and chorus without the main melodic singing voice, and video data composed of the lyrics displayed on the background of the music and the lyrics. Note that a plurality of karaoke apparatuses 100 may exist for the server apparatus 200. Further, a plurality of server devices 200 may exist for the karaoke device 100.

  FIG. 2 is a block diagram showing the hardware configuration of the karaoke apparatus 100. The karaoke apparatus 100 includes a control unit 10, a storage unit 20, an operation unit 30, a display unit 40, a communication control unit 50, an audio processing unit 60, a microphone 61, and a speaker 62, and these units are connected via a bus 70. Has been. The control unit 10 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), and the like. In the control unit 10, the CPU controls each unit of the karaoke apparatus 100 by reading out a computer program stored in the ROM or the storage unit 20, loading it into the RAM, and executing it.

  The operation unit 30 includes various operators and outputs an operation signal representing the content of an operation performed by the user to the control unit 10. The display unit 40 includes, for example, a liquid crystal panel, and displays various images such as lyrics telop and background video according to each karaoke piece under the control of the control unit 10. The communication control unit 50 connects the karaoke device 100 and the network NW by wire or wirelessly, and controls data communication between the karaoke device 100 and the server device 200 via the network NW.

  The server device 200 is a computer including a CPU and various memories (not shown), and particularly includes a network storage 210. The network storage 210 is an HDD, for example, and stores various data such as content data of karaoke music. In FIG. 2, the server device 200 includes one network storage 210, but the number of network storages is not limited to this, and the server device 200 may include a plurality of network storages. When content data of a karaoke song reserved by the user is stored in the network storage 210, the karaoke device 100 communicates with the server device 200 according to control by the communication control unit 50, and content read from the network storage 210 While downloading data via the network NW, streaming playback is performed in which playback is performed sequentially from the part where the download is completed.

  The microphone 61 outputs an analog audio signal representing the collected audio to the audio processing unit 60. The audio processing unit 60 includes an A / D (Analog / Digital) converter, converts the analog audio signal output from the microphone 61 into digital audio data, and outputs the digital audio data to the control unit 10. get. The audio processing unit 60 includes a D / A (Digital / Analog) converter, converts digital audio data received from the control unit 10 into an analog audio signal, and outputs the analog audio signal to the speaker 62. The speaker 62 emits a sound based on the analog audio signal received from the audio processing unit 60. In this embodiment, a case where the microphone 61 and the speaker 62 are included in the karaoke apparatus 100 will be described. However, the audio processing unit 60 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 61 to the speaker 62 is an analog audio signal. However, digital audio data may be input / output. In such a case, the audio processing unit 60 does not need to perform A / D conversion or D / A conversion. The same applies to the operation unit 30 and the display unit 40, and an external output terminal may be provided to connect an external monitor.

  The storage unit 20 is a storage unit for storing various data, and is, for example, an HDD or a nonvolatile memory. The storage unit 20 includes a plurality of storage areas such as an accompaniment data storage area 21, a video data storage area 22, a GM (Guide Melody) data storage area 23, a scoring data storage area 24, and a user singing voice data storage area 25. Yes.

  FIG. 3 is a schematic diagram showing the contents of the accompaniment data storage area 21. The accompaniment data storage area 21 stores information related to accompaniment data representing accompaniment sound in each music piece. The accompaniment data storage area 21 stores a plurality of accompaniment data records including a plurality of items such as “song number”, “song name”, “singer name”, “genre”, and “file storage location”. “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. The “file storage location” represents a storage location of a data file that is accompaniment data itself of each music piece. When a folder called server1 is included, a data file of accompaniment data is stored in the server device 200. If not included, it means that a data file of accompaniment data is stored in the karaoke apparatus 100. For example, in FIG. 3, a song whose song name is “BBB” indicates that a data file of accompaniment data is stored in the server device 200, and a song whose song name is “CCC” has a data file of accompaniment data as karaoke. It is stored in the storage unit 20 of the device 100. The accompaniment data file is, for example, a MIDI (Musical Instrument Digital Interface) format file.

  The video data storage area 22 stores lyrics data indicating the lyrics of each song and background video data representing a background video displayed on the background of the lyrics. The lyrics shown by the lyrics data are displayed on the display unit 40 as lyrics telop as the music progresses during karaoke singing. In addition, the background video represented by the background video data is displayed on the display unit 40 as the background of the lyrics telop as the music progresses during karaoke singing. In the GM data storage area 23, data indicating the melody of the vocal part of the music, that is, guide melody data (hereinafter referred to as “GM data”), which is data specifying the content of the constituent sound to be sung, is stored. The GM data is obtained when the control unit 10 performs the skill evaluation process of the singing by the user in a section other than the section in which the control unit 10 performs singing by rap or singing by imitation (hereinafter referred to as “specific singing”). It is used as a reference for comparison. Since the evaluation process performed by the control unit 10 will be described later, a detailed description thereof will be omitted here. The GM data is described in the MIDI format, for example.

  The scoring data storage area 24 stores data (hereinafter referred to as “scoring data”) used for scoring a section where a specific song is performed (hereinafter referred to as “specific section”). FIG. 4 is a diagram showing an example of the contents of scoring data. As shown in the figure, the scoring data includes items of “song number”, “specific section data”, “pitch displacement point data”, and “volume displacement point data”. Among these items, the “song number” is as described above. “Specific section data” is data indicating a specific section. In the example shown in FIG. 4, it is indicated that the music piece with the “song number” “1004-19” has two sections, a section from time t11 to t20 and a section from time t21 to time t30. ing. The number of specific sections included in one musical piece may be one or plural. In the case of a music piece that does not have a specific section, scoring data is not stored.

  In FIG. 4, “pitch displacement point data” is data generated from model sound data (for example, GM data) representing a model sound, and a tendency of a pitch change in the model sound data (in a graph representing a pitch). Time data indicating the time at which (tilt) changes, and pitch data indicating the value of the pitch at that time. In the example shown in FIG. 4, (time, pitch) = (t11, p11), (t12, p12),... In a specific section from time t11 to t20 of the music whose “song number” is “1004-19”. The plurality of locations shown are shown as pitch displacement points. The model sound data is not limited to GM data, and may be, for example, data representing a typical singing voice, or any data representing a model sound. The “volume displacement point data” is data generated from the model sound data, and the time data indicating the time when the tendency of the volume change (slope of the graph representing the volume) changes in the model sound data, and the volume of the sound at that time Volume data indicating a value. In the following description, when it is not necessary to distinguish the pitch displacement point data and the volume displacement point data from each other for convenience of explanation, these will be referred to as “displacement point data”. That is, the displacement point data is data indicating a location where the slope of the graph representing the temporal change in pitch (or volume) changes by a predetermined threshold value before and after the displacement.

  FIG. 5 is a diagram illustrating an example of the contents of the displacement point data. In the figure, the horizontal axis indicates time, and the vertical axis indicates pitch (or volume). A solid line 500 represents a change in the pitch (or volume) of the guide melody represented by the GM data, and is hereinafter referred to as a GM curve 500. In this embodiment, as a point at which the slope of the GM curve 500 changes greatly (hereinafter referred to as “displacement point”), a point where the pitch (or volume) has changed from an increase to a decrease (see, for example, time t19 in FIG. 5), an increase. Points that fall within a certain range, points that begin to rise (for example, see time t14), points that the pitch (or volume) starts to rise from descending (for example, see time t17), and that fall within a certain range that stops. The amount of change in the slope of the graph (GM curve 500) representing the pitch (or volume), such as a point (see, for example, time t13) or a point that starts to fall (see, for example, time t11), is equal to or greater than a predetermined threshold. Use a different position. In this embodiment, in addition to the location where the amount of change in the slope of the GM curve 500 is greater than or equal to a predetermined threshold, the location where the pitch detection at the start of singing is started (such as time t11 in FIG. 5) is also the displacement point. Used as In addition, it is good also as a structure which does not use as a displacement point not only this but the location where the detection of the pitch at the time of a singing start is started.

  In the user singing voice data storage area 25, the user's singing voice collected by the microphone 61 during the period in which the accompaniment data is being reproduced for each piece of karaoke music is converted into digital data by the voice processing unit 60. Audio data generated by the conversion is stored. This voice data is called user singing voice data. The user singing voice data is voice data representing a voice waveform, and is stored as a data file in a WAVE (RIFF waveform Audio Format) format, for example. The user singing voice data for each song is associated with the GM data of the song by the control unit 10.

  FIG. 6 is a block diagram illustrating an example of a functional configuration of the karaoke apparatus 100. In FIG. 6, the reproduction unit 11 and the scoring unit 12 are realized by the CPU of the control unit 10 reading a computer program stored in the ROM or the storage unit 20 and loading it into the RAM for execution. The reproducing unit 11 reproduces karaoke music. Specifically, the playback unit 11 causes the speaker 62 to emit sound based on the accompaniment data and the GM data, and causes the display unit 40 to display a video based on the video data.

  The scoring unit 12 scores data representing the singing voice of the singer (hereinafter referred to as “user singing voice data”). The scoring unit 12 determines whether the section being sung is a section where a specific song is performed (hereinafter referred to as a “specific section”) or another section (hereinafter referred to as a “standard section”), and the scoring data is used in the specific section. On the other hand, scoring is performed using GM data in the standard section. More specifically, the scoring unit 12 evaluates the singing according to the difference between the pitch of the singing voice and the pitch of the GM data in the standard section, while the amount of change in the singing pitch and the GM data in the specific section. Evaluation is performed such that the smaller the difference from the change amount of the pitch, the higher the evaluation.

  FIG. 7 is a block diagram illustrating an example of a functional configuration of the scoring unit 12. In FIG. 7, the pitch specifying unit 121 functions as a voice acquisition unit that acquires the user singing voice data stored in the user singing voice data storage area 25, analyzes the acquired user singing voice data, and analyzes the user singing voice data. It functions as a pitch specifying unit that specifies the pitch of the sound indicated by. The pitch specifying unit 121 outputs data representing the specified pitch (hereinafter referred to as “pitch data”) to the section determining unit 123. The volume specifying unit 122 specifies the volume of the user singing voice data stored in the user singing voice data storage area 25. The volume specifying unit 122 outputs data representing the specified volume (hereinafter referred to as “volume data”) to the section determining unit 123.

  The section determination unit 123 refers to the section data stored in the scoring data storage area 24 and determines whether or not the acquired user singing voice data is a specific section. If it is determined that the section is a specific section, the section determining unit 123 outputs the pitch data acquired from the pitch specifying unit 121 to the pitch displacement point specifying unit 124. On the other hand, in other cases, the section determination unit 123 outputs the pitch data acquired from the pitch identification unit 121 to the pitch comparison unit 126. Further, when it is determined that the section is a specific section, the section determining unit 123 outputs the volume data acquired from the volume specifying section 122 to the volume displacement point specifying section 125. On the other hand, in other cases, the section determination unit 123 outputs the volume data acquired from the volume identification unit 122 to the volume comparison unit 127.

  The pitch displacement point identifying unit 124 identifies, as a displacement point, a point where the slope of the graph representing the temporal change in the pitch represented by the pitch data supplied from the section determining unit 123 changes more than a predetermined threshold before and after that. . That is, the pitch displacement point specifying unit 124 specifies a time at which the amount of change in the slope of the graph represented by the pitch data supplied from the section determining unit 123 is equal to or greater than a predetermined threshold, and sets the pitch value at that time. Identify. The slope of the graph represented by the pitch data is obtained as follows, for example. The pitch displacement point specifying unit 124 may obtain an inclination from adjacent samples, or may obtain an inclination from an approximate curve of a plurality of samples. Further, the pitch displacement point specifying unit 124 may apply an LPF (low pass filter) to a column of slopes obtained from adjacent samples. In addition, the pitch displacement point specifying unit 124 may apply LPF to the sample before calculating the inclination. Further, the pitch displacement point specifying unit 124 may obtain the slope of the tangent by differentiating one point at a time. The pitch displacement point identification unit 124 outputs pitch displacement point data representing the identified time and pitch to the pitch comparison unit 126.

  The volume displacement point identification unit 125 identifies the volume displacement point in the user singing voice data from the volume data supplied from the section determination unit 123. That is, the volume displacement point specifying unit 125 specifies a time at which the slope of the graph represented by the volume data supplied from the section determining unit 123 changes by a predetermined threshold before and after that, and sets the volume value at that time. Identify. The volume displacement point identification unit 125 outputs volume displacement point data representing the identified time and volume to the volume comparison unit 127.

  The pitch comparison unit 126 performs different processing in the standard section and the specific section. In the standard section, the pitch comparison unit 126 acquires pitch data of the user singing voice output from the section determination unit 123 and GM data corresponding to the user singing voice. The GM data is a reference for comparison when the control unit 10 evaluates the skill of singing by the user, and is a reference that is predetermined for the song to be sung. The control unit 10 associates the user singing voice data with the GM data in the time axis direction, compares the pitch of the user singing voice data with the pitch of the GM data according to the correspondence result, and compares the result data representing the difference between the two. Is generated.

  On the other hand, in the specific section, the pitch comparison unit 126 acquires pitch displacement point data representing the pitch at the displacement point of the model sound data from the scoring data storage area 24. This pitch displacement point data is an example of model sound data indicating the pitch of a model sound. Further, the pitch comparison unit 126 is within a predetermined time difference from the displacement point indicated by the pitch displacement point data acquired from the scoring data storage area 24 (that is, the displacement point relating to the pitch indicated by the model sound data) and the time of this displacement point. And the displacement point indicated by the pitch displacement point data output from the pitch displacement point specifying unit 124 (that is, the displacement point relating to the pitch indicated by the user singing voice data) are associated with each other and related to the associated user singing voice data The difference between the change amount of the pitch at the displacement point and the change amount of the pitch at the displacement point related to the model sound data is specified. The association between the displacement points is performed as follows, for example. The pitch comparison unit 126 obtains a displacement point from the pitch sequence of the user singing voice data and the pitch sequence of the exemplary sound data. The temporal position of the displacement point is determined by the time when the beginning position of the song is 0 (zero). The pitch comparison unit 126 searches for a displacement point obtained from the pitch sequence of the user singing voice data in the vicinity of the displacement point obtained from the pitch sequence of the model sound data. As the vicinity of the displacement point, the range may be determined by time, for example, within 1 second before and after the displacement point, or may depend on the tempo, such as one beat. When there is no displacement point in the vicinity, it is assumed that there is no displacement point corresponding to the displacement point of the model sound data. On the other hand, when there is only one displacement point in the vicinity, the pitch comparison unit 126 assumes that the displacement point is a corresponding displacement point. In addition, when there are two or more displacement points in the vicinity, the pitch comparison unit 126 employs a plurality of displacement points that are temporally close as the corresponding displacement points. Note that the pitch comparison unit 126 compares the displacement points having the same pitch inclination change mode (for example, turning from rising to falling, turning from falling to rising, etc.).

The pitch comparison unit 126 generates comparison result data representing the specified difference. The pitch comparison unit 126 corresponds to an example of an association unit and a difference specifying unit according to the present invention. In this embodiment, the pitch comparison unit 126 performs the following processing to generate comparison result data. 8 and 9 are diagrams for explaining the processing contents of the pitch comparison unit 126 in the specific section. FIG. 8 is a diagram showing the contents of the calculation process of the difference in time variation at the displacement point, and FIG. 9 is the diagram showing the contents of the calculation process of the difference in pitch variation at the displacement point. 8 and 9, the horizontal axis indicates time, and the vertical axis indicates pitch. The GM curve 500 is similar to that shown in FIG. A solid line 300 represents a change in the pitch of the voice when the user sings represented by the user singing voice data, and is hereinafter referred to as a user singing voice curve 300. First, the pitch comparison unit 126 includes pitch displacement point data (gt _i , gp _i ) at the displacement point X _i (i is an integer from 1 to n (n> 1)) included in the scoring data, and immediately before that. Change amount (Δgt _i , Δgp _i ) = (gt _i −gt _i−1 , gp _i −gp _i− ) with respect to the pitch displacement point data (gt _i−1 , gp _i−1 ) at the displacement point X _i−1 . ₁ ) Calculate. Similarly, pitch comparison section 126, the user pitch displacement point data (ut _i, up _i) in displacement point UX _i of the singing voice data and, pitch displacement point data in displacement point UX _i-1 of the immediately preceding (ut _{i- 1} , up _i-1 ) (Δut _i , Δup _i ) = (ut _i −ut _i−1 , up _i −up _i−1 ) is calculated.

Next, the pitch comparison unit 126 calculates the amount of change (Δgt _i , Δgp _i ) obtained for each displacement point of the scoring data and the amount of change (Δut _i , Δup _i ) obtained for each displacement point of the user singing voice data. the difference value _{(Δt i, Δp i) =} (Δgt i -Δut i, Δgp i -Δup i) is calculated, the difference value Delta] t _i of time representing the calculation result, and the difference value Delta] p _i of the pitch, comparison Output as result data.

  The volume comparison unit 127 performs different processing in the standard section and the specific section. In the standard interval, the volume comparison unit 127 compares the volume of the user singing voice data stored in the user singing voice data storage area 25 with a predetermined volume reference value, and provides comparison result data representing the difference between the two. Is generated.

On the other hand, in the specific section, the volume comparison unit 127 uses the volume displacement point data included in the scoring data and the volume displacement point data of the user singing voice data, and provides comparison result data representing the difference between the changes in both. Generate. This comparison process is the same as the pitch comparison process performed by the pitch comparison unit 126. That is, the volume comparison unit 127 performs the volume displacement point data (gt _i , gv _i ) at the displacement point X _i for the volume displacement point data for each displacement point included in the scoring data stored in the scoring data storage area 24. And the amount of change (Δgt _i , Δgv _i ) between the volume displacement point data (gt _i-1 , gv _i-1 ) at the displacement point X _i-1 immediately before that and the displacement point of the user singing voice data The same processing is performed to calculate the amount of change (Δut _i , uv _i ), and the difference value (Δt _i , Δv _i ) = (Δgt _i −Δut _i , Δgv _i −Δuv _i ) is compared with the comparison result data Output as.

  The scoring output unit 128 performs singing voice evaluation processing based on the comparison result data output from the pitch comparison unit 126 and the comparison result data output from the volume comparison unit 127, and the evaluation result is displayed on the display unit 40 or the like. Output to. The scoring output unit 128 acquires specific section data representing the specific section from the scoring data storage area 24, and in the specific section indicated by the acquired specific section data, the pitch change amount of the singing voice data and the model sound data While the evaluation process is performed so that the smaller the difference from the change amount of the pitch, the higher the evaluation, the user singing voice is calculated based on the difference between the pitch of the user singing voice data and the pitch of the GM data in the sections other than the specific section. Evaluate the data. More specifically, for example, in the standard section, the scoring output unit 128 compares the change in the pitch of the voice indicated by the user singing voice data with the change in the pitch of the guide melody indicated by the GM data, and matches them. An evaluation value indicating the degree of is calculated. The evaluation value is 100% (that is, there is no deduction) if the difference between the two pitches is within a predetermined allowable range in a certain note, and the period during which the difference between the two pitches does not fall within the above range. However, in the case of GM data, it may be 50% if the note length is half of the note length in this note. That is, in a certain note, a value obtained by dividing a period in which the difference in pitch between the two is within the above range by the sound length in this note in GM data is used as an evaluation value. The control unit 10 determines a point to be deducted based on the calculated evaluation value. For example, when “2 points” is assigned to a certain note and the evaluation value is calculated as 50%, the control unit 10 determines “1 point” as a deduction point.

  The evaluation value is, for example, 100% (that is, no deduction point) if the difference between the two pitches is within a predetermined allowable range in a certain note, and the difference between the two pitches is not within the above range. For example, the period may be 50% if the period is half of the note length in this note in the GM data. Note that the volume reference value may be set for each note included in the GM, or may be set for each predetermined section such as for each measure.

On the other hand, in the specific section, the scoring output unit 128 may score, for example, for each displacement point of the model sound data, and obtain the average after the end of the specific section. More specifically, for example, as scoring output unit 128 calculates the deviation of x _i by the following equation (1), the number the higher the average value of deviation of x _i of all the displacement point smaller increases Scoring may be performed. In the following equation (1), Δt _i represents a time difference between pitch displacement points (difference value Δt _i calculated by the pitch comparison unit 126), and Δp _i represents a pitch difference between displacement points (calculated by the pitch comparison unit 126). Difference value Δp _i ) is shown. Α and β are weighting coefficients.
x _i = | Δt _i | * α + | Δp _i | * β (1)

In addition, the aspect of scoring is not limited to the above, but may be other aspects. For example, the scoring output unit 128 scores the deviation of the time difference and the deviation of the pitch difference for each displacement point, calculates the score s _i for each displacement point by the following equation (2), and scores all the displacement points Scoring may be performed such that the higher the average of, the higher the score. In the following equation (2), st _i denotes the score based on the deviation of the time difference in the displacement point X _i, sp _i denotes the score based on the deviation of the pitch difference in displacement point X _i.
s _i = st _i * α + sp _i * β (2)

<Operation>
FIG. 10 is a flowchart showing the flow of processing performed by the control unit 10. When the music is reserved by the user via the operation unit 30 (step S100; Yes), the control unit 10 searches for the reserved music from the storage unit 20 (step S102). Specifically, in step S102, the control unit 10 uses the song number or song name of the selected song from each of the accompaniment data storage area 21, the video data storage area 22, and the GM data storage area 23 as a key. Data related to the music is searched, and the search result data is read into the RAM.

  Next, the control unit 10 reproduces karaoke music based on the accompaniment data, video data, and GM data stored in the RAM (step S104). Specifically, in step S104, the control unit 10 causes the speaker 62 to emit sound based on the accompaniment data and the GM data, and causes the display unit 40 to display a video based on the video data. And the control part 10 memorize | stores in the user song voice data storage area 25 the user song voice data by which the user's song voice picked up by the microphone 61 is converted into digital data by the voice processing part 60 ( Step S106). When the reproduction of the karaoke music is completed, the control unit 10 scores the singing based on the user singing voice data, the GM data, and the scoring data stored in the user singing voice data storage area 25 (step S108). And the control part 10 displays a scoring result on the display part 40 (step S110).

  FIG. 11 is a flowchart showing the flow of the scoring process (step S108 in FIG. 10) performed by the control unit 10. First, the control part 10 specifies the pitch of the sound shown by user song voice data (step S200). Next, the control unit 10 specifies the pitch displacement point from the user singing voice data in the specific section (step S210). Next, the control unit 10 scores the user singing voice by performing the following steps S220 to S250 for each predetermined unit section. First, the control unit 10 determines whether the scoring target is a standard section or a specific section (step S220). When it is a standard section (step S220; NO), the control unit 10 compares the pitch of the user singing voice data with the pitch of the GM data, calculates the evaluation value according to the difference between the two, and the user singing voice. The volume of the data is compared with a predetermined volume reference value, and an evaluation value is calculated according to the difference between the two (step S230). On the other hand, when it is a specific section (step S220; NO), the control unit 10 changes the data representing the pitch displacement point specified from the user singing voice data and the pitch stored in the scoring data storage area 24. The change amount of the displacement point data is compared, an evaluation value corresponding to the difference between the two change amounts is calculated, and data representing the volume displacement point specified from the user singing voice data and stored in the scoring data storage area 24 The obtained volume displacement point data is compared, and an evaluation value corresponding to the difference between the two changes is calculated (step S240).

  The control unit 10 determines whether or not to end the process by determining whether there is an unscored section (step S250), and when there is a section to be scored (step S250; NO), While returning to step S220 and scoring the next section, if it is determined that the music has been scored to the end (step S250; YES), the scoring process is terminated.

  By the way, in rap singing, it is sung in a rhythmic manner without much melody, while rhyming at the end of a measure. For this reason, in scoring laps, the degree of coincidence of pitch is not so important, and intonation and rhythm are emphasized. In general singing grading, the goal is to sing a song sung on a melody, so the absolute value of the pitch is emphasized. On the other hand, in the lap, since the emphasis is not on the pitch value but on the inflection and rhythm, it is difficult to score with the conventional method. In this embodiment, scoring is performed according to the difference between the amount of change in the pitch of the user singing voice data and the amount of change in the pitch of the GM data. Can be scored.

<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, different evaluation processes are performed in the specific section and other sections with reference to the section data. However, the present invention is not limited to this, and the process for determining whether or not the specific section is performed is not performed. May be. In this case, scoring using the above scoring data may be performed in all sections of the music. More specifically, for example, for a music piece whose genre is “rap”, a scoring process using scoring data may be performed in all sections of the music piece. According to this aspect, the control unit 10 does not need to switch the scoring process and does not need to indicate a specific section in the music data.

<Modification 2>
In the above-described embodiment, the control unit 10 identifies the pitch displacement point and the volume displacement point from the user singing voice data, and compares the difference in the amount of change between adjacent displacement points between the user singing voice data and the scoring data. did. The aspect of the comparison process between the user singing voice data and the scoring data is not limited to that described above, and may be, for example, the following process. First, the control unit 10 calculates the amount of change in the pitch of the user singing voice data at a predetermined time interval and sets the amount of change in the pitch of the model sound data (for example, GM data) at the predetermined time interval. Calculate with Next, the control unit 10 compares the calculated amount of change in the pitch of the user singing voice data with the amount of change in the pitch of the model sound data, and calculates the difference between the two. In the example shown in FIG. 12, the difference (b−a), (c−b), (dc),... Of the pitch of the model sound data at regular intervals, and the pitch of the user singing voice data at regular intervals. The differences (b′−a ′), (c′−b ′), (d′−c ′),... Are compared, and the difference between them is calculated. It can be said that the smaller the difference obtained by this calculation process, the smaller the difference between the pitch of the user singing voice data and the pitch of the model sound data. Therefore, the evaluation process similar to the above-described embodiment is performed by performing an evaluation process in which the control unit 10 performs a higher evaluation as the calculated difference is smaller. Thus, instead of scoring according to the difference in the amount of change in pitch at the displacement point, the control unit 10 is predetermined with the pitch of the user singing voice data and the pitch of the sound indicated by the model sound data. Alternatively, the evaluation processing may be performed based on the difference between the two amounts of change. Also in this aspect, as in the above-described embodiment, singing by a singing method in which emphasis is placed on intonation and rhythmic feeling can be suitably evaluated. In short, the control unit 10 evaluates the sound indicated by the user singing voice data based on the difference between the change amount of the pitch of the user singing voice data and the change amount of the pitch of the sound indicated by the model sound data. Can be output.

<Modification 3>
Moreover, the following processes may be sufficient as the comparison process of user song voice data and the data for scoring. First, the control unit 10 obtains the slope of the GM curve 500, and divides the specific section into a plurality of sections according to the range of the slope value. For example, the control unit 10 may use a section having a positive slope (that is, a section where the pitch is increasing), a section having a negative value (that is, a section where the pitch is decreasing), zero (or preliminarily close to zero). You may divide | segment into the area (namely, area with little change of a pitch) which is a value which is in the defined threshold value. Similarly, the control part 10 calculates | requires the inclination of the user singing voice curve 300, and divides | segments a specific area into a some area according to the range of the value of this inclination. Next, the control unit 10 makes a high evaluation for a section in which the pitch change mode is the same (for example, a section in which both slopes are positive values) (see section A1 in FIG. 13), while the change in pitch changes. An evaluation process is performed so that a section having a different aspect (for example, a section in which one slope value is positive while the other slope value is negative) (see section A2 in FIG. 13) has a low evaluation. It may be. Since it can be said that the difference between the amount of change in the pitch of the user singing voice data and the amount of change in the pitch of the model sound data is large in the sections having different modes of change in the pitch, this evaluation process is similar to the above-described embodiment. The evaluation processing is performed such that the lower the difference between the pitch change amount of the singing voice data and the pitch change amount of the model sound data, the lower the evaluation. Therefore, also in this aspect, the singing by the singing method where emphasis is placed on the intonation and the rhythmic feeling can be suitably evaluated. Moreover, according to this aspect, it is not necessary to specify a displacement point when evaluating a song.

<Modification 4>
Moreover, the following processes may be sufficient as the comparison process of user song voice data and the data for scoring. First, the control part 10 specifies a pitch from user song voice data, and calculates the average value of a pitch. Further, the control unit 10 calculates an average value of the pitch of the model sound data, and a section where the difference between the pitch value and the calculated average value is equal to or greater than a predetermined threshold (sections A11, A12,... Reference section) is generated. The control unit 10 is not limited to generating the section data, and the section data may be previously included in the scoring data and stored in the scoring data storage area 24. The control unit 10 identifies a section (see sections A21, A22,... In FIG. 14) in which the difference between the pitch of the user singing voice data and the average value is equal to or greater than a predetermined threshold, and the identified section and the section data. The section shown is compared, and the evaluation process is performed so that the larger the overlapping part is, the higher the evaluation is. In this aspect, since the evaluation is performed according to the overlap amount of the section in which the difference from the average value of the pitch is equal to or greater than the threshold value, the singing by the singing method in which the inflection and the sense of rhythm are emphasized can be scored suitably.

  As another example, for example, the control unit 10 acquires a pitch at a predetermined time interval (for example, about 500 ms) in the user singing voice data, and calculates a total value of absolute values of a difference in pitch between adjacent samples. In addition, the absolute value of the difference between adjacent pitches when pitches are acquired at similar time intervals in the model sound data is calculated, and the total value in the user singing voice data is compared with the total value in the model sound data May be. In this case, the evaluation process may be performed such that the smaller the difference between the total value calculated from the user singing voice data and the total value calculated from the model sound data, the higher the evaluation. According to this aspect, the evaluation can be performed using the pitch of the model sound data, and the calculation amount required for the evaluation process can be reduced.

<Modification 5>
In the above-described embodiment, the control unit 10 performs scoring for each displacement point using the above-described equation (1). However, the scoring process is not limited to this, and for example, a scoring function as shown in FIG. The scoring value may be calculated using 400. FIG. 15 is a diagram showing the content of the score calculation process per displacement point, the horizontal axis shows the difference in the amount of change in time (or pitch) between the user singing voice data and the scoring data, and the vertical axis Indicates a score. In the example shown in FIG. 15, the control unit 10 calculates a score so that if the difference in the amount of change is within a certain range, the score will be full, the score will decrease thereafter, and the deviation over a certain amount will be the lowest score. To do. That is, the control unit 10 outputs the same evaluation result when the difference between the change amount of the user singing voice data and the change amount of the scoring data is within a predetermined threshold value, but otherwise Outputs a lower evaluation result as the difference is larger. According to this aspect, since it is possible to adopt a calculation method that allows a certain amount of deviation, a scoring result closer to hearing is obtained.

<Modification 6>
In the above-described embodiment, scoring is performed for each specific point by scoring for each displacement point and obtaining the average value of the scoring values for each displacement point. However, the scoring mode for the specific region is limited to this. is not. For example, if there is a pitch displacement point at a certain time in the scoring data, a high score will be obtained if there is a singing pitch displacement point in the immediate vicinity, and the score will decrease as the distance increases. May be. Further, for example, the control unit 10 may deduct points when the singing pitch takes a displacement point at a time when there is no displacement point of the scoring data pitch.

As another example, for example, the control unit 10 may perform statistical processing while viewing the entire specific section. More specifically, for example, the average and deviation of the difference between the time of the displacement point indicated by the scoring data and the time of the singing displacement point (see Δt50 in FIG. 16) are obtained. Higher scores may be obtained as the deviation is smaller. At each displacement point, the time difference between the displacement points is calculated between the scoring data and the user singing voice data, and the closer the mean of the time difference is to near zero and the closer the deviation is to zero, the more indicated by the scoring data. Singing the model singing. Therefore, the control unit 10 may calculate the score score by the following equation (3). In the following equation (3), A is an average value, D is a deviation value, and a, b, and c are coefficients.
score = aA + bD + c (3)

  In this aspect, the control unit 10 identifies a specific displacement point (for example, a point having a maximum value) by performing processing such as taking a histogram (see FIG. 17), and assigns a weight to the identified displacement point. May be performed. According to this aspect, the time shift is emphasized at the specific displacement point, and scoring closer to the feeling of listening is performed.

<Modification 7>
In the above-described embodiment, the control unit 10 performs the scoring using the scoring data in the specific section. However, the scoring may be performed without using the scoring data. In this case, for example, the evaluation process is performed using the difference between the time of the beat included in the music information (or the time to divide the beat into two or four) and the displacement point specified from the user singing voice data and the time. Also good. (See FIG. 18). That is, the control unit 10 may acquire beat data indicating the beat of the music, and evaluate the singing voice based on the time difference between the time indicated by the acquired beat data and the time of the pitch displacement point in the user singing voice. . In this case, the control unit 10 may perform the evaluation process so that the evaluation becomes lower as the deviation amount (that is, the time difference) between the beat and the displacement point is larger. According to this aspect, since the singing can be evaluated without using the model sound data, the labor of creating the model sound data can be saved. Moreover, the control part 10 may use together the evaluation process which concerns on this modification, and the evaluation process which concerns on the above-mentioned embodiment. That is, the control unit 10 performs evaluation based on the time difference between the time of the displacement point of the pitch of the user singing voice and the time indicated by the beat data of the music, and the change amount of the pitch of the user singing voice and the pitch of the model sound data You may make it evaluate based on the difference with the variation | change_quantity.

  Further, as another aspect not using the scoring data, for example, the control unit 10 may statistically process the maximum value among the displacement points, and the higher the score is, the smaller the variance is. That is, the amount of change in the pitch value of the peak value (see the displacement points p91, 92, and 93 in FIG. 19) that appears on the graph representing the temporal change in the pitch of the sound indicated by the user singing voice data. Deviation) may be calculated, and the evaluation process may be performed so that the smaller the calculated change amount (deviation), the higher the evaluation. As an aspect of calculating the score, for example, a value obtained by subtracting the deviation from 100 may be calculated as a score. Among the displacement points, the value of the point where the value becomes the maximum, in the case of singing such as rap, is gathered for the singers who are good at singing (see FIG. 19). Therefore, by performing such an evaluation process, a higher evaluation is obtained for a singing with the same displacement point value. According to this aspect, since the singing can be evaluated without using the model sound data, the labor of creating the model sound data can be saved.

  Moreover, as another aspect which does not use the scoring data, for example, the control unit 10 may specify a rising part of the volume of the user singing voice, and evaluate the singing rhythm using the specified rising part. In this case, for example, the control unit 10 performs the evaluation process using the difference between the time of the beat included in the music information (or the time when the interval between the beats is divided into two or four) and the time of the specified rising portion of the volume. You may go. In this case, the control unit 10 may perform the evaluation process so that the evaluation becomes lower as the deviation amount (that is, the time difference) between the beat and the rising portion is larger. According to this aspect, evaluation can be performed without using scoring data. Moreover, according to this aspect, since the singing can be evaluated without using the model sound data, the labor for creating the model sound data can be saved. Moreover, not only the rising part of a sound volume but the control part 10 may evaluate the rhythm of a song using the rising part of a pitch. In this case, the control unit 10 may specify the rising part of the pitch of the user singing voice, and perform the evaluation process using the difference between the time of the beat included in the music information and the time of the rising part of the specified pitch. . Moreover, not only the rising part of a pitch but the control part 10 may evaluate the rhythm of a song using the timing at which the detection of the user song voice was started. In this case, the control unit 10 may specify the timing when the detection of the user singing voice is started, and perform the evaluation process using the time difference between the time of the beat included in the music information and the specified timing. According to this aspect, since the singing can be evaluated without using the model sound data, the labor of creating the model sound data can be saved.

<Modification 8>
In the above-described embodiment, the control unit 10 obtains the section data indicating the specific section. Instead, the control unit 10 analyzes the user singing voice data according to a predetermined algorithm and performs analysis. You may make it identify a specific area according to a result. Specifically, for example, the control unit 10 may specify a section in which the pitch change mode satisfies a predetermined condition as the specific section. Also in this case, similarly to the above-described embodiment, the control unit 10 may perform scoring using scoring data in the specific section. According to this aspect, the trouble of describing the specific section in the song data in advance can be saved.

<Modification 9>
In the above-described embodiment, the control unit 10 performs the scoring process using the difference in pitch change and the difference in volume change. However, the control unit 10 may be configured not to take the change in volume into account. That is, the evaluation value may be calculated based on the change in pitch. In this case, it is not necessary to include data representing the volume displacement point in the scoring data.

  In the above-described embodiment, the scoring data is stored in advance in the scoring data storage area 24. However, the present invention is not limited to this, and the control unit 10 is a model song (hereinafter referred to as "model song"). ) (Hereinafter, “exemplary song data”) may be analyzed to generate scoring data.

  In the above-described embodiment, the control unit 10 stores the singing voice data in the user singing voice data storage area 25 and performs the scoring after the singing is finished. However, not limited to this, the scoring is performed in real time during the singing. Processing may be performed.

  In the above-described embodiment, the control unit 10 compares the pitch of the singing voice with the pitch of the GM data, and performs the evaluation process according to the comparison result. The aspect of this may be sufficient. For example, an evaluation value, that is, an evaluation result may be calculated for an evaluation item by using various known methods such as frequency analysis using FFT (Fast Fourier Transform) and sound volume analysis.

  In the above-described embodiment, the control unit 10 outputs the scoring result to the display unit 40. However, the present invention is not limited thereto, and the data indicating the scoring result may be output to an externally connected storage device. Further, for example, the scoring result may be output by transmitting to a server device connected via a communication network. In this embodiment, the user is informed by outputting the scoring result to the display unit 40. However, the manner of notification is not limited to this. For example, the scoring result may be notified by a voice message or a notification sound. Any information may be used as long as it notifies the user.

  Moreover, in the above-mentioned embodiment, although the control part 10 evaluated the song voice of a singer, it may replace with the song voice of a singer, and may evaluate the performance sound of the musical instrument by a player. 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.

  In the above-described embodiment, the control unit 10 uses the pitch and the volume as the characteristics of the sound. However, the characteristics of the sound are not limited to the pitch and the volume, and may be other characteristics. For example, the characteristics of the sound are the fluctuation of the power of a specific harmonic, the ratio of the power of the specific harmonic and the fundamental, the ratio of the total harmonic power and the power of the fundamental, the SN ratio, the loudness (the volume is the frequency characteristic of the hearing) A value that is corrected according to JIS C1509 (also referred to as “A-weighted sound pressure level” or “sound level”). May be.

  Moreover, in the above-mentioned embodiment, the control part 10 evaluated user singing voice data based on the difference of the variation | change_quantity of the time of the displacement point regarding user singing voice data, and the variation | change_quantity of the time of the displacement point regarding model sound data. However, the present invention is not limited to this, and regarding the time of the displacement point, although the absolute time is correct, the evaluation may be made high. That is, the control unit 10 specifies the difference between the time of the displacement point related to the user singing voice data and the time of the displacement point related to the model sound data, and evaluates the sound indicated by the user singing voice data based on the specified difference. Good. The control unit 10 is an example of a time displacement specifying unit according to the present invention.

<Modification 10>
In the above-described embodiment, the control unit 10 performs the evaluation based on the difference in the change amount of the sound feature of the user singing voice data in the specific section, while the sound of the user singing voice data is used in the section other than the specific section. The evaluation was performed based on the difference between the feature of the sound and the feature of the sound indicated by the GM data. Not only this but the control part 10 performs the evaluation based on the difference of the variation | change_quantity of the characteristic of a sound in the specific area, or performs the conventional singing evaluation mainly on the pitch (namely, the characteristic of the sound of user singing voice data) (Evaluation is based on the difference from the sound feature indicated by the GM data).

<Modification 11>
In the above-described embodiment, two or more devices connected by a communication network share the functions related to the karaoke device 100 of the above-described embodiment, and a system including the plurality of devices uses the karaoke device 100 of the same embodiment. It may be realized. 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 scoring processing is connected via a communication network. In this case, for example, the computer apparatus converts the sound collected by the microphone into an audio signal and transmits it to the server apparatus, and the server apparatus analyzes and scores the received audio signal, and the scoring result is calculated by the computer apparatus. May be sent to. According to this aspect, the processing load of the karaoke terminal is reduced, and statistical processing in the server becomes possible.

<Modification 12>
In addition to the evaluation apparatus, the present invention can be understood as a method for realizing these and a program for causing a computer to realize a voice evaluation 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).

DESCRIPTION OF SYMBOLS 10 ... Control part, 20 ... Storage part, 21 ... Accompaniment data storage area, 22 ... Video data storage area, 23 ... GM data storage area, 24 ... Scoring data storage area, 25 ... User singing voice data storage area, 30 ... Operation unit, 40 ... display unit, 50 ... communication control unit, 60 ... audio processing unit, 61 ... microphone, 62 ... speaker, 70 ... bus, 100 ... karaoke device, 200 ... server device, 210 ... network storage, 300 ... user Singing voice curve, 400 ... scoring function, 500 ... GM curve

Claims (6)

An audio acquisition unit for acquiring audio data indicating an audio waveform;
An exemplary sound acquisition unit that acquires exemplary sound data indicating characteristics of the exemplary sound;
A feature specifying unit for specifying the feature of the sound indicated by the voice data acquired by the voice acquisition unit;
A difference specifying unit that specifies a difference between the amount of change of the feature specified by the feature specifying unit and a change amount of each of the change amount of the feature of the sound indicated by the exemplary sound data;
An audio evaluation device comprising: an evaluation unit that evaluates a sound indicated by the audio data based on the difference specified by the difference specifying unit and outputs an evaluation result. When the point where the slope of the graph representing the temporal change in the characteristic of the sound changes more than a predetermined threshold is used as the displacement point,
A displacement point identifying unit that identifies a displacement point of the feature identified by the feature identifying unit;
An associating unit that associates a displacement point related to the feature indicated by the exemplary sound data with a displacement point of the audio data identified by the displacement point identifying unit that appears within a predetermined time difference from the time of the displacement point. And
The difference specifying unit specifies a difference between a feature change amount at a displacement point related to the audio data correlated by the association unit and a feature change amount at a displacement point related to the exemplary sound data in specifying the difference. The speech evaluation apparatus according to claim 1, wherein: A time displacement identifying unit that identifies the difference between the time of the displacement point related to the audio data and the time of the displacement point related to the exemplary sound data;
The evaluation unit evaluates the sound indicated by the audio data based on the difference specified by the difference specifying unit and the difference specified by the time displacement specifying unit, and outputs an evaluation result. 2. The voice evaluation device according to 2. The evaluation unit outputs the same evaluation result when the difference specified by the difference specifying unit is within a predetermined threshold value. In other cases, the evaluation unit increases the specified difference. The speech evaluation apparatus according to any one of claims 1 to 3, wherein an evaluation result indicating low evaluation is output. A section data obtaining unit for obtaining section data representing a section in which the voice data is sung in a predetermined singing manner;
In the section indicated by the section data acquired by the section data acquisition section, the evaluation section performs evaluation based on the difference specified by the difference specifying section, while in a section other than the section indicated by the section data. The speech evaluation apparatus according to claim 1, wherein the evaluation is performed based on a difference between the feature specified by the feature specifying unit and the feature of the sound indicated by the model sound data. Analyzing the voice data according to a predetermined algorithm, and according to the analysis result, comprising a section specifying unit for specifying a section for performing singing according to a predetermined singing mode,
The evaluation unit performs evaluation based on the difference specified by the difference specifying unit in the section specified by the section specifying unit, while the feature specifying unit performs the evaluation in a section other than the specified section. The speech evaluation apparatus according to claim 1, wherein the evaluation is performed based on a difference between the identified feature and the feature of the sound indicated by the model sound data.

Images