JP4647110B2 - Matheries similar sound conversion device and Matheries similar sound conversion method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention is, for example, similar to a mazary for obtaining sounds similar to a mazarise as a voice that particularly interests a 0-year-old child, particularly an infant up to about 10 months of age, by converting it from an input audio or audio signal. The present invention relates to a sound conversion method and a conversion device.
[0002]
[Prior art]
  Conventionally, toys with sound stimulation for babies have been widely used, such as toys whistle and drums.
  By the way, in recent years, the tendency of sensibility has been studied through sounds that infants pay attention to or listen to.
[0003]
  In other words, infants up to about 10 months of age are becoming particularly interested in such infants with regard to continuously changing sound stimulation as a characteristic of talking to mothers specific to this period. The characteristic of sound stimulation is the voice of the childcare personSpecialIt is called “Mazaly's”.
[0004]
[Problems to be solved by the invention]
  However, with respect to conventional toys that produce sounds such as whistle and drums, no particular consideration has been given to the point of interest, especially corresponding to the developmental stage of infants. For this reason, toys and the like that have been conceived so far can generate sounds similar to the Mazaries as described above to give appropriate auditory stimuli to infants and perform external actions in accordance with their interest trends. It was not even done.
  Further, in relation to the development process of the language ability of the infant and the voice stimulation, a means for generating a voice and selecting the appropriate voice to stimulate the development is not devised.
[0005]
  The present invention has been made to solve such a problem, and obtains a sound similar to Matheries, which can be particularly interesting to infants, by converting other sounds or sound signals input to the apparatus. It is an object of the present invention to provide a Matheries-like sound conversion device and a Matheries-like sound conversion method.
[0006]
[Means for Solving the Problems]
  The present inventionVoice or voice signal or information input means corresponding to the voice signal, basic sound converting means for converting the inputted voice or voice signal into a basic sound having a predetermined fundamental frequency and its harmonic structure, and a predetermined harmonic structure A change sound generating means for providing a change sound by giving a certain range of intonation change for the basic sound comprising:The fundamental sound converting means specifies the fundamental frequency of the input voice or audio signal, and the fundamental sound converting means shifts the fundamental frequency so as to increase or decrease, thereby converting the converted fundamental frequency. Means for generating a fundamental sound having a harmonic structure based on the fundamental frequency, and a resonance adjusting means for adjusting the sound pressure to become weaker as the frequency becomes higher while the frequency having the harmonic structure gradually increases.And converting the input voice or voice signal to generate a Mazaise-like sound.
[0007]
  the aboveAccording to the configuration, the input means is for inputting a voice or a voice signal or information corresponding to the voice signal to be converted into a Mazaise-like sound. The basic sound conversion means functions to match the basic frequency of the Mazaise-like sound by converting or specifying the basic frequency of the input voice or voice signal or information corresponding to the voice signal. Further, the change sound generation means gives an intonation change so as to conform to the characteristics of the mazalies described later.
  Here, the “fundamental sound” of the present invention is a sound having a fundamental frequency suitable for a Mazaise-like sound and having a harmonic structure of the fundamental frequency. Further, the “change sound” is a sound having a certain range of intonation changes corresponding to the mazalies.
  “Information corresponding to a voice signal” means a language that is not a voice or a voice signal but is input instead of voice, for example, and changes to voice input using a keyboard or the like.
  Further, for example, spectrum analysis can be performed on the voice or voice signal taken in by the input means, and the fundamental frequency can be specified from a number of contained frequencies.
  Furthermore, the basic sound conversion means can generate a basic sound having a frequency suitable for the characteristics of the mazalies by increasing or decreasing the basic frequency.
  Then, a resonance adjustment unit that adjusts the frequency to gradually increase and the sound pressure to become weaker as the frequency becomes higher can realize a harmonic structure that conforms to the characteristics of the Matheries.
[0008]
[0009]
[0010]
[0011]
[0012]
  Preferably,The fundamental sound converting means is configured to shift the fundamental frequency of the input voice or speech signal up or down to a fundamental frequency of approximately 300 to 500 Hz.
[0013]
  the aboveAccording to the configuration, the basic sound converting means can generate a basic sound having a frequency that is more suitable for the characteristics of Mazalies by raising or lowering the fundamental frequency in a range of approximately 300 to 500 hertz.
  If the fundamental frequency is less than 300 Hz, the infant is interestedZIf it is extremely low, it becomes an unpleasant stimulus. It has been confirmed that when the fundamental frequency is higher than 500 Hz, the infant is excited more than necessary, and in extreme cases, it may start crying.
[0014]
  Preferably,The change sound generation means applies an intonation change that rises and / or falls within a predetermined range with respect to the input sound or sound signal.
[0015]
  the aboveAccording to the configuration, the change sound generation means can generate an appropriate mazaise similar sound by giving an intonation change that rises or falls within a predetermined range, or rises and falls.
[0016]
  Preferably,The intonation change is set in a range of approximately one octave.
[0017]
  the aboveAccording to the configuration, when the intonation change by the change sound generating means exceeds approximately one octave, it becomes a stimulus that is too strong for an infant and may cry against such a stimulus.
[0018]
  In addition, the present invention includes a step of inputting sound or a sound signal or information corresponding to the sound signal from the outside, a step of identifying a sample sound as a processing unit from the input sound or sound signal, and a basis from the sample sound Identifying a frequency, generating a change sound by applying an intonation change that rises and / or falls within a predetermined range with respect to a sound including the fundamental frequency, and raising or lowering the fundamental frequency with respect to the change sound Shifting, and then adjusting a predetermined harmonic structure to the shifted fundamental frequency,At the stage of adjusting the harmonic structure, the frequency is gradually increased, and the sound pressure is decreased as the frequency becomes higher.Mazaries similar sound conversion methodIt is characterized by being.
[0019]
  the aboveAccording to the configuration, the step of inputting sound or a sound signal or information corresponding to the sound signal from the outside is a step necessary for acquiring the sound or sound signal of the conversion source. The specification of the sample sound is a stage required for specifying the processing unit for conversion. The fundamental frequency is specified for this sample sound. The specified fundamental frequency is given an intonation change that matches the characteristics of the mazalies. Next, make the fundamental frequency conform to the characteristics of Matheries.Adjust the frequency so that the sound pressure becomes weaker as the frequency increases..
[0020]
  Preferably,In specifying the sample sound, the sample sound is specified in units of words based on language recognition.
[0021]
  the aboveAccording to the configuration, when word extraction is performed using language recognition such as Japanese language recognition for input speech or speech signal or information corresponding to the speech signal, a processing unit can be determined relatively easily. .
[0022]
  Preferably, in the step of generating the change sound, the intonation change is set in a range of approximately one octave.
[0023]
[0024]
[0025]
  The present invention also includes a step of inputting sound or a sound signal or information corresponding to the sound signal from the outside, and a step of specifying a sample sound as a processing unit from the input sound or sound signal or information corresponding to the sound signal. A step of specifying a fundamental frequency from the sample sound, a step of generating a change sound by applying an intonation change that rises and / or falls within a predetermined range with respect to the sound including the fundamental frequency, and the change sound Shifting the fundamental frequency to increase or decrease, and then adjusting a predetermined harmonic structure to the shifted fundamental frequency;In the stage of adjusting the harmonic structure, the frequency is gradually increased and the sound pressure is decreased as the frequency becomes higher.It is a computer-readable recording medium storing a program.
[0026]
DETAILED DESCRIPTION OF THE INVENTION
  Preferred embodiments of the present invention will be described below in detail with reference to the accompanying drawings.
  The embodiments described below are preferable specific examples of the present invention, and thus various technically preferable limitations are given. However, the scope of the present invention particularly limits the present invention in the following description. As long as there is no description of the effect, it is not restricted to these aspects.
[0027]
  FIG. 1 is a perspective view illustrating an example of the appearance of a Mazaise-like sound converter according to an embodiment of the present invention, and FIG. 2 is a block configuration diagram illustrating an electrical configuration thereof.
  In FIG. 1, the Mothers-like sound conversion device 10 includes a main body 12, for example, a microphone 11 as input means connected to the main body 12, and a speaker 23 as output means connected to the main body 12.
[0028]
  The configuration of the Mazaise-like sound converter 10 will be described in detail with reference to FIG.
  The mazaise-similar sound conversion apparatus 10 includes a sound or sound signal input means 11, a basic sound conversion means 30 for converting the input sound or sound signal into a basic sound having a predetermined fundamental frequency and its harmonic structure, and a predetermined sound. And a change sound generation means 40 for changing the intonation of a certain range to generate a change sound for the basic sound having the overtone structure.
[0029]
  In the present embodiment, for example, as shown in FIG. 2, the basic sound conversion means 30 includes a speech recognition means 14, a memory 15, a voiceprint analysis means 16, a pitch shift unit 18, and a harmonic filter. The change sound generation means 40 includes the voice recognition means 14, a memory 15, and an intonation shift unit 17.
[0030]
  More specifically, in the figure, the input means 11 is provided with a voice or voice signal input means 11 for inputting a voice or voice signal to be converted or information corresponding to the voice signal. A processing means to which the input means 11 is connected is provided. Such a configuration of the main body 12 may form, for example, an electric or electronic circuit as a synthesizer having each function to be described later. For example, a necessary storage means is mainly provided around a central processing unit (CPU). 2 may be realized by the CPU and the storage means by storing and executing a software program having a processing procedure to be described later. In this case, all the functions may be realized by software, a part may be realized by a dedicated circuit, and the other configuration may be realized by software. Alternatively, by using an information processing terminal such as a personal computer to store software having a processing procedure at each stage described later, the Matheries similar sound conversion apparatus 10 may be realized on the information processing terminal. In this case, such software can be supplied by being stored in a computer-readable recording medium such as a floppy disk, hard disk, or magneto-optical disk.
[0031]
  The main body 12 is provided with a control unit 13. The control unit 13 is realized by, for example, a partial function of the CPU or a control board, and is connected to each unit in the main body to control the operation of each unit.
  Various modes can be considered for the input means 11 connected to the main body 12. For example, when the Mazaise-like sound conversion device 10 is a portable device in the form as shown in FIG. 1 or incorporated in a handy loudspeaker or the like, a sound collecting microphone is used as the input means 11. Further, when the Mazaise-like sound conversion device 10 is realized by software stored in a personal computer, such a case is obtained when it is an input terminal of a computer main body or connected to a communication link such as the Internet. Terminals used for connection are used. Further, in the case where the Mazaries similar sound conversion device 10 is realized in a mode of being stored in the device attached to the function of another device such as a video deck, for example, a video signal input terminal or a video in the device. A signal input terminal that separates the audio signal from the signal is used.
[0032]
  The voice recognition means 14 is connected to the input means 11 directly or via the control means 13. The voice recognition means 14 is a means for specifying the processing unit of the input voice or voice signal. For this reason, the voice recognition means 14 simply cuts out a voice or a voice signal by using a time constant, or preferably, an A / D (analog-digital) so that a language recognition function such as Japanese language recognition can be used. A conversion means is provided for recognizing the language from the input speech signal and specifying the conversion decision in accordance with word unit, phrase unit, predetermined unit decision rule by part-of-speech decomposition, or the like. From the input means 11DesiWhen an audio signal is sent as a tall signal, the A / D conversion means is not used.
[0033]
  As the memory 15, for example, a storage device built in the device, a RAM of a computer and / or a built-in hard disk, a specific area of an external storage unit, or the like is used. The memory 15 is connected to the voice recognition means 14 via the control means 13 or directly, and stores voice signal data separated for each processing unit determined by the direct voice recognition means 14. In processing, the voice signal data is sent to the voiceprint analysis means 16 in accordance with an instruction from the control unit 13.
  The voiceprint analysis unit 16 is connected to the memory 15 via the control unit 13 or further to the intonation shift unit 17 and the pitch shift unit 18. As will be described later, the voiceprint analysis means 16 determines the fundamental frequency of the voice signal data for each processing unit read from the memory 15 or sent directly from the voice recognition means 14 by voiceprint analysis. ing. Therefore, the voiceprint analysis means 16 can use, for example, an equalizer or software that performs a frequency spectrum analysis function.
[0034]
  The intonation shift unit 17 is directly connected to the voice print analysis unit 16 via the control unit 13, and further connected to the pitch shift unit 18 and the speed adjustment unit 19. The intonation shift unit 17 is the center of the change sound generation means 40 of the present invention. As will be described later, for example, a fundamental frequency is determined for an audio signal as a processing unit, and then the later-described mazalies for the audio signal is determined. In order to conform to the above characteristics, a change sound to which an intonation change that rises and / or falls within a predetermined range, for example, a range of one octave is generated. In other words, with respect to the voice that is the processing unit, the specific portion is pitch-shifted up and / or down within a range of one octave.
[0035]
  The pitch shift unit 18 is connected to the control unit 13 or directly to the intonation shift unit 17 and further to the speed adjustment unit 19. For example, the pitch shifter 18 may be configured using software having a function of processing so that the frequency gradually changes along a time series, and having a function approximate to this. The pitch shifter 18 is responsible for the central function of the basic sound converting means 30, and the fundamental frequency of the change sound to which the intonation change is applied is set to, for example, 300 hertz or so so as to conform to the characteristics of later-described Mothers. A predetermined frequency shift is performed so as to obtain a relatively high frequency of about 500 hertz.
[0036]
  Specifically, the pitch shift unit 18 shifts the frequency of the sound as a processing unit by, for example, increasing the sound pressure of a predetermined frequency component selected using a bandpass filter using an amplifying unit. It is also possible to perform a certain frequency shift for all the components constituting this sound.
  Further, after the frequency shift, as described later, a harmonic structure is added by the harmonic filter 24, and it is more preferable that the harmonic pressure gradually decreases as the frequency increases. A harmonic structure that lowers the sound pressure is generated.
[0037]
  The speed adjustment unit 19 is connected to the pitch shift unit 18 or the intonation shift unit 17 and adjusts the sound length of the audio signal subjected to the intonation change.
[0038]
  Since the speed adjusting unit 19 outputs a sound signal of a mazaise-like sound as described later, preferably, a sound signal of a mazaise-like sound output from the speed adjusting unit 19 is provided in the main body 12. A memory 20 is provided for storing according to 13 instructions. The memory 20 may use the same storage means as the memory 15.
[0039]
  2, the output means 21 is formed only by the speaker 23 and is connected to the D / A converter 22 on the main body 12 side. The D / A converter 22 converts the audio signal of the Mazaise-like sound output from the speed adjusting unit 19 or the memory 20 into an analog signal in the main body 12 and sends it to the speaker 23. The speaker 23 is configured to output a Mazaise-like sound based on the analog-converted audio signal. The D / A converter 22 may be provided on the output means side.
[0040]
  The Mazaise-like sound conversion device 10 of this embodiment is configured as described above. Before describing a method for converting an input voice or audio signal into a Mazaise-like sound using this device, first, the characteristics of the Mazalies will be described. Will be described.
  FIG. 3 is an oscilloscope waveform showing the mazaries corresponding to the voices etc. when a mother or other caregiver speaks to the infant. Specifically, when the mother talks to the infant as “Oishii Ne”. Is. FIG. 4 is a graph in which the fundamental frequency characteristics are plotted, and FIG. 5 is a graph in which the sound pressure, that is, the loudness of the sound in FIG. 4 is plotted.
[0041]
  In particular, as shown in FIG. 4, the fundamental frequency is a voice belonging to a range of approximately 200 Hz to 500 Hz, and is shown at a relatively high pitch (frequency). (Basic sound).
[0042]
  As shown in FIGS. 4 and 5, this frequency is a changing sound that changes within a relatively short sound length of 0.4 to 1.0 seconds, and is within approximately one octave. The range has changed.
[0043]
  Further, referring to FIGS. 6 and 7, the overtone structure of Mazalies can be understood. That is, as shown in FIG. 6, this sound has not only the sound of S1 corresponding to FIG. 4 but also a harmonic structure such as second overtone S2, third overtone S3,. Here, although this sound has a harmonic structure, it is at most up to about S4 and S5, and is not a sound that resonates excessively. FIG. 7 is a graph showing the frequency scale of FIG. 6 changed, and is a diagram for facilitating understanding of the state of the basic sound S1 and the state of its harmonics. Further, FIG. 8 is a graph showing a gentle change by processing the data of FIG.
[0044]
  9 to 12 are diagrams for explaining the characteristics of the Mazarez harmonic structure as human speech.
  In these figures, the horizontal axis indicates the frequency increasing toward the right, and the vertical axis indicates the sound pressure in dB.
  FIG. 9 shows a sound source wave spectrum of speech. As shown in the figure, there are many overtones, and the sound pressure of overtones has a curve (PP line) connecting the peaks. Draw down. In FIG. 9, the sound shifted so that the sound with the highest sound pressure has a predetermined high frequency is the basic sound of the present invention.
[0045]
  FIG. 10 schematically shows the transfer characteristics based on the transfer function of the human vocal tract when the sound of “a” is uttered, that is, in a state where the vocal tract is not so narrowed. Compared with FIG. 11 showing the transfer function, it is characteristic that the sound pressure of overtones with a constant interval is increased. As a result, the spectrum of the basic sound shown in FIG. 9 becomes as shown in FIG. 12 due to the transfer characteristics shown in FIG. That is, in the harmonic structure, as the frequency increases, the PP line connecting the peak of the sound pressures of the respective harmonics becomes a synthesized sound showing a curvilinear descent as in FIG. Such a sound is referred to as a reference sound, and such characteristics are given as described later.
[0046]
  On the premise of the above, in the present embodiment, when generating a Mazaise-like sound, the sound has the following characteristics.
  (1) Mazalies is a voice with a high pitch, that is, a high fundamental frequency (basic frequency = F0) and a high key, and the frequency is about 300 Hz to 500 Hz, and the average is about 350 Hz. .
  (2) Mazalies have a harmonic structure and are therefore reverberant sounds. For this reason, in the present embodiment, in order to generate the Matheries-like sound, as will be described later, based on the function of the basic sound conversion means 30 so as to meet the above-mentioned condition (1) and the condition of this overtone, Generate basic sounds. However, this overtone, if the high frequency range is emphasized too much, may become a sound that resonates excessively and may not be a “friendly voice”. For this reason, in particular, the basic sound is changed to be a reference sound having the characteristics shown in FIG.
  (3) In addition, the mazalies are sounds with a large difference in sound level, a fundamental frequency changing, and a sound rich in intonation. In the present embodiment, this is called “change sound” and is realized by the function of the change sound generation means 40 as described later. However, the change is preferably within one octave, and if it exceeds this, there is a possibility that the voice is not “friendly”.
[0047]
  FIG. 13 is a flowchart showing an example of a method for converting an input voice or audio signal into a Mazaise-like sound having the above-mentioned Mazalies characteristics by the Mazaise-like sound converter 10 based on the configuration of FIG. .
  With reference to FIG. 13, an example of a method for converting a Mazaise-like sound will be described.
[0048]
  A voice or a voice signal is input to the input means 11 of FIG. 2 (ST1).
  For example, in the case of a device that directly inputs sound, the sound is input from a microphone 11 or a handy loudspeaker as shown in FIG. In the case of a conversion device built in a video deck or a conversion device connected to the video deck, an audio signal separated from the video signal is input. When the conversion apparatus 10 is connected to a network or the like, audio signals such as various contents distributed via the network are input. In addition, an audio signal corresponding to a reproduction signal reproduced from a recorded tape recorder, compact disk, magneto-optical disk or the like may be input. Further, when the Mazaise-like sound conversion device is built in a television or the like, an audio signal separated from a broadcast wave sent from a broadcast station is input.
[0049]
  When the voice signal from the input means is an analog signal, the voice recognition means 14 in FIG.DesiAfter being converted into a tall signal, for example, a predetermined processing unit such as a word or a phrase is identified by a Japanese language recognition program, and the processing unit 13 stores the processing unit in accordance with an instruction from the control unit 13. . Audio signals divided into processing units (hereinafter referred to as “sample signals”). FIG. 14 shows the state in which such a sound signal is input as an oscillograph waveform, in which the horizontal axis represents time and the vertical axis represents amplitude. In step 2, the audio signal VS is cut out from t1 to t2 to be a sample signal (sample sound) (ST2).
[0050]
  Next, the voiceprint analysis means 16 performs the pitch analysis of the sample sound, that is, the frequency analysis, and specifies the fundamental frequency F0 (ST3). As shown in FIG. 15, the basic sound is a sound having the specified basic frequency F0 and its harmonic structure VS2, VS3,.
  In this embodiment, for the sake of illustration, only the voice VS1 corresponding to the fundamental frequency F0 is extracted and described. However, the following processing is also performed simultaneously for each harmonic structure, except for the case where it is specifically described below. .
[0051]
  Next, an operation for generating a change sound is performed. Here, the intonation shift unit 17 shown in FIG. 2 raises and / or lowers within a predetermined range, for example, one octave range, in order to emphasize the change in the fundamental frequency of the audio VS1 using, for example, a synthesizer function. Grants intonation shift. For example, in FIG. 16, the rate of change is increased to about 0.8 octave, which is CVS1.
[0052]
  Subsequently, the operation shown in FIG. 17 is performed. In FIG. 17, the pitch shift unit 18 in FIG. 2 performs pitch analysis (frequency analysis) on the voice CVS1 to which the intonation change is added, and the average frequency is obtained.AF0Is obtained (ST5).
  This average frequencyAF0Based on the above, the pitch shifter 18 further performs a pitch shift (frequency shift) so as to be AF′0 shown in FIG. 18 (ST6).
  This pitch shift is intended to have a fundamental frequency that approximates to Mothers, and the range of change is preferably set in the range of 300 Hz to 500 Hz. In this embodiment, it is set to 350 Hz. Such a pitch shift can be frequency shifted by increasing the sound pressure of a predetermined frequency component, or a constant frequency shift may be performed for all components constituting the sound. In the present embodiment, the entire CVS1 is raised (see FIG. 18) to obtain the voice SCVS1.
[0053]
  Next, with respect to the voice SCVS1 after the pitch shift, that is, after the frequency change, resonance adjustment of the harmonic part is performed (ST7). Here, the harmonic overtone parts VS2, VS3... With respect to the fundamental frequency are simultaneously performed in the conversion process of the voice VS1 corresponding to the fundamental frequency F0 with respect to the intonation adjustment (ST4) and the pitch shift (ST6).
  The pitch shift unit 18 is preferably connected to a harmonic filter 24 having a resonance adjustment function. As a result, for example, for each overtone that has been pitch-shifted into the intonation change corresponding to the audio SCVS1, a harmonic structure as shown in FIG. 12 is formed using, for example, a transfer function as shown in FIG. That is, the PP line connecting the peak of the sound pressure of each overtone whose frequency increases gradually is set as a reference sound that is a synthesized sound indicating a curvilinear descent.
  In this case, the transfer function is not used directly, and the filter is such that the higher the harmonics, the lower the sound pressure.TYou may adjust using.
[0054]
  Lastly, as a result of the intonation shift, the speed of the voice SCVS1 to which the tone length or speed change is given is adjusted by the speed adjusting unit 19 and returned to the original tone length (ST8).
  In this case, in the case of Mothers, in response to the tendency to speak slowly, the speed may be adjusted to be slower than the input voice by taking this into account when adjusting the speed.
  As a result, an audio signal that is well suited to the Mazaise-like sound is generated. Therefore, the audio signal is stored in the memory 20 as necessary, or is directly converted into an analog signal by the D / A converter 22, and the Mazaise-like sound is output from the speaker 23. Sound is output.
[0055]
  As described above, the Mothers-like sound conversion apparatus 10 according to the present embodiment converts other sounds or sound signals that are input into sounds similar to Mothers that can attract the attention of infants and outputs the sound. Can do. Therefore, for infants or infantsWhenFor example, the present invention can be applied to a very wide range of products as follows.
  For example, it is possible to talk to the infant with a muzzle-like sound from a handy loudspeaker or the like to which voice is directly input.
  In the case of a conversion device with a built-in video deck, the audio signal of the recorded video signal can be reproduced as a mazaise-like sound.
  In addition, when the Mazaise-like sound conversion device 10 is connected to a network or the like, audio signals such as various contents distributed via the network are converted to the Mazaise-like sound, so that it can be reproduced as a Mazaise-like sound. .
  Further, a reproduction signal reproduced from a recorded tape recorder, compact disk, magneto-optical disk or the like is reproduced as a Mazaise-like sound.
  Furthermore, when the Mazaise-like sound converter 10 is built in a television or the like, an audio signal separated from a broadcast wave sent from a broadcasting station can be reproduced as a Mazaise-like sound.
  Furthermore, those who cannot speak Mazarez well can listen to their voice converted to the Mazaries sound by inputting their recorded voice into the input means, and practice speaking and uttering by Mazaries. Can do.
  Further, by using a language input means such as a personal computer, a word processor keyboard or the like as an input means, even a person who cannot speak a language or a person with a disability can speak by the Mothers.
  Furthermore, by using a communication line such as a telephone or the Internet, even if you are in a remote place, you can talk to Mazaries from a remote place.
[0056]
  The present invention is not limited to the above-described embodiment. For example, even if the sound quality is not questioned even if the approximation with the Mazaise similar sound is inferior, the input audio is input to the pitch shift unit, and the frequency is changed so that it immediately matches the Mazalies, A simple apparatus configuration or a conversion method may be used as it is or after adding a predetermined intonation change.
  For example, when Japanese language is recognized by the language recognizing means, a Mazaise-like sound is stored in advance in a memory as table data or the like for each predetermined unit such as a word or a phrase, and such data It is good also as a structure converted using.
  Therefore, for example, when Japanese is recognized by this language recognition means, the intonation in each word is stored in advance in the memory as, for example, table data, and the intonation adjustment is performed with reference to this data. Also good.
  Further, in the above-described embodiment, each configuration may be omitted or may be realized in combination with other configurations not described in the case where it is allowed in relation to the quality of the Mazaise-like sound. May be.
[0057]
【The invention's effect】
  As described above, according to the present invention, a sound similar to a mazarise that can attract the attention of an infant in particular, a mazaise-like sound that can be obtained by converting other sounds or audio signals input to the apparatus. It is possible to provide a conversion device and a method for converting a Mazaise-like sound.
[Brief description of the drawings]
FIG. 1 is a schematic perspective view showing an example of the appearance of an embodiment of a Mazaise-like sound converter.
FIG. 2 is a block diagram showing a configuration of the Mazaise-like sound conversion device of FIG. 1;
FIG. 3 is an oscilloscope waveform chart showing a muzzle's voiceprint.
4 is a graph showing the fundamental frequency of the voiceprint of FIG. 3;
5 is a graph showing the sound pressure of the voice in FIG.
6 is a graph showing overtones of the sound in FIG. 3. FIG.
7 is a graph showing an overtone of the voice of FIG. 3 with an enlarged frequency scale. FIG.
8 is a graph showing the sound pressure of the voice in FIG.
FIG. 9 is a diagram showing a harmonic structure of a basic sound in relation to a human voice in a sound source wave spectrum.
FIG. 10 is a diagram showing a transfer function of human production.
FIG. 11 is a diagram showing a transfer function of an open space.
FIG. 12 is a diagram for explaining a voice spectrum of a Mazaise-like sound.
FIG. 13 is a flowchart showing an example of a conversion method by the Matheries similar sound converter in FIG. 2;
FIG. 14 is an oscilloscope waveform diagram of input voice.
FIG. 15 is a graph showing a fundamental frequency obtained by cutting a processing unit from an input voice.
FIG. 16 is a graph showing a state where a change in intonation is applied to a voice in a processing unit.
FIG. 17 is a graph showing how an average frequency is obtained for speech in units of processing.
FIG. 18 is a graph showing how a frequency shift is performed on audio in units of processing.
[Explanation of symbols]
  DESCRIPTION OF SYMBOLS 10 ... Mazaise similar sound conversion apparatus, 11 ... Input means, 12 ... Main body, 13 ... Control means, 21 ... Output means, 30 ... Basic sound conversion means, 40 ... Change sound generation means

Claims (8)

Input means of information corresponding to voice or voice signal or voice signal;
A basic sound converting means for converting an input voice or audio signal into a basic sound having a predetermined fundamental frequency and its harmonic structure;
A change sound generating means for giving a change in intonation within a certain range for the basic sound having a predetermined harmonic structure;
Means for specifying the fundamental frequency of the input sound or sound signal by the basic sound converting means;
Means for generating a basic sound having a harmonic structure based on the converted fundamental frequency by shifting the fundamental frequency so as to raise or lower the fundamental frequency;
Resonance adjustment means for adjusting the frequency with the harmonic structure to gradually increase and adjusting the sound pressure to become weaker as the frequency becomes higher ,
A Mazaries similar sound conversion device characterized in that the input voice or audio signal is converted to generate a Mazaries similar sound. 2. The mazalies according to claim 1, wherein the basic sound converting means is configured to shift the basic frequency of the input voice or voice signal to a basic frequency of approximately 300 to 500 Hz. Similar sound conversion device. 3. The mazaise-like sound conversion device according to claim 1 or 2, wherein the change sound generation means gives an intonation change that rises and / or falls within a predetermined range with respect to the input sound or sound signal. 4. The Mazaise-like sound converter according to claim 1 , wherein the intonation change is set in a range of approximately one octave. Inputting information corresponding to voice or voice signal or voice signal from outside;
Identifying a sample sound as a processing unit from the input voice or voice signal or information corresponding to the voice signal;
Identifying a fundamental frequency from the sample sound;
Generating a change sound by applying an intonation change that rises and / or falls within a predetermined range with respect to the sound including the fundamental frequency;
Shifting the fundamental frequency to increase or decrease with respect to the change sound;
And adjusting a predetermined harmonic structure to the shifted fundamental frequency,
A method of converting a mazaise-like sound, wherein in the step of adjusting the overtone structure, the frequency is gradually increased and the sound pressure is decreased as the frequency becomes higher . 6. The method of converting a mothers-like sound according to claim 5, wherein the sample sound is specified in units of words based on language recognition. The method for converting a mazaise-like sound according to claim 5 or 6 , wherein intonation change is set in a range of approximately one octave in the step of generating the change sound. Inputting information corresponding to voice or voice signal or voice signal from outside;
Identifying a sample sound as a processing unit from the input voice or voice signal or information corresponding to the voice signal;
Identifying a fundamental frequency from the sample sound;
Generating a change sound by applying an intonation change that rises and / or falls within a predetermined range with respect to the sound including the fundamental frequency;
Shifting the fundamental frequency to increase or decrease with respect to the change sound;
Next, adjusting a predetermined harmonic structure to the shifted fundamental frequency;
With
A computer-readable recording medium characterized by storing a program for adjusting the harmonic structure so that the frequency gradually increases and the sound pressure decreases as the frequency increases .

Images