JP4244706B2 - Audio playback device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a sound reproducing device, a sound reproducing apparatus capable of reproducing the high quality for a particular phrase (such as "Hello") in particular by speech synthesis.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, a character string speech conversion device has been devised that converts character string information such as electronic mail into speech and outputs it. As a conventional character string speech conversion device, there is a device that divides character string information into phrase units and outputs the sound at the same time as displaying the content (for example, see Patent Document 1).
[0003]
[Patent Document 1]
JP 2001-7937 A
[0004]
[Problems to be solved by the invention]
However, in the conventional character string speech conversion device, the character string information is divided into phrases and output as speech. However, since the speech output is a set of sounds in pronunciation units (or character units), reproduction of the joint of the pronunciation units is performed. (Sound output) is strange. That is, in the conventional character string speech conversion system may be output by changing the tone of voice with good sound quality for the entire phrase, i.e. there is a problem that can not be output as a natural voice (e.g. "Hello").
[0005]
In order to solve this problem, for example, a method is conceivable in which a voice for a phrase (or phrase) is sampled in advance and stored as voice data, and is output as a voice waveform during reproduction. However, with this method, in order to improve the quality of audio output, the sampling frequency must be increased, and it is necessary to store a large volume of audio data.
[0006]
The present invention has been made to solve the above problems, be reproduced desired phrase consisting of a character string information (such as "Hello") to change the tone of voice as good speech quality (output) An audio reproducing apparatus that can be used is provided.
[0007]
[Means for Solving the Problems]
In order to solve the above problems, the present invention has the following configuration.
That is, Book The present invention provides a speech reproduction apparatus that has a synthesis dictionary that is a database that holds formant frame data corresponding to pronunciation units in advance, and that synthesizes speech using the synthesis dictionary by providing information in which pronunciation units are listed. The replacement unit replaces the formant frame data of the pronunciation unit held in the synthesis dictionary with arbitrary user data, and the stored data is replaced by the replacement unit when the information listing the pronunciation units is given. And speech synthesis means for synthesizing speech using the synthesis dictionary.
[0008]
Also, Book invention Before The user data is formant frame data acquired in units of phrases.
[0009]
Also, Book invention Before The user data is added to a timbre parameter for processing formant frame data held in the synthesis dictionary.
[0010]
Also, Book invention Before When the timbre parameter to which the user data is added is given and the timbre parameter is designated at the time of reproduction, the replacement means adds formant frame data, which is held in the synthesis dictionary, to the user data. The replacement, the speech synthesizing means, synthesizes speech using the synthesis dictionary replaced by the timbre parameter when the enumeration information in units of speech is given.
[0011]
Also, Book invention Is easy The user exchange is included in a data exchange format that defines an information structure for reproducing desired data by synchronizing music and voice, and voice is synthesized using the data exchange format .
[0012]
Also, Book invention Before The music piece information included in the information configured as the data exchange format is reproduced as it is, and the audio information included in the information is reproduced using the replacement unit and the voice synthesizing unit. To do.
[0013]
Also, Book invention Before The data exchange format includes information in which the user data is added to a voice parameter as a constituent element.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of an audio reproducing apparatus according to an embodiment of the present invention. First, the basic configuration of the audio reproduction device according to the present embodiment will be described.
[0015]
The audio playback device 1 includes an application 14, a middleware API 15, a converter 16, a driver 17, a default tone color parameter 18, a default synthesis dictionary 19, and a sound source 20, and includes a script 11, a user tone color parameter 12, a user phrase synthesis dictionary (variable length). The audio is reproduced by inputting 13.
[0016]
The sound reproduction device 1 is based on a method of reproducing sound by formant synthesis using a CSM (Composite Sine Wave Model) speech synthesis method using FM sound source resources. And in this embodiment, the user phrase synthetic | combination dictionary 13 is defined and the audio | voice reproduction apparatus 1 allocates a user phrase per phoneme to a timbre parameter. When the data of the user phrase synthesis dictionary 13 is assigned to the timbre parameter at the time of reproduction, the voice reproduction device 1 replaces the phoneme in the default synthesis dictionary 19 with the user phrase, and performs voice synthesis using the replaced data. The “phoneme” is a minimum unit of pronunciation, and there are two types of vowels and consonants in Japanese. Next, the details of the audio reproduction device 1 will be described.
[0017]
The script 11 defines a data format for reproducing “HV (Human Voice: voice synthesized by the above method)”. That is, the script 11 has a format for synthesizing speech composed of a message such as a synthesized character string including prosodic symbols, setting of sound to be generated, and a reproduction application. It has become. The definition of the data format in the script 11 is language-dependent and can be defined in various languages. In this embodiment, only the definition in Japanese is taken as an example.
[0018]
The user phrase synthesizing dictionary 13 and the default synthesizing dictionary 19 sample and analyze an actual voice in units of pronunciation characters (for example, “A”, “I”, etc.), thereby using eight formant frequencies, formant levels, and pitches as parameters. It is a database that is indexed and holds these parameters in advance as phonetic character data as formant frame data. The user phrase synthesizing dictionary 13 is a database constructed outside the middleware, and the user can arbitrarily create such a database, and the retained contents are completely replaced with the retained contents of the default synthetic dictionary 19 via the middleware API 15. Can do. That is, the entire contents of the default synthesis dictionary 19 can be replaced with the contents of the user phrase synthesis dictionary 13. On the other hand, the default synthesis dictionary 19 is a database constructed in the middleware.
[0019]
The user phrase synthesis dictionary 13 and the default synthesis dictionary 19 preferably have two types of male voice and female voice, respectively. Further, the quality of the output sound of the audio reproduction device 1 varies depending on the interval of the frame data held by the user phrase synthesis dictionary 13 and the default synthesis dictionary 19, but the interval of the frame data is set to 20 ms, for example.
[0020]
The user tone color parameter 12 and the default tone color parameter 18 are a group of parameters for controlling the voice quality in the output sound of the sound reproducing device 1. The user tone color parameter 12 and the default tone color parameter 18 are, for example, eight sets of formant frequencies and formant level changes (formant frequencies registered in the user phrase synthesis dictionary 13 and the default synthesis dictionary 19 and the amount of change from the formant level). Designation), and a basic waveform for formant synthesis can be designated, and various timbres can be created.
[0021]
The default timbre parameter 18 is a timbre parameter set that is previously stored in the middleware by default. The user tone color parameter 12 is a parameter that can be arbitrarily created by the user and is held outside the middleware. The user tone color parameter 12 extends the default tone color parameter 18 via the middleware API 15.
[0022]
The application 14 is software for reproducing the script 11.
A middleware API (Application Program Interface) 15 is an interface between an application 14 made of software, a converter 16 made of middleware, a driver 17, a default tone color parameter 18, and a default synthesis dictionary 19.
[0023]
The converter 16 interprets the script 11 and uses the driver 17 to finally convert it into formant frame string data in which frame data is continuously formed.
The driver 17 generates a formant frame sequence based on the pronunciation characters included in the script 11 and the default synthesis dictionary 19, interprets the timbre parameters, and processes the formant frame sequence.
The sound source 20 outputs a sound signal corresponding to the data output from the converter 16, and the sound signal is output to a speaker to become a sound.
[0024]
Next, features of the audio playback device 1 according to the present embodiment will be described in detail.
First, the user tone color parameter 12 includes a parameter for assigning a phrase ID held by the user phrase synthesis dictionary 13 to an arbitrary pronunciation unit. FIG. 2 is a diagram showing an example in which a phrase ID is assigned to each pronunciation unit. That is, FIG. 2 shows allocation of mora and phrase ID.
The mora means a beat, and in Japanese, it is a kana character unit.
[0025]
By assigning a phrase ID to each pronunciation unit, it is specified that the pronunciation unit specified by the user tone color parameter 12 is not the default synthesis dictionary 19 but the user phrase synthesis dictionary 13 is used. The user tone color parameter 12 preferably has an arbitrary number of sounding units that can be specified in one tone color parameter. As described above, the configuration in which the phrase ID is assigned to each sounding unit in the user tone color parameter 12 is an example of this embodiment, and any method can be used as long as it can be replaced with the sounding unit.
[0026]
Next, details of the user phrase synthesis dictionary 13 will be described. FIG. 3 is a diagram showing an example of the contents of the user phrase synthesis dictionary 13. The user phrase synthesis dictionary 13 stores frame data composed of eight sets of formant frequencies, formant levels, and pitches for each phrase ID. The “phrase” in FIG. 3 is a phrase having a single unit such as “good morning”. The “phrase” means an arbitrary lump without defining a group such as a word, a syllable, or a sentence.
[0027]
The tool for creating the user phrase synthesis dictionary 13 is equipped with an analysis engine that analyzes and generates frame data consisting of eight formant frequencies, formant levels, and pitches from normal sound files (* .wav, * aif, etc.). There is a need to.
[0028]
The script 11 has an event for changing voice quality, and the user tone color parameter 12 can be designated by this event.
[0029]
For example, the description of the script 11 is “TJK12 everyone X10 Aka”.
In this example, “K” is an event for specifying the default timbre parameter 18, and “X” is an event for specifying the user timbre parameter 12. Further, “X10” designates the user tone color parameter shown in FIG.
[0030]
And, in this case, the reproduced sound is "Hello everyone is Suzuki".
"Everyone" becomes a voice using the default tone color parameter 18 and default synthesis dictionary 19, also, "Suzuki" and "Hello" is the voice using a user tone color parameter 12 and a user phrase synthesis dictionary 13. In other words, "you" becomes a voice that was synthesized by reading each of the formant frame data of "only" and "Do not" and "of", "I" from the default synthesis dictionary 19, "Suzuki" and "Hello", respectively The phrase-based formant frame data is read from the user phrase synthesis dictionary 13 and synthesized.
[0031]
In the above example, “a”, “i”, and “ka” are used, but any character and symbol that can be expressed in text are acceptable. In addition, in the above example, "X10" and later, "A" is "Hello", because "or" is pronounced "Suzuki", the next time you want to pronounce the original of the "A" is the default synthesis dictionary What is necessary is just to put the symbol to return (for example, XOO).
[0032]
Next, a data exchange format of music reproduction sequence data (SMAF: Synthetic music Mobile Application Format) used in the audio reproducing apparatus 1 according to the present embodiment will be described with reference to FIG. FIG. 4 is an explanatory diagram showing the format of the SMAF file according to the present embodiment. SMAF is one of the data exchange formats for distributing and mutually using data for expressing music using a sound source, and is a data format specification for expressing multimedia contents on mobile terminals and the like. It is.
[0033]
The SMAF file 30 in the data exchange format shown in FIG. 4 has a basic structure of data chunks called “chunks”. The chunk is composed of a fixed length (8 bytes) header part and an arbitrary length body part. The header part is divided into a 4-byte chunk ID and a 4-byte chunk size. The chunk ID is used as a chunk identifier, and the chunk size indicates the length of the body part. The SMAF file 30 has a chunk structure for itself and various data included therein.
[0034]
As shown in FIG. 4, the SMAF file 30 includes a contents info chunk 31, an optional data chunk 32, a track track chunk 33, and an HV chunk. (HV Chunk) 36.
[0035]
The content info chunk 31 stores various management information about the SMAF file 30. For example, the content class, type, copyright information, genre name, song name, artist name, song / composer name, and the like are stored. Stored. The optional data chunk 32 stores information such as copyright information, genre name, song name, artist name, and lyrics / composer name. The optional data chunk 32 may not be provided in the SMAF file 30.
[0036]
The track chunk 33 is a chunk for storing a sequence track of music to be sent to a sound source, and includes a setup data chunk (Setup Data Chunk (option)) 34 and a sequence data chunk (Sequence Data Chunk) 35. Yes.
[0037]
The setup data chunk 34 is a chunk for storing timbre data of the sound source portion and the like, and stores a sequence of exclusive messages. The exclusive message is, for example, a tone color parameter registration message.
[0038]
The sequence data chunk 35 stores actual performance data, and stores a mix of HV (Human Voice) note-on that determines the playback timing of the script 11 and other sequence events. Here, HV and other music events are distinguished by HV channel designation.
[0039]
The HV chunk 36 includes an HV Setup Data Chunk (HV Setup Data Chunk (option)) 37, an HV User Phrase Dictionary Chunk (option) 38, and an HV-S chunk 39. Contains.
[0040]
The HV setup data chunk 37 stores HV user tone color parameters and a message for specifying an HV channel. The HV-S chunk 39 stores HV-script data.
[0041]
The contents of the user phrase synthesis dictionary 13 are stored in the HV user phrase phrase dictionary chunk 38. Also, the HV user tone color parameters stored in the HV setup data chunk 37 require parameters for assigning mora and phrase IDs shown in FIG.
[0042]
By applying the SMAF file 30 shown in FIG. 4 to the audio reproduction device 1, audio (HV) can be reproduced in synchronization with the music, and the contents of the user phrase synthesis dictionary 13 can also be reproduced. Is possible.
[0043]
Next, an HV authoring tool, which is a tool for creating the user phrase synthesis dictionary 13 in FIG. 1 and the SMAF file 30 shown in FIG. 4, will be described with reference to FIG. FIG. 5 is a functional image diagram showing an example of the HV authoring tool.
[0044]
When creating the SMAF file 30, the HV authoring tool 42 reads an SMF (Standard MIDI File) file 41 (including note-on that determines the HV sounding timing) created in advance by a MIDI sequencer, and creates an HV script UI 44 and an HV voice. Based on the information obtained from the editor 45, conversion processing to the SMAF file 43 (corresponding to the SMAF file 30) is performed.
[0045]
The HV voice editor 45 is an editor capable of editing the HV user tone color parameter (corresponding to the user tone color parameter 12) included in the HV user tone color file 48. The HV voice editor 45 can assign a user phrase to an arbitrary mora in addition to editing various HV tone parameters.
[0046]
The HV voice editor 45 has an interface for selecting a mora and a function for assigning an arbitrary sound file 50 to the mora. The sound file 50 allocated by the interface of the HV voice editor 45 is analyzed by the waveform analyzer 46 to generate eight sets of formant frequency, formant level and pitch frame data. These frame data can be input / output as individual files (HV user tone color file 48, HV user synthesis dictionary file 49).
[0047]
The HV script UI 44 can directly edit the HV script. This HV script can also be input / output as an individual file (HV script file 47). Further, the HV authoring tool 40 according to the present embodiment may include the HV authoring tool 42, the HV script UI 44, the HV voice editor 45, and the waveform analyzer 46.
[0048]
Next, an example in which the audio reproduction device 1 is applied to a mobile communication terminal will be described with reference to FIG. FIG. 6 is a block diagram illustrating a configuration example of the mobile communication terminal 60 including the audio reproduction device 1.
[0049]
The mobile communication terminal 60 includes, for example, a mobile phone, and includes a CPU 61, a ROM 62, a RAM 63, a display unit 64, a vibrator 65, an input unit 66, a communication unit 67, an antenna 68, an audio processing unit 69, a sound source 70, a speaker 71, and a bus. 72. The CPU 61 controls the entire mobile communication terminal 60. The ROM 62 stores various communication control programs, control programs such as a music reproduction program, and various constant data.
[0050]
The RAM 63 is used as a work area and stores music files and various application programs. The display unit 64 includes a liquid crystal display device (LCD). Vibrator 65 vibrates when an incoming call is received. The input unit 66 includes a plurality of buttons. The communication unit 67 includes a modem unit and the like, and is connected to the antenna 68.
[0051]
The voice processing unit 69 is connected to a transmission microphone and a reception speaker, and has a function of encoding and decoding a voice signal for a call. The sound source 70 reproduces music based on the music file stored in the RAM 63 or the like, reproduces sound, and outputs it to the speaker 71. The bus 72 is a transmission path for transferring data among the constituent elements of the CPU 61, ROM 62, RAM 63, display unit 64, vibrator 65, input unit 66, communication unit 67, audio processing unit 69, and sound source 70.
[0052]
Further, the communication unit 67 can download the HV-script file or the SMAF file 30 shown in FIG. 4 from the content server or the like and store it in the RAM 63. The ROM 62 also stores the application 14 and middleware program of the audio reproduction device 1 shown in FIG. The application 14 and middleware program are read and activated by the CPU 61. Further, the CPU 61 generates formant frame data by interpreting the HV-script stored in the RAM 63, and sends the formant frame data to the sound source 70.
[0053]
(Operation)
Next, the operation of the audio playback device 1 will be described. First, a method for producing the user phrase synthesis dictionary 13 will be described. FIG. 7 is a flowchart showing a method for producing the user phrase synthesis dictionary 13.
[0054]
First, by using the HV authoring tool 42 shown in FIG. 5, the HV tone color using the user phrase synthesis dictionary 13 is selected, and the HV voice editor 45 is activated (step S1).
Next, using the HV voice editor 45, a mora to be applied is selected and a sound file is pasted. Then, the HV voice editor 45 outputs a user phrase dictionary (corresponding to the HV user synthesis dictionary file 49) (step S2).
[0055]
Next, the HV voice parameter is edited using the HV voice editor 45. Then, the HV voice editor 45 outputs a user tone color parameter (corresponding to the HV user tone color file 48) (step S3).
[0056]
Next, using the HV script UI 44, a voice quality change event for designating the corresponding HV tone is described in the HV-script, and a mora to be reproduced is described. Then, the HV script UI 44 outputs an HV-script (corresponding to the HV script file 47) (step S4).
[0057]
Next, the reproduction | regeneration operation | movement of the user phrase dictionary in the audio | voice reproduction apparatus 1 is demonstrated with reference to FIG. FIG. 8 is a flowchart showing the reproduction operation of the user phrase synthesis dictionary in the audio reproduction device 1.
First, the user tone color parameter 12 and the user phrase synthesis dictionary 13 are registered in the middleware of the sound reproducing device 1. Then, the script 11 is registered in the middleware of the audio playback device 1 and playback is started (steps S11 and S12).
[0058]
In the reproduction, it is monitored whether there is a voice quality change event (X event) for designating the user tone color parameter 12 in the script 11 (step S13).
When a voice quality change event is found in step S13, the phrase ID assigned to the mora is searched from the user tone parameter 12, and the data corresponding to the phrase ID is read from the user phrase synthesis dictionary 13 and the default synthesis managed by the HV driver. Of the data in the dictionary 19, the dictionary data of the corresponding mora is replaced with the data of the user phrase synthesis dictionary 13 (step S14).
The replacement process in step S14 may be performed in advance before reproduction.
[0059]
When step S14 ends and when no voice quality change event is found in step S13, converter 16 interprets the mora of script 11 (or the script after the replacement process in step S14 if step S14 is performed). Then, the data is finally converted into formant frame sequence data using the HV driver (step S15).
Next, the data converted in step S15 is reproduced by the sound source 20 (step S16).
[0060]
Next, it is determined whether or not the script 11 is finished (step S17). If not finished, the process returns to step S13. If finished, the reproduction operation of the user phrase dictionary is finished.
[0061]
Next, a method for producing the SMAF file 30 shown in FIG. 4 will be described with reference to FIG. FIG. 9 is a flowchart showing a method for producing the SMAF file 30.
First, the user phrase synthesis dictionary 13, the user tone color parameter 12, and the script 11 are produced according to the procedure shown in FIG. 7 (step S21).
[0062]
Next, the SMF file 41 including an event for controlling the music data and the pronunciation of the HV script is produced (step S22).
Next, the SMF file 41 is inputted to the HV authoring tool 42 shown in FIG. 5, and the SMF file 41 is converted into the SMAF file 43 (corresponding to the SMAF file 30) by the HV authoring tool 42 (step S23).
[0063]
Then, the user tone color parameter 12 created in step S21 is put into the HV setup data chunk 37 of the HV chunk 36 of the SMAF file 30 shown in FIG. 4, and the user phrase synthesis dictionary 13 created in step S21 is stored in the same SMAF. It is put into the HV user / phrase / dictionary chunk 38 of the HV chunk 36 of the file 30 and outputted as the SMAF file 30 (step S24).
[0064]
Next, a method for reproducing the SMAF file 30 will be described with reference to FIG. FIG. 10 is a flowchart showing a method of reproducing the SMAF file 30.
First, the SMAF file 30 is registered in the middleware of the audio reproduction device 1 shown in FIG. 1 (step S31).
Here, the audio reproducing apparatus 1 normally registers the music data portion in the SMAF file 30 in the middleware music reproducing unit and prepares for reproduction.
[0065]
Next, the audio reproducing device 1 determines whether or not the HV chunk 36 is present in the SMAF file 30 (step S32).
If there is an HV chunk 36 in step S32, the audio reproducing device 1 interprets the contents of the HV chunk 36 (step S33).
Next, the sound reproducing device 1 registers user tone color parameters, registers a user phrase synthesis dictionary, and registers a script (step S34).
[0066]
When there is no HV chunk 36 in step S32, or when the registration in step 34 is completed, the audio reproducing device 1 interprets the chunk of the music part (step S35).
Next, the audio reproducing device 1 performs music reproduction by starting interpretation of the sequence data (actual performance data) in the sequence data chunk 35 in response to the “start” signal (step S36).
[0067]
In this reproduction, in the process of sequentially interpreting events in the sequence data, the audio reproduction device 1 determines whether or not the event is HV note-on (step S37).
If it is determined in step S37 that the HV note is on, the audio reproduction device 1 starts reproducing the HV script data of the HV chunk designated by the HV note on (step S38).
[0068]
After this step S38, the audio reproducing device 1 performs the reproducing operation of the user phrase dictionary shown in FIG.
That is, the audio reproducing device 1 monitors whether or not there is a voice quality change event (X event) that specifies the user tone color parameter 12 in the reproduction in step S38 (step S39).
[0069]
If a voice quality change event is found in step S39, the phrase ID assigned to the mora is searched from the user tone parameter 12 and the data corresponding to the phrase ID is read from the user phrase synthesis dictionary 13 and the default synthesis managed by the HV driver. Of the data in the dictionary 19, the corresponding mora dictionary data is replaced with user phrase dictionary data (step S40).
The replacement process in step S40 may be performed in advance before reproduction.
[0070]
When step S40 is completed, and when no voice quality change event is found in step S39, the converter 16 interprets the script mora and finally converts it to formant frame sequence data using the HV driver (step S40). S41).
[0071]
Next, the audio reproducing device 1 reproduces the data converted in step S41 in the HV portion of the sound source 20 (step S42).
Next, the audio reproducing device 1 determines whether or not the music has ended (step S43). When the music has ended, the audio reproducing device 1 ends the reproduction of the SMAF file 30, and when the music has not ended, the process returns to step S37.
[0072]
In step S37, when the event is not HV note-on, the audio reproducing device 1 converts the event into music source reproduction event data as music data (step S44).
Next, the audio reproducing device 1 reproduces the data converted in step S44 on the music portion of the sound source 20 (step S45).
[0073]
Thus, according to the present embodiment, the method of reproducing by formant synthesis using the resources of the FM sound source has the following three advantages.
First, according to this embodiment, a user can assign a favorite phrase. Thereby, it is possible to perform reproduction closer to the favorite voice color without depending on the fixed dictionary.
Secondly, according to the present embodiment, a part of the default synthesis dictionary 19 is replaced with the user phrase synthesis dictionary 13, so that it is possible to avoid an excessive increase in the data capacity in the audio reproduction device 1. In addition, since a part of the default synthesis dictionary 19 can be replaced with an arbitrary phrase, it is possible to pronounce in units of phrases, and to eliminate the uncomfortable feeling at the joints of the pronunciations that occur in the synthesized speech of conventional pronunciation units. it can.
Thirdly, according to the present embodiment, since an arbitrary phrase can be designated in the HV script, synthesis in mora units and pronunciation in phrase units can be used in combination.
[0074]
Furthermore, according to the present embodiment, it is possible to change the voice color at the formant level as compared with the method of reproducing the waveform data configured by sampling the phrase in advance. According to the present embodiment, although the data size and quality depend on the frame rate, high-quality reproduction can be performed with a much smaller data capacity than the sampling waveform data. Therefore, for example, the audio reproducing device 1 of the present embodiment can be easily incorporated into a mobile communication terminal such as a mobile phone, and the contents of an e-mail can be reproduced with high quality audio.
[0075]
As mentioned above, although embodiment of this invention was explained in full detail with reference to drawings, the specific structure is not restricted to this embodiment, The design change etc. of the range which does not deviate from the summary of this invention are included.
[0076]
【The invention's effect】
As described above, according to the present invention, data held in the pronunciation unit in the synthesis dictionary can be replaced with arbitrary user data, so that a desired phrase can be reproduced with high quality sound.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an audio reproduction device according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating an example in which a phrase ID is assigned to each sound generation unit.
FIG. 3 is a diagram showing an example of the contents of a user phrase synthesis dictionary.
FIG. 4 is a diagram showing a format of a SMAF file.
FIG. 5 is a functional image diagram showing an example of an HV authoring tool.
FIG. 6 is a block diagram illustrating an example of a mobile communication terminal including the audio reproduction device according to the present embodiment.
FIG. 7 is a flowchart of a method for creating a user phrase synthesis dictionary.
FIG. 8 is a flowchart of a reproduction operation of a user phrase synthesis dictionary.
FIG. 9 is a flowchart illustrating a method for producing a SMAF file.
10 is a flowchart of a method for reproducing a SMAF file 30. FIG.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Voice reproduction apparatus, 11 ... Script, 12 ... User tone color parameter, 13 ... User phrase synthetic | combination dictionary (variable length), 14 ... Application, 15 ... Middleware API, 16 ... Converter, 17 ... Driver, 18 ... Default tone color parameter, 19 ... Default composition dictionary, 20 ... Sound source, 30 ... SMAF file, 31 ... Content info chunk, 32 ... Optional data chunk, 33 ... Track chunk, 34 ... Setup data chunk, 35 ... Sequence data Chunk, 36 ... HV chunk, 37 ... HV setup data chunk, 38 ... HV user phrase / dictionary chunk, 39 ... HV-S chunk, 41 ... SMF file, 42 ... HV authoring tool, 43 ... SMAF file, 44 ... HV script UI 45 ... HV voice editor 46 ... waveform analyzer 47 ... HV script file 48 ... HV user tone file 49 ... HV user synthesis dictionary file 50 ... sound file

Claims (6)

In a voice reproduction apparatus that has a synthesis dictionary that is a database that holds formant frame data corresponding to pronunciation units in advance, and that synthesizes speech using the synthesis dictionary by being given information in which pronunciation units are listed,
Substitution means for replacing formant frame data in pronunciation units held in the synthesis dictionary with user data that is formant frame data acquired in phrase units ;
And a voice synthesizing unit that synthesizes a voice using the synthesis dictionary in which retained data is replaced by the replacing unit when the information in which the pronunciation units are arranged is given. The user data, sound reproducing apparatus according to claim 1, characterized in that it is added to the tone color parameters for processing the formant frame data held in the synthesis dictionary. When the timbre parameter to which the user data is added is given and the timbre parameter is designated at the time of reproduction, the replacement means converts formant frame data, which is data held in the synthesis dictionary, to the user data. Replace,
Said speech synthesis means, when the enumeration information of the speech units given according to any one of claims 1 or 2, characterized in that the speech synthesized using synthesis dictionary is replaced by the tone color parameter Audio playback device. The audio playback device
The user exchange is included in a data exchange format that defines an information structure for reproducing desired data by synchronizing music and voice, and voice is synthesized using the data exchange format. The audio reproduction device according to any one of claims 1 to 3 . The audio playback device
The music piece information included in the information configured as the data exchange format is reproduced as it is, and the audio information included in the information is reproduced using the replacing unit and the voice synthesizing unit. The sound reproducing device according to claim 4 . 6. The audio reproducing apparatus according to claim 4 , wherein the data exchange format includes information obtained by adding the user data to an audio parameter.

Images