JP5625321B2 - Speech synthesis apparatus and program - Google Patents

Description

  The present invention relates to a technique for setting a control variable applied to speech synthesis.

  Techniques for synthesizing sounds (typically singing sounds) in which musical expressions such as the intensity of pronunciation and the strength of breath components are controlled in various ways have been proposed. Patent Document 1 discloses an image (piano roll image) showing a time series of sounds designated as synthesis targets (hereinafter referred to as “designated sounds”) and a plurality of types of control variables representing facial expressions given to each designated sound. There has been proposed a technique for displaying a graph showing temporal transitions of the above with a common time axis. The user can edit the temporal transition of the control variable by selecting one of a plurality of types of control variables for which temporal transition is set in advance and operating the graph.

JP 2008-165130 A

  However, in order to change the facial expression given to the synthesized sound under the technique of Patent Document 1, it is necessary for the user to edit a control variable prepared in advance. Therefore, there is a problem that the burden of work is large for a user who is not familiar with the correlation between the control variable and the musical expression that changes in accordance with the control variable. In a configuration in which a plurality of types of control variables can be edited, it is necessary to select and edit any one of the plurality of types of control variables sequentially one by one. Therefore, the problem of an increase in the work load becomes particularly serious. . In view of the above circumstances, an object of the present invention is to reduce a burden on a user's work for setting a plurality of types of control variables.

In order to solve the above-described problems, the speech synthesizer according to the present invention variably applies an application section in a time series of a designated sound (sound designated as a synthesis target) indicated by music information according to an instruction from a user. The section selection means for setting, the variable selection means for selecting the variable information according to the instruction from the user among the plurality of variable information indicating the time series of the control variables applied to the speech synthesis, and the variable selection means selected Variable setting means for setting the time series of control variables in the applicable section according to the variable information, and means for synthesizing the specified sound indicated by the music information, wherein the time series of the control variable set by the variable setting means is applied to the applicable section Voice synthesizing means applied to the synthesis of the designated sound. Note that the music information and the variable information are stored in a separate storage area set in a single storage device (for example, the storage device 12 in FIG. 1) or a storage area set in each separate storage device. obtain.

  In the above configuration, variable information selected in response to an instruction from the user among a plurality of variable information indicating temporal transitions of control variables is applied to speech synthesis of each designated sound in the application section. This is used to set the time series of the control variable X. Therefore, in order to change the musical expression assigned to the synthesized sound, the user needs to change (edit) the time series of the control variables to the desired value for the synthesized sound. There is an advantage that the burden on the user's work necessary for giving a facial expression is reduced. However, in addition to the configuration in which the variable selection unit selects any one of the plurality of pieces of variable information, the configuration in which the user edits the control variable after setting by the variable setting unit in the same manner as in Patent Document 1 is also within the scope of the present invention. It can employ | adopt as a suitable aspect of.

  In a preferred aspect of the present invention, each of the plurality of variable information indicates a time series of the control variable for each of a plurality of types of control variables applied to speech synthesis. In the above aspect, since each variable information indicates a time series of a plurality of types of control variables, an advantage that synthetic sounds with various facial expressions can be generated as compared with a configuration in which the variable information indicates only one type of control variable. There is.

  A speech synthesizer according to a preferred aspect of the present invention uses a common time axis for a time series of sound indicators corresponding to each designated sound indicated by music information and a time series of control variables set by the variable setting means. Display control means for displaying on the display device. In the above aspect, since the time series of the sound indicator of each designated sound and the time series of the control variable are displayed with the time axis in common, the musical sound given to each designated sound according to the control variable There is an advantage that the user can easily confirm the facial expression.

In a preferred aspect of the present invention, the variable selecting means includes a user among a plurality of pieces of variable information corresponding to different combinations of options relating to each of a plurality of attributes (for example, sound quality, genre, song part, and key) of the synthesized sound. Selects variable information corresponding to the combination of options indicated for each attribute . According to the above aspect, since variable information is selected according to an instruction of an attribute familiar to the user, it is appropriate even if the user has no detailed knowledge about the musical expression indicated by the variable information. Advantageous variable information can be applied to speech synthesis.

  In a preferred aspect of the present invention, the section setting means variably sets the editing section in the application section in response to an instruction from the user, and the variable setting means is a control variable in the editing section of the application section. A series and a time series of control variables in a section other than the editing section (non-editing section) can be set independently. In the above aspect, the time series of the control variables is set independently of the editing section corresponding to the instruction from the user in the applied section, in the section other than the editing section. That is, it is possible to partially change the time series of control variables in the application section. Therefore, compared with the configuration in which only the applicable section is set, there is an advantage that synthesized sounds with various facial expressions that reflect the intention to the user can be generated. A specific example of the above aspect will be described later as a second embodiment, for example.

  By the way, in the configuration in which the time series of the control variable is set independently in the editing section and the non-editing section, the time series of the control variable may be discontinuous at the boundary between the editing section and the non-editing section. Therefore, in a preferred aspect of the present invention, the variable setting means executes control variable interpolation so that the control variable is continuous inside and outside the editing section in the application section. In the above aspect, since the discontinuous change of the control variable at the boundary between the editing section and the non-editing section is suppressed, the advantage that an unnatural (abrupt) change in the musical expression of the synthesized sound is prevented. There is. In addition, the specific example of the above aspect is later mentioned, for example as 3rd Embodiment.

  The speech synthesizer according to a preferred aspect of the present invention includes adjustment value setting means for variably setting an effect adjustment value in accordance with an instruction from a user, and the variable setting means has a degree according to the effect adjustment value. A time series of control variables in the application section is set so that the variable information is reflected in the synthesis of the designated sound in the application section. In the above aspect, since the variable information is reflected in the speech synthesis to a degree corresponding to the effect adjustment value set by the instruction from the user, the synthesized sound of various expressions reflecting the user's musical intention There is an advantage that can be generated.

  A speech synthesizer according to a preferred aspect of the present invention comprises adjustment value setting means for individually setting an effect adjustment value according to an instruction from a user in an editing section of an applied section and a section other than the editing section. The variable setting means reflects the variable information in the synthesis of the specified sound in the editing section to a degree according to the effect adjustment value in the editing section, and the section in the degree according to the effect adjustment value in the section other than the editing section. The time series of the control variables in the applicable section is set so that the variable information is reflected in the synthesis of the designated sound in the. In the above aspect, the time series of the control variable is set independently for the editing section and the non-editing section in accordance with the effect adjustment value set for the editing section and the non-editing section of the applied section. Therefore, there is an advantage that a synthesized sound with various expressions according to the intention can be generated for the user.

  In the configuration in which the effect adjustment value is set for the editing section and the non-editing section, the musical expression of the synthesized sound is caused by the fact that the effect adjustment value is extremely different between the editing section and the non-editing section. It may be unnatural. Therefore, the adjustment value setting means according to a preferred aspect of the present invention provides an instruction from the user while maintaining the ratio between the effect adjustment value of the editing section and the effect adjustment value other than the editing section of the application section. Change accordingly. In the above aspect, since the ratio between the effect adjustment value of the editing section and the effect adjustment value of the non-editing section is maintained, the musical expression of the synthesized sound is prevented from changing unnaturally inside and outside the editing section. There is an advantage that you can. In addition, the specific example of the above aspect is later mentioned as 4th Embodiment, for example.

The music information processing apparatus according to each of the aspects described above is realized by hardware (electronic circuit) such as a DSP (Digital Signal Processor) dedicated to processing music information, and a general-purpose device such as a CPU (Central Processing Unit). This is also realized by cooperation between the arithmetic processing unit and the program. The program of the present invention includes a section setting means for setting a computer to variably set an application section in a time series of a designated sound (sound designated as a synthesis target) indicated by music information in accordance with an instruction from a user, speech synthesis Variable selection means for selecting variable information according to an instruction from the user from among a plurality of variable information indicating a time series of control variables applied to the control, and control within the application section according to the variable information selected by the variable selection means Variable setting means for setting a time series of variables, and means for synthesizing a specified sound indicated by music information, and applying the time series of control variables set by the variable setting means to the synthesis of the specified sound in the application section It functions as a speech synthesis means. According to the above program, the same operation and effect as the speech synthesizer according to the present invention are realized. The program of the present invention is provided to a user in a form stored in a computer-readable recording medium and installed in the computer, or provided from a server device in a form of distribution via a communication network and installed in the computer. Is done.

It is a block diagram of the speech synthesizer concerning a 1st embodiment. It is a schematic diagram of an edit image. It is a schematic diagram of an edited image when an application section is instructed. It is a schematic diagram of a management image. It is a schematic diagram of an operation image. It is a schematic diagram of an operation image when an interval condition and an effect adjustment value are instructed. It is a schematic diagram of an edited image in a state where a variable transition image is displayed. It is a schematic diagram of the edited image in 2nd Embodiment. It is a schematic diagram of the management image in 2nd Embodiment. It is a schematic diagram of the management image in 2nd Embodiment. It is a schematic diagram for demonstrating the interpolation of the control variable by the area setting part of 3rd Embodiment. It is a schematic diagram of the management image in 4th Embodiment.

<A: First Embodiment>
FIG. 1 is a block diagram of a speech synthesizer 100 according to the first embodiment of the present invention. The speech synthesizer 100 is a device that synthesizes various sounds such as singing sounds (hereinafter referred to as “synthesized sounds”). As shown in FIG. 1, the control device 10, the storage device 12, the input device 14, and the display device 16. And a sound emitting device 18. In the following, it is assumed that the speech synthesizer 100 is used for synthesizing a song singing sound.

  The control device (CPU) 10 executes a plurality of functions (display control unit 22, information generation unit 24, section setting unit 26, variable processing necessary for generating the audio signal SOUT by executing the program PG stored in the storage device 12. Unit 32 and speech synthesis unit 34). The audio signal SOUT is a signal representing the waveform of the synthesized sound. A configuration in which each function of the control device 10 is realized by a dedicated electronic circuit (DSP) or a configuration in which each function of the control device 10 is distributed over a plurality of integrated circuits may be employed.

  The input device 14 is a device (for example, a mouse or a keyboard) that receives an instruction from a user. The display device (for example, a liquid crystal display device) 16 displays an image instructed from the control device 10. A sound emitting device (for example, a speaker or a headphone) 18 emits a sound wave corresponding to the sound signal SOUT generated by the control device 10.

  The storage device 12 stores a program PG executed by the control device 10 and various data (phoneme information DV, music information DS, variable information DP) used by the control device 10. A known recording medium (or a combination of a plurality of types of recording media) such as a semiconductor recording medium or a magnetic recording medium is arbitrarily employed as the storage device 12. A configuration in which the program PG and each data (DV, DS, DP) are distributed and stored in a plurality of recording media may be employed.

  The phoneme information DV is a data group used as a synthetic sound material, and includes a large number of unit data corresponding to different speech units (for example, data indicating a time waveform or a feature amount of a speech unit). Is done. The phoneme segment is a phoneme corresponding to a minimum unit that can be audibly distinguished from a voice, or a phoneme chain in which a plurality of phonemes are connected.

  The music information DS is information (score data) indicating the time series of each designated sound that constitutes the music. Specifically, the music information DS includes the pitch (note number) of the specified sound, the pronunciation period (for example, the start time and duration of pronunciation), and the pronunciation characters (for example, syllables and phonemes corresponding to the words of the lyrics). Specify for each specified sound.

  The variable information DP is a data group indicating temporal changes in a plurality of types of control variables (control parameters) X applied to speech synthesis. The control variable X is a variable for controlling a musical expression given to the synthesized sound. Specifically, the sound intensity (velocity), volume (dynamics) of the specified sound, breathness intensity (breathness), clarity (brightness, clearness), opening degree during pronunciation, gender ( genderfactor), time point of continuous pitch change (portamento) (portamento-timing), minute pitch change (pitch-bend), maximum pitch change (pitch-bend sensitivity), etc. A known variable applied to the above is arbitrarily adopted as the control variable X.

  As shown in FIG. 1, the storage device 12 stores a plurality of variable information DP. One variable information DP includes a plurality of variable transition data V corresponding to different types of control variables X (X1, X2,...). The variable transition data V of each control variable X is a data string indicating a time series (temporal transition) of the control variable X over a predetermined time. The mode of change of the control variable X indicated by each variable transition data V may be different for each variable information DP even in the variable transition data V of the same type of control variable X. The type of the control variable X indicated by each variable transition data V in the variable information DP is basically common to a plurality of variable information DP, but the type of the control variable X is different for each variable information DP. It is also possible to adopt.

  The display control unit 22 in FIG. 1 generates images (edited image 60, management image 70, operation image 80) visually recognized by the user in order to generate and edit music information DS and edit musical expressions given to synthesized sounds. ) Is displayed on the display device 16. FIG. 2 is a schematic diagram of an edited image 60 used for creating and editing the music information DS. As shown in FIG. 2, the edited image 60 includes a score area 62 that displays a time series of designated sounds, and a variable area 64 that displays changes over time of the control variable X.

  The score area 62 is a piano roll type image area in which a vertical axis (pitch axis) corresponding to pitch and a horizontal axis (time axis) corresponding to time are set. The user designates the pitch of the designated sound and the sound generation period (start point and end point) by appropriately operating the input device 14 while viewing the score area 62. The display control unit 22 arranges a sound indicator 622 corresponding to the designated sound instructed by the user in the score area 62. The position of the sound indicator 622 in the direction of the pitch axis is determined in accordance with the pitch instructed by the user, and the position and size of the sound indicator 622 in the direction of the time axis is in accordance with the pronunciation period instructed by the user. It is determined. In addition, the user instructs the pronunciation characters (lyrics) of each designated sound by appropriately operating the input device 14. It is also possible to employ a configuration in which a score image in which a musical score of a designated sound is recorded in a staff score is arranged in the score area 62.

  The information generation unit 24 in FIG. 1 stores the pitch, the pronunciation period, and the pronunciation character of the designated sound instructed by the user for the score area 62 in the music information DS of the storage device 12. By repeating the above processing, music information DS indicating the time series of the designated sound instructed by the user is generated in the storage device 12, and the time series of the sound indicator 622 of each designated sound is illustrated in FIG. As shown in FIG.

  The section setting unit 26 in FIG. 1 uses a section SA (hereinafter referred to as “applied section”) SA to which the variable information DP is applied in the time series (songs) of the designated sound indicated by the music information DS for the input device 14. It is variably set according to instructions from the person. For example, the section setting unit 26 specifies a section extending from the start point to the end point instructed by the user with respect to the score area 62 as the application section SA. The section setting unit 26 sequentially specifies a plurality of application sections SA that do not overlap in time in the music in accordance with an instruction from the user. The start point and the end point of each application section SA can be changed at any time according to an instruction from the user to the input device 14.

  As shown in FIG. 3, the display control unit 22 arranges a section indicator (shaded portion) 624 indicating the application section SA set by the section setting unit 26 in the edited image 60. The section indicator 624 is, for example, a band-shaped image extending along the time axis from the start point to the end point of the application section SA in the score area 62. Further, identification information (for example, the name “Part A” in FIG. 3) of the application section SA designated by the user is added to the section indicator 624.

  In addition, the display control unit 22 causes the display device 16 to display the management image 70 of FIG. 4 used for management of each application section SA. As shown in FIG. 4, the management image 70 is a chart in which records 72 are arranged for each application section SA. Each record 72 includes identification information (name) of the application section SA and a duration. The duration is specified by the start point and end point of the application section SA. The method for specifying the duration is arbitrary. For example, in addition to the method of specifying the time of the start point and the end point as illustrated in FIG. 4, a method of specifying the start point and end point of the duration by the number of the bar or beat point in the music may be employed. The user can also instruct the application section SA by direct input to the management image 70 (continuation period or identification information input). That is, the section setting unit 26 sets an application section SA corresponding to the duration period input by the user for the management image 70.

  The variable processing unit 32 of FIG. 1 generates a plurality of variable transition data W corresponding to different control variables X (X1, X2,...) From the variable information DP (each variable transition data V) in the storage device 12. . The variable transition data W corresponding to each control variable X is a data string indicating a time series (temporal transition) of the control variable X to be applied to speech synthesis in the application section SA. A specific configuration and operation of the variable processing unit 32 will be described later.

  The voice synthesizer 34 synthesizes the designated sound indicated by the music information DS stored in the storage device 12 to generate a voice signal SOUT. Specifically, the speech synthesizer 34 generates the segment data corresponding to the pronunciation character (speech segment) of each designated sound indicated by the music information DS among the phoneme information DV of the storage device 12. The audio signal SOUT is generated by adjusting the pitch and the tone generation period and connecting them to each other. A plurality of variable transition data W generated by the variable processing unit 32 is applied to the synthesis of each designated sound within the application section SA set by the section setting unit 26 in the time series (musical piece) of the designated sound indicated by the music information DS. Is done. In other words, the synthesized speech signal SOUT is generated by giving a musical expression corresponding to the variable transition data W of each control variable X to the designated sound indicated by the music information DS. For each designated sound in the sections other than the application section SA, a configuration in which a musical expression is not applied, or a configuration in which each control variable X is fixed to a predetermined value (initial value) and a facial expression is applied can be employed. A known technique is arbitrarily employed for speech synthesis in accordance with the music information DS and the control variable X.

  As shown in FIG. 1, the variable processing unit 32 includes a variable selection unit 42, an adjustment value setting unit 44, and a variable setting unit 46. The variable selection unit 42 selects the variable information DP corresponding to the instruction from the user for each application section SA among the plurality of variable information DP stored in the storage device 12. Specifically, the variable selection unit 42 acquires variable information DP corresponding to a condition (hereinafter referred to as “section condition”) instructed by the user from the input device 14 for the application section SA. The adjustment value setting unit 44 in FIG. 1 variably sets the effect adjustment value A indicating the degree to which the variable information DP selected by the variable selection unit 42 is reflected in the speech synthesis in accordance with an instruction from the user to the input device 14. To do.

  The display control unit 22 causes the display device 16 to display the operation image 80 of FIG. 5 that is used to instruct the section condition and the effect adjustment value A. As shown in FIG. 5, the operation image 80 includes a condition instruction area 82 used for instructing a section condition and an adjustment value instruction area 84 used for instructing an effect adjustment value A.

  The section condition is defined by a plurality of attributes (sound quality, genre, song part) regarding the application section SA. The condition indicating area 82 is an image in which a plurality of options (candidates) that can be selected by the user are listed for each of a plurality of attributes that define the section condition. Specifically, as illustrated in FIG. 5, a plurality of options relating to the sound quality of the application section SA (male voice, female voice, robot voice) and a plurality of options relating to the genre of the application section SA (rock, pops, jazz,... ) And a plurality of options (intro, A melody, B melody, rust,...) Related to the music part of the application section SA are arranged in the condition indicating area 82. As can be understood from the above examples, the music part corresponds to the structural positioning of the application section SA in the music.

  The user instructs one of the options arranged in the condition instruction area 82 for each attribute of the application section SA by appropriately operating the input device 14. FIG. 5 illustrates a case where the user selects “female voice” for the voice quality, selects “Bossa Nova” for the genre, and selects “A melody” for the song part. A combination of options designated by the user for each attribute is instructed to the variable selection unit 42 as a section condition.

  The storage device 12 stores the variable information DP for each section condition that can be instructed by the user (that is, for each combination of options that can be selected for each attribute). Each variable transition data V of the variable information DP corresponding to each section condition has a temporal transition of a musical expression expressed in a time series of the control variable X indicated by the variable transition data V, and the section condition (sound quality, It is created so as to be musically compatible with the melody that satisfies the genre and the song part. For example, each variable transition data V of the variable information DP corresponding to the section condition (female voice, bossa nova, A melody) indicated in the example of FIG. 5 is the temporal transition of the musical expression indicated by each variable transition data V. However, it is created so as to be musically adapted to the singing sound when the melody of “A melody” is uttered with “woman voice” in the music of “Bossa Nova”. The variable selection unit 42 in FIG. 1 selectively acquires, from the storage device 12, variable information DP that matches or approximates the section condition instructed by the user among the plurality of variable information DP stored in the storage device 12.

  On the other hand, a slider-type operation element image 842 that moves in response to an operation on the input device 14 is arranged in the adjustment value instruction area 84 of FIG. The adjustment value setting unit 44 sets the effect adjustment value A according to the position of the operator image 842 moved by the user by operating the input device 14. For example, when the operator image 842 is positioned at the lower end of the movable range, the adjustment value setting unit 44 sets the effect adjustment value A to the minimum value (for example, 0%) and operates the upper end of the movable range. When the child image 842 is positioned, the effect adjustment value A is set to the maximum value (for example, 100%). When the effect adjustment value A is the minimum value (0%), for example, the figure “O” with “OFF” written in FIG. 5 is turned on, and no musical expression is given to the synthesized sound in the application section SA. Is displayed and the effect adjustment value A exceeds the minimum value, the figure “O” with “ON” in FIG. 5 is turned on to give a musical expression to the synthesized sound in the application section SA. Is displayed.

  As shown in FIG. 6, the section condition designated by the user for the condition designation area 82 and the effect adjustment value A designated for the adjustment value designation area 84 are also reflected in the management image 70. Note that the user can also directly input the section condition and the effect adjustment value A to the management image 70 by appropriately operating the input device 14. That is, the variable selection unit 42 selects the variable information DP according to the section condition input to the management image 70, and the adjustment value setting unit 44 sets the effect adjustment value A according to the input to the management image 70.

  The variable setting unit 46 in FIG. 1 adjusts the variable transition data W of each control variable X (X1, X2,...) Applied to speech synthesis in the application section SA with the variable information DP selected by the variable selection unit 42. It is generated according to the effect adjustment value A set by the value setting unit 44. Specifically, the variable setting unit 46 executes a first process and a second process exemplified below.

  Each variable transition data V of the variable information DP stored in the storage device 12 indicates a time series of the control variable X over a predetermined time selected in advance regardless of the application section SA. The first process is a process for expanding and contracting each variable transition data V of the variable information DP so as to match the time length of the application section SA set by the section setting unit 26. For example, the process of expanding / contracting the time series of the control variable X indicated by each variable transition data V by, for example, interpolation (decimation), or connecting the variable transition data V along the time axis so that the time series of the control variable X is repeated. The process to perform can be adopted as the first process.

  The second process is a process of generating each variable transition data W by adjusting each variable transition data V after the first process according to the effect adjustment value A. Specifically, the variable setting unit 46 approximates the time series of the control variable X in the variable transition data W and the time series of the control variable X in the variable transition data V after the first processing (that is, the variable The variable transition data W of each control variable X is generated in the second process so that the degree of reflection of the variable transition data V with respect to the transition data W changes according to the effect adjustment value A. For example, as the effect adjustment value A is closer to the maximum value (100%), the variable setting unit 46 sets the time series of the control variable X in each variable transition data W after the second process to the control variable X of the variable transition data V. The closer to the time series and the closer the effect adjustment value A is to the minimum value (0%), the control variable X in each variable transition data W after the second processing is a predetermined value (eg, zero) unrelated to the variable transition data V. So that the variable transition data W of each control variable X is generated. In addition, although the 2nd process was performed after execution of a 1st process above, the structure which performs a 1st process after execution of a 2nd process is also employ | adopted.

  The voice synthesis unit 34 applies the time series of the control variable X of each variable transition data W generated by the variable processing unit 32 (variable setting unit 46) by the above procedure to the synthesis of each designated sound in the application section SA. An audio signal SOUT is generated. On the other hand, as shown in FIG. 7, the display control unit 22 converts an image 642 indicating the time series of the control variable X indicated by each variable transition data W (hereinafter referred to as “variable transition image”) 642 into a variable area 64 in the edited image 60. For each control variable X. Specifically, a graph indicating the transition of the control variable X (for example, a line graph) is displayed in the variable area 64 as a variable transition image 642. The display control unit 22 uses the same time axis for the time series of the control variable X indicated by the variable transition image 642 in the variable area 64 and the time series of each sound indicator 622 (designated sound) in the score area 62 (match). ), The variable transition image 642 is displayed. That is, the control variable X at each time point on the time axis in the variable transition image 642 in the variable region 64 is applied to the synthesis of the designated sound existing at that time point in the score region 62.

  As shown in FIG. 7, the variable transition image 642 is displayed in a different mode (display color, line width, line type) for each control variable X. For example, FIG. 7 illustrates a case where the variable transition image 642 of the control variable X1 is displayed with a solid line and the variable transition image 642 of the control variable X2 is displayed with a broken line. Note that a configuration in which only the variable transition image 642 of one or more types of control variables X instructed by the user among the plurality of control variables X is selectively arranged in the variable area 64 may be employed. Further, when an instruction to change (edit) each variable transition image 642 in the variable area 64 is input from the input device 14, the variable transition data W corresponding to the variable transition image 642 (time series of the control variable X) A configuration in which the variable setting unit 46 updates in response to a change instruction is also suitable.

  As described above, in the first embodiment, the variable information DP selected according to the instruction (section condition) from the user among the plurality of variable information DP indicating the temporal transition of the control variable X is the variable information DP. This is used for setting the time series of the control variable X (generation of variable transition data W) applied to speech synthesis of each designated sound in the application section SA. That is, it is possible to change the musical expression of the synthesized sound according to the selection result (variable information DP) by the variable selection unit 42. Therefore, in order to change the musical expression of the synthesized sound, compared with the technique of Patent Document 1 in which the user needs to edit the variable information, the use necessary for giving the desired expression to the synthesized sound. There is an advantage that the burden on the worker's work is reduced. Furthermore, in the above example, since the variable information DP indicates the temporal transition of the plurality of control variables X, compared with the configuration in which the variable information DP indicates only one type of control variable X, various facial expressions can be obtained. The special effect of being able to generate synthesized sounds is realized.

  In addition, since the degree to which the variable information DP is reflected in the speech synthesis is variably controlled according to the instruction (effect adjustment value A) from the user, when the variable information DP prepared in advance is used for the speech synthesis. Nevertheless, there is an advantage that various synthesized sounds reflecting the musical intention of the user can be generated as compared with the configuration in which the reflection degree of the variable information DP is fixed.

<B: Second Embodiment>
Next, a second embodiment of the present invention will be described. In the following examples, elements having the same functions and functions as those of the first embodiment are denoted by the same reference numerals, and detailed descriptions thereof are omitted as appropriate.

  FIG. 8 is a schematic diagram of an edited image 60 in the second embodiment. The section setting unit 26 of the second embodiment variably sets the editing section SB in the application section SA in accordance with an instruction from the user to the input device 14. The section setting unit 26 can sequentially specify a plurality of editing sections SB that do not overlap in time within each application section SA in accordance with an instruction from the user. Note that the same method as the setting of the application section SA is applied to the setting of the editing section SB.

  As shown in FIG. 8, the display control unit 22 includes a section indicator 626 indicating the editing section SB set by the section setting section 26, and identification information (“phrase A” in FIG. 8) designated by the user for the editing section SB. Is placed in the edited image 60. Further, as shown in FIG. 9, the display control unit 22 manages the record 72 including the identification information and the duration of the editing section SB set by the section setting unit 26 separately from the record 72 of the application section SA. 70. Note that the processing related to the sections other than the editing section SB (hereinafter referred to as “non-editing sections”) in the application section SA is the same as the processing related to the application section SA of the first embodiment, and thus description thereof will be omitted as appropriate.

  The user selects a desired editing section SB and appropriately manipulates the operation image 80, so that the section condition and effect adjustment value A of the editing section SB and the section condition and effect adjustment value A of the non-editing section and Can be set independently. The variable selection unit 42 selects the variable information DP corresponding to the section condition specified for the editing section SB, and the adjustment value setting section 44 notifies the variable setting section 46 of the effect adjustment value A specified for the editing section SB. To do. The display control unit 22 reflects the section condition and effect adjustment value A instructed by the user in the record 72 of the editing section SB in the management image 70. For example, in FIG. 10, when the effect adjustment value A in the editing section SB is increased from the effect adjustment value A (80% in FIG. 9) including the editing section SB (80%), as indicated by shading. → 90%) is assumed.

  The variable setting unit 46 independently sets the variable transition data W in the editing section SB and the variable transition data W in the non-editing section in the application section SA. That is, the variable setting unit 46 selects the variable transition data W of the editing section SB while the variable selection section 42 selects the editing section SB while maintaining the variable transition data W already generated for the non-editing section of the application section SA. It is generated according to the variable information DP and the effect adjustment value A set by the adjustment value setting unit 44 for the editing section SB. For example, the same method as the generation of the variable transition data W in the application section SA in the first embodiment is adopted to generate the variable transition data W in the editing section SB.

  As shown in FIG. 8, the display control unit 22 displays the variable transition image 642 for the non-editing section in the application section SA as in the first embodiment, and also displays the variable corresponding to the variable transition data W in the editing section SB. The transition image 642 is arranged in an area corresponding to the editing section SB in the variable area 64. The voice synthesizer 34 generates the voice signal SOUT by applying the variable transition data W set by the variable setting unit 46 for the editing section SB to the synthesis of each designated sound in the editing section SB.

  In the second embodiment, the variable transition data W is generated independently of the non-editing section for the editing section SB corresponding to the instruction from the user in the application section SA. That is, the user can partially edit the time series of the control variable X in the application section SA. Therefore, compared to the configuration in which only the application section SA is set (first embodiment), there is an advantage that it is possible to generate synthetic sounds with various facial expressions that highly reflect the user's intention.

<C: Third Embodiment>
When the variable transition data W is individually set in the editing section SB and the non-editing section in the application section SA as in the second embodiment, the variable transition is understood as understood from the contents of the variable area 64 in FIG. The time series of the control variable X indicated by the data W may be discontinuous at the boundary between the editing section SB and the non-editing section. Therefore, the variable setting unit 46 of the third embodiment allows the variable transition data W in the non-editing section and the variable transition data in the editing section SB so that the control variable X continues (smoothly transitions) inside and outside the editing section SB. Perform interpolation with W.

  For example, as shown in FIG. 11, the variable setting unit 46 includes the time series α1 of the control variable X indicated by the variable transition data W in the editing section SB and the sections immediately before and immediately after the editing section SB (non-editing section). By overlapping the time series α2 of the control variable X indicated by the variable transition data W on the time axis and cross-fading, the time series α3 (dashed line portion) of the control variable X at the start point and end point of the editing section SB ) Is calculated. The speech synthesizer 34 applies the interpolated variable transition data W to speech synthesis within the application section SA (the editing section SB and the non-editing section).

  According to the above configuration, since the discontinuous change of the control variable X at the boundary between the editing section SB and the non-editing section is suppressed, an unnatural (abrupt) change in the musical expression of the synthesized sound is prevented. The Therefore, there is an advantage that a natural synthesized sound in which a musical expression smoothly transitions can be generated. However, since there may be a song (a song that becomes unnatural when the control variable X is interpolated) for which it is desirable to discontinuously change the musical expression inside and outside the editing section SB, the variable transition data W is interpolated. A configuration in which the user can instruct the presence or absence of this is preferable.

<D: Fourth Embodiment>
In the second embodiment, the effect adjustment value A of the editing section SB and the effect adjustment value A of the non-editing section are set independently. However, if the effect adjustment value A is extremely different between the editing section SB and the non-editing section, the synthesis is performed. The transition of the musical expression of the sound may be unnatural. Therefore, in the fourth embodiment, one of the effect adjustment value A in the editing section SB and the effect adjustment value A in the non-editing section is changed in conjunction with the other.

  Specifically, the adjustment value setting unit 44 responds to an instruction from the user while maintaining the ratio (relative ratio) between the effect adjustment value A of the editing section SB and the effect adjustment value A of the non-editing section. Change. For example, assume that the effect adjustment value A of the non-editing section (application section SA) is set to 80% and the effect adjustment value A of the editing section SB is set to 90% as illustrated in FIG. As illustrated in FIG. 12, when the user changes the effect adjustment value A in the non-editing section (part A) from 80% to 70%, the adjustment value setting unit 44 displays the effect adjustment value A in the editing section SB. Change from 90% of 10 to 78% (≒ 90 × 70/80). Similarly, when the user changes the effect adjustment value A of the editing section SB, the adjustment value setting unit 44 changes the effect adjustment value A of the non-editing section so that the ratio between the two before the change is maintained. .

  According to the above configuration, each change is made while maintaining the ratio between the effect adjustment value A of the editing section SB and the effect adjustment value A of the non-editing section, so that the musical expression is not good inside and outside the editing section SB. There is an advantage that it is possible to suppress natural changes.

<E: Modification>
Each of the above forms can be variously modified. Specific modifications are exemplified below. Two or more aspects arbitrarily selected from the following examples can be appropriately combined.

(1) Modification 1
The attributes constituting the section condition are not limited to the above examples (sound quality, genre, song part). For example, a configuration is adopted in which the variable information DP is selected according to the section condition including the key (key) of the music. However, the instruction of the section condition is not essential for the selection of the variable information DP, and the method for selecting the variable information DP is appropriately changed. For example, a configuration in which the user directly instructs any of the plurality of variable information DP from the input device 14 (that is, not via the input of the section condition) may be employed. As can be understood from the above description, the variable selection unit 42 in each of the above forms is included as an element for selecting the variable information DP according to the instruction from the user among the plurality of variable information DP. However, according to each of the above forms in which section conditions familiar to the user such as sound quality, genre, and song part are used for selection of the variable information DP, detailed knowledge about the musical expression given by the variable information DP is obtained. Even if it is not, a special effect that the variable information DP desired by the user can be applied to speech synthesis is realized.

(2) Modification 2
A method of generating the variable transition data W according to the effect adjustment value A is arbitrary. For example, a configuration in which the variable setting unit 46 generates the variable transition data W by adding or multiplying the effect adjustment value A to the variable transition data V is also employed. If the effect of imparting various facial expressions is unnecessary, the setting of the effect adjustment value A (adjustment value setting unit 44) can be omitted.

(3) Modification 3
In the third embodiment, the control variable X is interpolated between the editing section SA and the non-editing section. However, a configuration in which the control variable X is interpolated for two application sections SA that are in succession on the time axis is also employed. For the interpolation of the control variable X between the application sections SA, the same method as in the third embodiment in which the control variable X is interpolated between the editing section SA and the non-editing section is employed. According to the configuration in which the control variable X is interpolated for each application section SA, the discontinuous change of the control variable X at the boundary of each successive application section SA is suppressed, so that the musical expression changes smoothly. There is an advantage that a simple synthesized sound can be generated.

(4) Modification 4
In each of the above embodiments, the speech synthesizer 100 in which the speech synthesizer 34 generates the speech signal SOUT according to the variable transition data W (time series of the control variable X) generated by the variable processing unit 32 (variable setting unit 46). However, the present invention can also be implemented as a device (variable processing device) that generates variable transition data W applied to speech synthesis.

  Specifically, the variable processing device includes a storage device 12 that stores a plurality of variable information DP indicating a time series of the control variable X, and a section setting unit that variably sets the application section SA according to an instruction from the user. 26, and a variable processing unit 32 that sets a time series (variable transition data W) of the control variable X in the application section SA using the variable information DP according to the instruction from the user among the plurality of variable information DP. It comprises. That is, the information generation unit 24, the speech synthesis unit 34, and the display control unit 22 can be omitted as appropriate. The variable transition data W generated by the variable processing device (variable processing unit 32) is separated from the variable processing device via a portable recording medium or communication network (having a speech synthesizing unit 34). Applied to speech synthesis.

100 ... speech synthesizer, 10 ... control device, 12 ... storage device, 14 ... input device, 16 ... display device, 18 ... sound emitting device, 22 ... display control unit, 24 ... information generation , 26... Section setting section, 32... Variable processing section, 34... Speech synthesis section, 42... Variable selection section, 44 ....... adjustment value setting section, 46. 62 …… Score area, 622 …… Sound indicator, 624,626… Section indicator, 64 …… Variable area, 70 …… Management image, 72 …… Record, 80 …… Operation image, 82 …… Condition Designated area, 84... Adjustment value designated area, 642... Variable transition image, SA... Applicable section, SB... Edit section, DS ... Music information, DV ... Phoneme information, DP ... Variable information, V, W: Variable transition data.

Claims (6)

Section setting means for variably setting the application section in the time series of the designated sound indicated by the music information according to an instruction from the user;
A plurality of variable information indicating the time sequence of the control variables applied to speech synthesis, among a plurality of variable information corresponding to a plurality of respective different combinations of the options for the attributes of the synthesized speech, wherein the user each Variable selection means for selecting variable information corresponding to the combination of options instructed for the attribute ;
Variable setting means for setting a time series of control variables in the application section according to the variable information selected by the variable selection means;
A voice synthesizer comprising: a voice synthesizing unit for synthesizing a designated sound indicated by the music information, the voice synthesizing unit applying a time series of control variables set by the variable setting unit to synthesis of the designated sound in the application section; . Section setting means for variably setting the application section in the time series of the designated sound indicated by the music information and the editing section in the application section according to an instruction from the user;
Variable selection means for selecting variable information according to an instruction from a user among a plurality of variable information indicating a time series of control variables applied to speech synthesis ;
A means for setting a time series of control variables in the applied section according to the variable information selected by the variable selecting means, except for the time series of the control variables in the editing section and the editing section other than the applied section. Variable setting means capable of independently setting the time series of control variables in the interval of
A voice synthesizer comprising: a voice synthesizing unit for synthesizing a designated sound indicated by the music information, the voice synthesizing unit applying a time series of control variables set by the variable setting unit to synthesis of the designated sound in the application section; . The speech synthesis apparatus according to claim 2 , wherein the variable setting unit performs interpolation of the control variable so that the control variable is continuous inside and outside the editing section in the application section. Comprising adjustment value setting means for variably setting the effect adjustment value in accordance with an instruction from the user;
The variable setting means sets a time series of control variables in the application section so that the variable information is reflected in the synthesis of the designated sound in the application section to a degree according to the effect adjustment value. The speech synthesizer according to claim 3 .   Computer
  Section setting means for variably setting the application section in the time series of the designated sound indicated by the music information according to an instruction from the user,
  A plurality of variable information indicating a time series of control variables applied to speech synthesis, and the user has instructed each of the attributes among the variable information corresponding to the combination of options for each of the plurality of attributes of the synthesized sound. Variable selection means for selecting variable information corresponding to the combination of options,
  Variable setting means for setting a time series of control variables in the application section according to the variable information selected by the variable selection means; and
  Means for synthesizing a designated sound indicated by the music information, the voice synthesizing means for applying a time series of control variables set by the variable setting means to synthesis of the designated sound in the application section;
  Program to function as Computer
  Section setting means for variably setting the application section in the time series of the designated sound indicated by the music information and the editing section in the application section according to an instruction from the user,
  Variable selection means for selecting variable information according to an instruction from a user among a plurality of variable information indicating a time series of control variables applied to speech synthesis;
  A means for setting a time series of control variables in the applied section according to the variable information selected by the variable selecting means, except for the time series of the control variables in the editing section and the editing section other than the applied section. Variable setting means capable of independently setting the time series of control variables in the interval of
  Means for synthesizing a designated sound indicated by the music information, the voice synthesizing means for applying a time series of control variables set by the variable setting means to synthesis of the designated sound in the application section;
  Program to function as.

Images