JP2009294570A - Speech creation device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a speech creation device for creating speech and singing voice by vibration of a mechanical vibrator instead of an electrical signal. <P>SOLUTION: The speech creation device includes: a plurality of vibrators 101 for creating a sound wave with an optional band width centering on each predetermined folmant frequency; a drum 102 which is provided around the vibrator 101 so as to freely be rotated; a plurality of pins 104 fixed on the drum 102 in order to vibrate the plurality of vibrators 101 corresponding to each set of folmant frequencies of each phoneme of the speech to be created, by sequentially picking them; and a handle 103 for rotating the drum 102. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a sound generation device that generates sound regardless of an electrical signal.

A mechanical music box is composed of a plurality of vibration valves arranged in a comb-like shape and a drum adjacent to the vibration valve, and the vibration valve is played by pins arranged on the drum according to the rotation of the drum, thereby playing music. (For example, see Patent Document 1). Each vibration valve has a different fundamental frequency, and by adjusting the shape of the vibration valve so that the difference between the fundamental frequencies of adjacent vibration valves is 1/12 octave, the sound of any scale including semitones can be obtained. It is configured to be generated. It is possible to set the scale and the timing of sound generation by setting the pin arrangement on the drum, so you can play any song by setting the pins appropriately according to the score of the song It is.
Japanese Patent No. 2835363

  In the conventional mechanical music box, music can be played, but there is a problem that it is difficult to generate voice and singing voice by vibration of a mechanical vibrator regardless of an electrical signal.

  Therefore, in view of the above problems, the present invention provides a sound generation device that generates sound regardless of an electrical signal.

  The present invention includes a plurality of transducers generated by vibrating a sound wave having an arbitrary bandwidth centered on each predetermined formant frequency, a substrate provided in the vicinity of the plurality of transducers, A plurality of pins implanted in the substrate to vibrate the plurality of transducers corresponding to the combination of the formant frequencies of each phoneme of the generated speech simultaneously or successively in order; A drive unit configured to move the substrate or the vibrator relative to each other and vibrate the plurality of vibrators in the order by the plurality of pins to generate the sound. It is a generation device.

  The present invention also provides a plurality of transducers generated by vibrating a sound wave having an arbitrary bandwidth centered on each predetermined formant frequency, and each formant frequency of each phoneme of the generated speech. A plurality of transducers corresponding to a combination of the substrate and a substrate implanted in an order to vibrate simultaneously or continuously, and generating a sound by vibrating the plurality of transducers by playing in the order A voice generation device having a member.

  According to the present invention, it is possible to generate sound by vibration of a vibrator regardless of an electrical signal.

  Hereinafter, an audio generation device according to an embodiment of the present invention will be described.

(First embodiment)
A voice generation device according to a first embodiment will be described with reference to FIGS.

(1) Configuration of Speech Generation Device FIG. 1 is an overview diagram of a speech generation device according to this embodiment. This sound generation device includes a diaphragm 105, a vibrator 101, a drum 102, a handle 103, and a pin 104. The diaphragm 105 and the vibrator 101 are fixed on a base 106, and the drum 102 is rotatably attached. It has been.

  The diaphragm 105 is fixed horizontally on the base 106. A plurality of long plate-like vibrators 101 are connected to the diaphragm 105 in a comb-like shape at equal intervals, and the end points of the tips of the vibrators 101 are arranged in a straight line. In the case of FIG. 1, the length of the vibrator 101 is increased in order. The vibrator 101 generates a sound wave by flipping its end point and vibrating it, so that its power spectrum is concentrated in the vicinity of a predetermined center frequency and the power of the harmonics is reduced. Material and shape are set. The material of the vibrator 101 and the diaphragm 105 is, for example, metal or resin. The center frequency of each transducer 101 is set at equal intervals so as to cover the fluctuation range of the voice formant frequency. In the present embodiment, the center frequency of each transducer 101 is set so as to be 100 Hz from 100 Hz to 3000 Hz (3 kHz). That is, each vibrator 101 vibrates at a frequency of an arbitrary bandwidth around the formant frequency.

  The drum 102 has a cylindrical shape, is rotatably installed between a pair of support plates 107 and 107 erected from the base 106, and the outer peripheral surface of the drum 102 is a straight line in which end points of the vibrator 101 are arranged. Adjacent to each other.

  In addition, a handle 103 is connected to the center of the side surface of the drum 102 as a drive unit for rotating the drum 102 around the axis. By rotating the handle 103, the outer peripheral surface of the drum 102 and the vibrator 101 can be rotated. The drum 102 can be rotated while keeping the distance from the end point constant.

  The outer peripheral surface of the drum 102 serves as a substrate, and a plurality of pins 104 are implanted in the substrate that is the outer peripheral surface. The pin 104 has the same thickness as the width of the vibrator 101, and the height when planted is set to be larger than the distance between the end point of the vibrator 101 and the drum 102. . When the handle 103 is rotated to rotate the drum 102, the pin 104 comes into contact with the corresponding one of the vibrators 101 and repels the vibrator 101 to vibrate.

(2) Arrangement of Pins 104 Next, the arrangement of the pins 104, which is a feature of this embodiment, will be described.

  FIG. 2A shows a state in which the side surface of the drum 102 is viewed as a bus bar, and FIG. 2B shows a developed view assuming that the drum 102 is cut open by one of the bus bars. In FIG. 2B, the position of the pin on the horizontal axis (the rotating direction of the drum 102) corresponds to the timing at which the vibrator 101 is played, and the position on the vertical axis (the direction of the bus) in FIG. Corresponds to the center frequency of. FIG. 2 shows the arrangement of the pins 104 in the case of saying “Yes”.

  The phoneme of speech is characterized by a combination of formant frequencies. FIG. 3 shows a typical formant frequency of a Japanese vowel. F1 represents the frequency of the first formant, and F2 represents the frequency of the second formant. In FIG. 2, the pins 104 are arranged in accordance with the female formant frequency. In order to utter the phoneme “A”, a pin that simultaneously plays the vibrator 101 having the center frequency close to F1 and F2 of “A” and simultaneously plays the vibrator 101 having the center frequency close to F1 and F2 again after a certain time. 104 arrangement.

  When the phoneme changes from “A” to “I”, it is known that the formant frequency transitions smoothly, and the arrangement of the pin 104 is changed from F1 of “A” to F1 of “I”, and The center frequency of the vibrator 101 smoothly changes from “A” F2 to “I” F2.

  FIG. 4 shows the sound waves generated when the transducers 101 corresponding to F1 and F2 are played with respect to “A” and “I” (see FIG. 4A and FIG. 4B) and play them simultaneously. The waveform of the sound wave (see FIG. 4C) generated when

  Since the pins 104 are arranged at regular intervals in the time direction, the waveform generated when the drum 102 is rotated at a constant speed is, for example, the waveform of FIG. The voice waveform shown in FIG. 5 obtained repeatedly is generated.

(3) Effect According to the sound generation device of the present embodiment, when the user rotates the handle 103 and rotates the drum 102, the plurality of pins 104 implanted in the drum 102 respectively play the vibrator 101. Since the speech waveform is generated as described above, it is possible to generate speech without giving an electrical signal as in the conventional case.

(4) Modification 1
In the above embodiment, the interval between the pins 104 in the rotation direction of the drum 102 has a correlation with the pitch period of the sound to be generated. Therefore, the interval may be set according to the pitch period of the sound to be generated. The shorter the interval, the shorter the pitch period (higher voice), and the longer the interval, the longer the pitch period (lower voice).

(5) Modification 2
In the above embodiment, the center frequency interval of the vibrator 101 is set to 100 Hz, but the present invention is not limited to this.

  By making the interval finer, it is possible to express smoother formant transitions and to generate high-quality speech.

  Further, the frequency interval may be equal in the logarithmic region, or may be set to 1/20 octave interval that is finer than 1/12 octave of the conventional music box.

(6) Modification 3
In the above embodiment, when the formant frequency does not change at a continuous pitch cycle, the next pin 104 comes into contact with the vibrator 101 where the vibration has not converged, and the vibration of the previous cycle is stopped. Vibration will start. In this case, a phenomenon in which waveforms of adjacent periods overlap as in the case of actual speech is not reproduced.

  Therefore, a plurality of vibrators 101 having the same center frequency may be provided. As a result, even when the formant frequency does not change, it is possible to reproduce a phenomenon in which waveforms of adjacent periods overlap by implanting pins 104 so as to alternately play a plurality of transducers 101 having the same center frequency. It becomes possible, and a more natural voice can be generated.

(7) Modification 4
In the above embodiment, the drum 102 is driven to rotate by the handle 103, but instead, the drum 102 may be rotated by forming a drive unit by a motor or a spring.

(8) Modification 5
Instead of implanting the pins 104 on the drum 102 as in the above embodiment, the pins 104 are implanted on a substrate of flat plate or more, and the diaphragm 105 having a plurality of vibrators 101 is moved along the substrate, Audio may be generated.

(Second Embodiment)
Next, a voice generation device according to a second embodiment of the present invention will be described with reference to FIGS.

(1) Configuration of Speech Generation Device The speech generation device according to the present embodiment includes a substrate 201, a vibrator 202, and a spatula 203 on the upper surface of a kamaboko base 200, as shown in FIG.

  The substrate 201 is provided on the upper surface of the semi-cylindrical base 200, and as shown in FIG. 6, the t-axis direction is a straight line and the f-axis direction is a curved surface.

  A plurality of pin-shaped vibrators 202 are implanted in the substrate 201. The sound wave generated when the vibrator 202 is played is the same as that of the vibrator 101 in the first embodiment. The vibrator 202 is implanted so that the end points at the upper end of the vibrator 202 are aligned on one plane, and the lengths of the vibrators 202 are different so that the generated sound waves are different. Then, sound is generated by sequentially playing the end points at the upper end of the vibrator 202 while moving the plate-like spatula 203 in one direction.

  As shown in FIG. 7, an arbitrary interval (f1, f2, f3) is formed on a straight line 204 parallel to the f axis so that the vibrator 202 having a formant frequency corresponding to the phoneme of the generated speech is played at the same time. ,...), And a certain distance (t1, t2, t3,...) Is placed in the direction of the t-axis orthogonal to the straight line 204 to generate a speech waveform of the next period. A vibrator 202 for this purpose is arranged on a straight line 205. Similarly, other vibrators 202 are periodically arranged on the substrate 201.

(2) Operational State As shown in FIG. 6, the oscillator 202 is sequentially played while moving the spatula 203 at a constant speed in the direction of the t-axis, whereby a periodic characteristic having a formant frequency characteristic corresponding to a phoneme is obtained. A waveform is generated. In the transitional part where one phoneme changes to the next phoneme, the combination of the center frequencies of the transducers 202 corresponding to each period may be determined so that the formant frequency changes smoothly.

(3) Modification 1
Since the interval between the positions where the transducers 202 are arranged, such as the interval between the straight line 204 and the straight line 205, has a correlation with the pitch period of the generated voice, it is set to an interval according to the pitch period of the generated voice. May be. The shorter the interval, the shorter the pitch period (higher voice), and the longer the interval, the longer the pitch period (lower voice). In the example of FIG. 6, when playing from the left to the right of the plan view (from t1 to t6), since the interval is gradually increased, the pitch period is gradually increased (the voice gradually becomes lower). Generated.

(4) Modification 2
In the present embodiment, the spatula 203 is used to sequentially play the vibrator 202. However, the sound generating member is not limited to the spatula 203. For example, a finger or a coin can be used to play the vibrator 202. Anything can be used as long as it can start vibration.

  When it is assumed that an object whose playing part is not linear, such as a coin, is used, the end points of the vibrator 202 may be set to be aligned on a curved surface instead of on a plane.

(Example of change)
The present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the gist thereof.

(1) Modification 1
In each of the above-described embodiments, the pin 104 and the vibrator 202 are arranged so as to simultaneously play the vibrator 202 corresponding to the combination of formant frequencies. However, the pins 104 and the vibrator 202 do not necessarily have to be at the same time. The pins 104 and the vibrator 202 may be arranged so that they can be repelled.

  In the case of the second embodiment, the vibrators 202 may be arranged on a straight line or a curve, and the vibrators 202 corresponding to the combination of formant frequencies may be continuously played.

(2) Modification example 2
In each of the above-described embodiments, examples of generating vowels have been described, but consonants can also be generated. A voiced consonant is a periodic waveform similar to a vowel, and its phonology is characterized by a combination of formant frequencies and its change pattern, and thus can be generated in the same manner as a vowel.

  The arrangement of the pins 104 or the vibrators 202 at this time may be set based on a formant frequency change pattern obtained by analyzing actual speech.

  In addition, the unvoiced consonant is characterized by a combination of formant frequencies, but has a non-periodic waveform. Therefore, in the first embodiment, the arrangement of the pins 104 in the time direction (the rotation direction of the drum 102) is non-periodic. In addition, in the second embodiment, it can be generated by making the arrangement of the vibrators 202 non-periodic.

(3) Modification 3
In actual speech, the power of each formant is not uniform, and the power of each formant is also characteristic for each phoneme.

  Therefore, in the first embodiment, the length, shape, material, and the like of the pin 104 may be set so that the formant with higher power plays the vibrator 202 more strongly.

  For example, by increasing the length (protrusion amount) of the pin 104 as the formant has a higher power, the amplitude of the vibrator 202 increases and the formant power increases.

  In the second embodiment, the length, shape, material, and the like of the corresponding vibrator 202 may be set according to the formant power.

  In each of the above-described embodiments, it has been described that the end points of the vibrator 202 are arranged on one plane. However, the vibrator 202 corresponding to a formant having a large power is set to have a long length so as to protrude from the plane. And may be played with a large amplitude.

It is a general-view figure of the audio | voice production | generation apparatus of the 1st Embodiment of this invention. (A) shows a state in which the side surface of the drum is viewed as a bus bar, and (b) shows a development view assuming that the drum is cut open at one of the bus bars. It is a table | surface figure which shows the typical formant frequency of a Japanese vowel. (A) and (b) are waveform diagrams of sound waves generated when the transducers corresponding to F1 and F2 are played for “A” and “I”, and (c) shows that they are played simultaneously. It is a wave form diagram of the sound wave generated at the time. It is a wave form diagram of the sound produced | generated by the audio | voice production | generation apparatus of 1st Embodiment. It is a perspective view of the audio | voice production | generation apparatus of 2nd Embodiment. It is explanatory drawing which shows arrangement | positioning of the vibrator | oscillator of 2nd Embodiment.

Explanation of symbols

101 vibrator 102 drum 103 handle 104 pin 105 diaphragm

Claims (6)

A plurality of transducers generated by vibrating a sound wave having an arbitrary bandwidth centered on each predetermined formant frequency;
A substrate provided in the vicinity of a plurality of the vibrators;
A plurality of pins implanted in the substrate to vibrate and vibrate a plurality of the transducers corresponding to each formant frequency combination of each phoneme of the generated speech simultaneously or sequentially,
A drive unit that generates the sound by relatively moving either the substrate or the vibrator, and vibrating the plurality of vibrators in the order by the plurality of pins;
A voice generation device. A plurality of transducers generated by vibrating a sound wave having an arbitrary bandwidth centered on each predetermined formant frequency;
A plurality of the transducers corresponding to the combination of formant frequencies of each phoneme of the generated speech, and a substrate that is implanted in the order of vibrating simultaneously or continuously;
A generating member that vibrates by playing a plurality of the vibrators in the order, and generates the sound;
A voice generation device. A plurality of the pins are periodically implanted on the substrate at intervals corresponding to the pitch period of the sound.
The speech generation device according to claim 1.   The speech generation apparatus according to claim 1, wherein a difference between the formant frequencies corresponding to a plurality of the vibrators is smaller than 1/12 octave.   The speech generation apparatus according to claim 1, comprising a plurality of the vibrators having the same formant frequency.   The speech generation device according to claim 2, wherein the plurality of transducers are periodically implanted on the substrate at intervals corresponding to the pitch period of the speech.

Images