JPH1056679A - Sound wave detecting device and sound wave processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To convert sound waves to parallel electric signals in a frequency area at high speed in simple configuration by separately detecting a maximum amplitude for each of different frequency components in the plane of an advanced wave type diaphragm generated by sound waves. SOLUTION: A voice 2 is detected by plural electrodes, for which the amplitude at a maximum amplitude position for each frequency component differentially existent in the major axis direction of a diaphragm 14 of an advanced wave type microphone 3 is provided corresponding to the amplitude position, and converted to parallel voice signals S2 in the frequency area at high speed. A maximum value detection part 11 simultaneously detects the maximum values of S2 for each parallel signal in every prescribed sampling term, sends these values to an interface part 12 as voice signals S3 of parallel sampled amplitude signals, encodes these signals according to the format for voice information transmission of a computer 6 and generates serial voice data S1 through time division multiplexing. Based on the data S1, the computer 6 decodes a synthetic value for every sampling term and defines it as the synthetic signal of time sequence. Thus, processing at the computer can be shortened and simplified in simple configuration.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[Table of Contents] The present invention will be described in the following order. BACKGROUND OF THE INVENTION Problems to be Solved by the Invention Means for Solving the Problems (FIG. 2) Embodiments of the Invention (FIGS. 1 to 7) Effects of the Invention

[0002]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound wave detecting device and a sound wave processing device, and is preferably applied to, for example, a microphone.

[0003]

2. Description of the Related Art A conventional microphone, when given a sound, generates a minute electric sound signal in a time domain corresponding to the sound. For this reason, when amplifying a minute electrical audio signal generated by a microphone or transmitting the waveform of the electrical audio signal correctly, great care must be taken to suppress noise and the like.

[0004] By the way, an electrical voice signal generated by a microphone is input to a computer, and from this electrical voice signal, a so-called voice recognition in which the computer is used to mechanically extract the tone content of a word by a computer. Conceivable. In this case, it is necessary for the computer to perform so-called vector quantization, which divides the electrical audio signal supplied from the microphone for each sampling period, and converts time-domain audio data into frequency-domain audio data for each sampling period. At the time of this vector quantization, the computer has performed fast Fourier transform of the audio data in the time domain for each sampling period by software.

[0005]

However, there is a problem that it takes a lot of time to vector quantize the electric voice signal in the time domain obtained from the microphone by the fast Fourier transform. For this reason, vector quantization in real time of an electrical audio signal in the time domain generated by a human continuously uttered at a normal speed is a heavy load on a computer. Further, if the function of vector quantizing the electrical audio signal in the time domain at a predetermined period is to be constituted only by hardware, an enormous number of parallel channels are required and the structure becomes large.

The present invention has been made in view of the above points, and provides a sound wave detecting device capable of converting sound waves into parallel electric signals in the frequency domain at a high speed with a simple configuration, and a sound wave processing device using the same. It is something to propose.

[0007]

According to the present invention, there is provided a vibrating membrane vibrating with a sound wave, and a plurality of amplitude detecting means for detecting respective amplitudes at a plurality of positions on the vibrating membrane. The vibrating membrane is vibrated in the thickness direction by receiving a sound wave using a fluid as a medium, and the maximum amplitude positions of all frequency components included in the sound wave are different from each other in the plane. The plurality of amplitude detection means are arranged corresponding to the plurality of maximum amplitude positions, respectively, and separately detect the respective amplitudes of the plurality of maximum amplitude positions.

The vibrating membrane is vibrated in the thickness direction by a sound wave,
The sound waves are converted into parallel electric signals in the frequency domain by separately detecting the respective amplitudes at the maximum amplitude positions for the respective frequency components differently present in the plane of the vibrating membrane by a plurality of amplitude detecting means. High-speed conversion can be achieved with a simpler configuration than in the past.

According to another aspect of the present invention, there is provided a sound processing electronic apparatus comprising a sound wave electric signal converting means for converting a sound wave into an electric signal, and a signal processing means for processing an output signal of the sound wave electric signal converting means. The means includes a vibrating membrane vibrating with a sound wave and a plurality of amplitude detecting means for detecting respective amplitudes at a plurality of positions on the vibrating membrane. The vibrating membrane is vibrated in the thickness direction by receiving a sound wave using a fluid as a medium, and the maximum amplitude positions of all the frequency components included in the sound wave are different from each other in the plane. The plurality of amplitude detection means are arranged corresponding to the plurality of maximum amplitude positions, respectively.
The respective amplitudes of the plurality of maximum amplitude positions are separately detected.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings.

In FIG. 1, reference numeral 1 denotes a voice recognition device as a sound processing device as a whole, and a voice 2 as a sound wave is given to a traveling wave microphone 3. The traveling wave microphone 3 directly converts the sound 2 into a parallel electric sound signal in the frequency domain in real time, and generates, for example, time-division multiplexed serial sound data S1 based on the electric sound signal. The traveling wave microphone 3 supplies the audio data S1 to a computer 6 as a signal processing means via an interface cable 5.

The computer 6 decodes the vector-quantized audio data based on the audio data S1, compares the audio data with reference data, and recognizes the audio 2 in real time. Further, the computer 6 extracts the meaning and content from the recognized voice 2 and displays the meaning and content of the voice 2 on the display screen 8 of the display unit 7 based on the extraction result.

The traveling wave type microphone 3 receives the sound 2 transmitted through the air as a fluid by the cone-shaped sound collection unit 9 and converts the mechanical vibration of the sound 2 into a sound electric signal as a sound wave detecting device. Give to part 10. Audio electric signal converter 1
0 directly converts this mechanical vibration, which is a signal in the time domain, into an audio signal S2, which is a parallel electrical signal in the frequency domain, in real time, and converts this audio signal S2 into a maximum as an electrical amplitude signal generating means. The value is given to the value detection unit 11.

The maximum value detecting section 11 simultaneously detects the maximum value of each parallel signal of the audio signal S2 for each predetermined sampling period. The maximum value detector 11 supplies the maximum value of each parallel signal of the electric signal S2 as an audio signal S3 as a parallel sampling amplitude signal to an interface unit 12 as analog-to-digital conversion means. The interface unit 12 encodes the audio signal S3 in accordance with the audio information transmission format of the computer 6, and performs time division multiplexing to generate serial audio data S1.

The audio-electrical signal converter 10 has an outer box 13 having a substantially triangular cross section parallel to the plane of the drawing, in which a traveling wave type vibrating film 14 as a vibrating film and an electrode support plate 15 are arranged. Has been established.

As shown in FIG. 2 (A), the outer box 13 has a rectangular top and bottom surfaces, triangular sides, and is formed in a wedge shape as a whole. The outer case 13 includes a traveling wave type diaphragm supporting plate 16 that supports the traveling wave type diaphragm 14,
Both sides are connected by a rectangular electrode support plate (not shown) 15. As a result, the inner space of the outer case 13 that fills the air is divided into two, and the traveling wave type vibrating membrane 1 is divided.
An upper floor room 17 and a lower floor room 18 are respectively arranged above and below the 4. The outer box 13 is provided with a hole 19 at the innermost position, and the upper hole room 17 and the lower floor room 18 are formed by the hole 19.
The inside air is communicated so as not to generate a static air pressure difference.

The outer case 13 is formed of a sufficiently hard material that does not resonate with the vibration of the sound 2 and has a predetermined thickness and a size that accommodates all the traveling wave type vibration film 14. Outer box 1
3 is that the cross-sectional area in a plane orthogonal to the long axis decreases at a predetermined rate toward the back, and the back of the sound is terminated by the hole 19, thereby substantially eliminating the reflection of the sound 2 against the vibration. ing.

The traveling wave type diaphragm supporting plate 16 is made of a sufficiently hard material that does not resonate with the vibration of the sound 2. The traveling wave type vibrating film 14 is formed in an elongated rectangular shape, and two opposing long sides are supported by a traveling wave type vibrating film support plate 16. The traveling wave type vibration film 14 has a low selectivity Q with respect to the vibration frequency of the sound 2.

As a result, the sound 2 is given to the opening 20 of the upper floor room 17 and passes through the hole 19 and the lower floor room 18 to the opening 21 of the lower floor room 18.
Causes vibration in the thickness direction. This vibration is generated by a traveling wave traveling along the long axis of the traveling wave type vibration film 14.

As shown in FIG. 2B, the traveling-wave vibrating membrane 14 has a width of the short side 22 on the opening 20 side of the upper floor chamber 17.
0.1 [mm], width of short side 23 on the innermost side of upper floor room 17 is 0.5 [m
m], and the entire length 24 is formed of a metal thin film of about 35 [mm]. As a result, the traveling wave generated in the traveling wave type vibrating film 14 advances by 35 [mm] every 100 [μsec] and has a frequency component within the human audible frequency band. Traveling wave type vibrating membrane 1
4, the maximum amplitude positions of all the frequency components in this frequency band (20 [Hz] to 20 [KHz]) are present in order of frequency along the long axis of the traveling wave type vibration film 14.

Incidentally, the maximum amplitude position of each of the frequency components included in the traveling wave increases as the frequency increases.
It is close to the short side 22 of the traveling wave type vibration film 14. These shapes and dimensions are set based on data obtained by modeling a human cochlea. Thus, it is considered that the speech 2 can be recognized with the same audible characteristics as a human in the above-described speech recognition.

As shown in FIG. 2C, the electrode support plate 15
On the upper side, a plurality of electrodes 25 as a plurality of amplitude detecting means electrically separated from each other by a fixed minute gap are arranged at equal intervals along the longitudinal direction of the traveling wave type vibrating film 14 (in FIG. Is 2
0 is shown). The entire length of the plurality of electrodes 25 is equal to the total length 24 of the traveling wave type vibration film 14. All the electrodes 25 are opposed in parallel in proximity to the lower surface of the traveling wave type vibration film 14. Further, as shown in FIG. 2D, the plurality of electrodes 25 are formed in an elongated rectangular shape, and the longitudinal direction is arranged in parallel with the short axis direction of the traveling wave type vibration film 14. Thereby, each position of the plurality of electrodes 25 is
It corresponds to the maximum amplitude position for each frequency component existing on the traveling wave type vibration film 14 in order of frequency.

The plurality of electrodes 25 generate electrostatic voltages corresponding to the amplitudes of a plurality of opposing partial regions of the traveling wave type vibrating film 14 and supply the generated electrostatic voltages to the maximum value detecting unit 11. That is, as shown in FIG.
Each potential is applied to the maximum value detection unit 11 at the same time as a parallel audio signal S2. As shown in FIG. 4, the plurality of electrodes 25 are respectively connected to one ends of the plurality of micro capacitors 26 in the maximum value detection unit 11. The other ends of the plurality of minute capacitors 26 are commonly connected to the anode of the constant voltage source 27 in the maximum value detection unit 11. The cathode of the constant voltage source 27 and the traveling wave type vibration film 14 are grounded and have the same potential.

As a result, a small floating capacitance (indicated by a dotted-line capacitor in the figure) is generated between the plurality of electrodes 25 and the traveling-wave-type vibrating membrane 14. Therefore, the plurality of electrodes 25 generate an electrostatic voltage V _C according to the distance from the traveling wave type vibration film 14. The electrostatic voltage V _C is a constant voltage source 27
, The stray capacitance is C _float , the fixed capacitance of the micro capacitor 26 is C _fix , and V _C = E × C
_fix / (C _fix + C _float ).

The stray capacitance C generated at each electrode 25
_{The float} is determined according to the distance between each electrode 25 and the traveling wave type vibrating film 14. This distance is determined by the number of electrodes 2
5 is increased or decreased in accordance with the amplitude of each partial region of the traveling wave type vibration film 14 opposed to 5. Thereby, the plurality of electrodes 25
Generates an electrostatic voltage V _C according to the amplitudes of a plurality of opposing partial regions of the traveling wave type vibration film 14.

As described above, each position of the plurality of electrodes 25 sequentially corresponds to the maximum amplitude position for each frequency component existing on the traveling wave type vibrating membrane 14 in order of frequency. Accordingly, the plurality of electrodes 25 separately discriminate the frequency components of the frequency band corresponding to the traveling wave type vibration film 14 in real time in the order of frequency. Further, as described above, the entire length of the plurality of electrodes 25 is the same as that of the traveling wave type vibration film 14. Thereby, the plurality of electrodes 25 discriminate the same number of frequency components as the electrodes 25 including the frequency components near the upper and lower ends of the frequency band corresponding to the traveling wave type vibrating film 14.

Accordingly, the plurality of electrodes 25 separately discriminate in real time the same number of frequency components as the electrodes 25 including the frequency components near the upper and lower ends of the frequency band of the sound 2 in real time, and each of the discriminated frequency components Voltage V _C according to the amplitude of
Will be generated. A plurality of electrodes 25 gives the generated respective electrostatic voltage V _C separately to a plurality of amplifier 28 as parallel audio signal S2.

The plurality of amplifiers 28 amplify the parallel audio signal S2, respectively, and supply the amplified signals to the maximum value detection circuit 29 as a plurality of sampling circuits as the parallel audio signal S4. The maximum value detection circuit 29 detects the maximum value of the audio signal S4 every predetermined sampling period, for example, every 10 [ms], and holds this maximum value until the next sampling period.
Further, the maximum value detection circuit 29 supplies the plurality of stored maximum values to the interface unit 12 as parallel audio signals S3. The interface unit 12 encodes the audio signal S3 supplied from the maximum value detection circuit 29 into digital data corresponding to the audio signal S3, and performs time division multiplexing to generate serial audio data S1.

Here, it is assumed that f ₁ > f _n and n frequencies f ₁ , f ₂ ,... Of all the frequency components included in the traveling wave.
, F _n−1 , f _n are provided, and a process of obtaining a parallel audio signal S3 from the audio 2 by providing n electrodes 25 will be described. Here, it is assumed that f ₁ is the frequency component having the maximum amplitude near the opening 20, and f _n is the frequency component having the maximum amplitude near the hole 19. And n is
It is set according to the required resolution.

The sound 2 when it reaches the sound collecting section 9 of the traveling wave type microphone 3 has, for example, a signal waveform 30 in the time domain shown in FIG. As shown in FIG. 5B, at a certain time t ₁ , the amplitude of the traveling wave type vibration film 14 due to the low frequency component 31 in the frequency domain of the traveling wave on the traveling wave type vibration film 14 changes at the opening 20 side. It becomes smaller as it moves toward the hole 19 side. The amplitude of the traveling wave type vibration film 14 due to the low frequency component 31 is the same as that of the hole 19 of the traveling wave type vibration film 14.
It becomes the maximum value at the position where there is a shift.

On the other hand, the amplitude of the traveling wave type vibrating membrane 14 due to the high frequency component 32 in the traveling wave frequency domain is large on the opening 20 side and becomes smaller as it moves toward the hole 19 side. Further, the amplitude of the traveling wave type vibration film 14 due to the high frequency component 32 becomes maximum at a position close to the opening 20.

As described above, the amplitude of the traveling wave type diaphragm 14 at the time t ₁ is represented by the envelope 33 of the combined amplitude of all the frequency components included in the traveling wave. Further, the amplitude of each position along the long axis direction of the traveling wave type vibration film 14 at the time t ₁ has a magnitude corresponding to the intensity of each frequency component of the traveling wave. Thereby, the n electrodes 25
Changes the respective stray capacitances C _float in accordance with the amplitude indicated by the envelope 33, and generates an electrostatic voltage V _C corresponding to the stray capacitance C _float . Therefore, as shown in FIG. 5C, n maximum value detection circuits 2
9 detects n maximum values e _i corresponding to the amplitudes of the positions of the respective frequency components of the envelope 33.

The composite value e _MAX of the n maximum values e _i detected by the maximum value detection unit 11 is i _MAX = (e ₁ , e ₂ , n), where i = 1, 2,... ... e _n-1 , e _n ). This is a so-called set in which the synthesized value e _MAX is arranged in the order of frequency using, as elements, intensity values of a plurality of frequency components each representing a plurality of frequency bands obtained by dividing the frequency band of the sound 2 into n pieces for each sampling period. Means a vector.

The maximum value detecting section 11 supplies the detected n maximum values e _i in parallel to the interface section 12 as an audio signal S3, and encodes the n maximum values e _i . Thus, the encoded synthesized value e _MAX is converted into vector quantized data for each sampling period that can be easily processed by the computer 6.

In the above configuration, when the user gives the voice 2, the traveling wave microphone 3 converts the voice 2 into a parallel voice signal S2 in the frequency domain in real time, and performs a sampling period based on the voice signal S2. Composite value e for each
Generate _MAX . Subsequently, the traveling wave microphone 3
N maximum values e _i of the composite value e _MAX for each sampling period
, And time-division multiplexes the encoded data to generate serial audio data S1. Computer 6
Decodes the synthesized value e _MAX for each sampling period based on the audio data S1, and immediately obtains the time-series synthesized value e _MAX .

Thus, the computer 6 performs the process of dividing the audio data in the time domain for each sampling period, and the operation of performing the fast Fourier transform by software to quantize the vector in the time domain audio data for each sampling period. It becomes unnecessary. This means that, of the three steps required to display the character of the meaning and the content of the voice 2 given by the user on the display screen 8 of the display unit 7, the continuously provided voice 2 is converted into the computer 6 itself. Means that the processing of the first stage for converting into the vector-quantized audio data which is a signal form that can be processed by the computer 6 can be eliminated from the computer 6 side.

As a result, the computer 6 immediately shifts to the second stage processing for quantifying the similarity between the audio data in the frequency domain and the reference pattern of the audio registered in the computer 6 based on, for example, the formant position and appearance frequency. be able to. In the third stage, the computer 6 selects a word corresponding to the voice data to be quantified based on the result of the quantification in the second stage, and displays the selected word on the display screen 8. Thus, the computer 6
Can concentrate on only the processing of the second and third stages, so that the time required for recognizing the voice 2 can be significantly reduced as compared with the related art.

According to the above configuration, the traveling wave generated in the major axis direction of the traveling wave type vibration film 14 is generated by the sound 2, and the frequency components existing differently from each other in the major axis direction of the traveling wave type vibration film 14 are generated. By separately detecting the amplitude of each maximum amplitude position by the plurality of electrodes 25, the sound 2 can be converted into a parallel sound signal S2 in the frequency domain at a higher speed with a simpler configuration than in the past. Can be converted.

The traveling wave type microphone 3 is n in real time.
By generating n maximum values e _i for each sampling period from the audio signal S2 generated in parallel for each sampling period and encoding them to generate vector quantized data, a vector for each sampling period that occupies most of the voice recognition processing is generated. The quantization process is completed by the traveling wave microphone 3. Thereby, the burden of the voice recognition processing on the computer 6 side can be reduced step by step. Therefore, the computer 6 can handle and recognize only the semantic portion of the speech 2 at a higher level.

In the above-described embodiment, the case where the traveling wave microphone 3 generates the serial audio data S1 based on the parallel audio signal S2 generated by converting the audio 2 in real time has been described. However, the present invention is not limited to this, and can be applied to a case where parallel audio data that can be easily handled by the computer 6 is generated based on the parallel audio signal S2. In this case, the same effect as described above can be obtained. Incidentally, the interface system is not particularly limited, but the communication speed needs to correspond to the resolution of the traveling wave microphone 3.
For this reason, a high-speed interface system such as ultra fast SCSI or serial SCSI is desirable.

In the above-described embodiment, the case where a DC voltage is applied to the plurality of electrodes 25 when detecting a change in the capacitance of the plurality of electrodes 25 has been described. However, the present invention is not limited to this. It is also applicable when applying a high frequency voltage. Thereby, DC balance can be improved.

Further, in the above-mentioned embodiment, the case where the traveling wave type vibration film 14 is formed of a metal thin film and is connected to the ground line when the potential is applied to the traveling wave type vibration film 14 has been described. The present invention is not limited to this. A vibrating film of any material is formed and a thin metal film is adhered on the vibrating film surface to be connected to a ground line, or the vibrating film is electrostatically charged such as a film for electret charging. The present invention can also be applied to a case where the member is formed of a member or the member is attached to the surface of the diaphragm.

Further, in the above-described embodiment, the case where the plurality of electrodes 25 are arranged only below the traveling wave type vibration film 14 when detecting the amplitude of the traveling wave type vibration film 14 has been described. The present invention is not limited to this, and can be applied to the case where the electrodes 25 are arranged on both surfaces of the traveling wave type vibration film 14 as shown in FIG. In this case, the detection sensitivity can be increased by bridging the electrode 26 with the capacitor 26 in the maximum value detection unit connected to the electrode 25.

For example, as shown in FIG. 7, the voltage at the two connection midpoints between the stray capacitance 34 between the electrode 25 and the traveling wave type vibrating film 14 and the capacitor 26 in the maximum value detection unit is converted into two high input impedance buffers. 35 and 36 respectively. The buffers 35 and 36 respectively amplify the voltage at the connection midpoint and supply the amplified voltage to the voltage amplifier 37 for differential amplification. The voltage amplifier 37 supplies an output via the capacitor 38 to the rectifier circuit with only the AC component.

In the rectifier circuit, the diodes 39 and 40 and the capacitors 41 and 42 are bridge-connected, and the voltage at the connection point between the diode 39 and the capacitor 41 is
The voltage at the midpoint of connection between the diode 40 and the capacitor 42 is supplied to the maximum value detection circuit 43. Maximum value detection circuit 43
Detects the maximum value of the output of the rectifier circuit for each sampling period, and supplies the detection result to the analog-to-digital conversion circuit 44. The analog-to-digital conversion circuit 44 supplies the digitized audio data to, for example, a micro computer.

Further, in the above-described embodiment, each position in the long axis direction of the traveling wave type vibration film 14 is determined on the basis of the capacitance variation of the plurality of electrodes 25 which are close to and opposed to the traveling wave type vibration film 14. We described the case of detecting the amplitude of
The present invention is not limited to this, and can be applied to a case where the amplitude at a predetermined position of the vibrating membrane is detected by an arbitrary method, for example, an optical phenomenon, an electromagnetic phenomenon, or a piezoelectric phenomenon.

Further, in the above-described embodiment, the case where the traveling wave type microphone 3 is arranged outside the computer 6 and the audio data S1 generated by the traveling wave type microphone 3 is supplied to the computer 6 via the interface cable 5 is described. As described above, the present invention is not limited to this, and can be applied to a case where the traveling wave microphone 3 is provided inside the computer. In this case, the audio-electrical-signal conversion unit 10 samples and encodes the electric audio signal generated in parallel in real time at predetermined time intervals, and generates and encodes the parallel audio data without passing through the computer interface means. It can be directly expanded to memory cells.

Further, in the above embodiment, the case where the traveling wave type vibrating film 14 is formed of a metal thin film having a total length 24 of about 35 [mm] has been described. A vibrating film may be formed in a small size on an arbitrary substrate, for example, a substrate on which an integrated circuit is formed, in an arbitrary processing step, for example, the same step as the integrated circuit, so that vibration due to sound waves in a frequency band is generated.

Further, in the above embodiment, the outer box 13
Although the case where the inside is filled with air has been described, the present invention is not limited to this, and can be applied to a case where any fluid other than air is filled.

Further, in the above-described embodiment, the case where the sound 2 in the human audible frequency band is converted into the parallel sound signal S2 has been described, but the present invention is not limited to this.
The present invention is also applicable to a case where a sound wave in an arbitrary frequency band provided with a fluid as a medium is converted into a parallel electric sound signal.

Further, in the above-described embodiment, the case where the traveling wave type vibrating film 14 is formed in an elongated rectangular shape has been described. However, the present invention is not limited to this, and the present invention is not limited to this. As long as the maximum amplitude positions of all the frequency components of the vibration in the film thickness direction generated by the sound wave are different from each other in the in-plane direction, the vibration film may have an arbitrary shape, for example, a spiral shape.

Further, in the above-described embodiment, the case where the plurality of electrodes 25 continuously discriminate the frequency components including the frequency components near the upper and lower ends of the frequency band corresponding to the traveling wave type vibration film 14 has been described. The present invention is not limited to this,
The present invention can also be applied to a case where the interval between a plurality of vibration detecting means is widened to detect a minimum frequency component in a specific frequency band. Further, the present invention can be applied to a case where the amplitude of a frequency component of, for example, only the middle sound range in the human audible frequency band is detected with a minimum resolution by several amplitude detecting means. This allows
At the time of human voice recognition processing, for example, a parallel electrical voice signal is generated only from four frequency bands, and quantized data is generated based on the four parallel electrical voice signals, thereby enabling computer voice recognition. The processing can be speeded up and the overall configuration can be simplified. By setting the minimum frequency band and the minimum resolution, the diaphragm can be shortened.

Further, in the above embodiment, the outer box 13
Has been described as a whole in the form of a wedge, but the present invention is not limited to this, and all frequencies of vibration in the film thickness direction generated in the vibration film by sound waves in an arbitrary frequency band using a fluid as a medium are described. As long as the maximum amplitude positions of the components are formed so as to be different from each other in the in-plane direction of the diaphragm, the container means for storing the diaphragm may have any shape, for example, a spiral shape or a cylindrical shape having openings at both ends. .

Further, in the above-described embodiment, the case where the present invention is applied to the voice recognition device 1 for recognizing the voice 2 by the computer 6 has been described, but the present invention is not limited to this.
The audio electric signal converter 10 generates control data corresponding to the meaning and content of the audio extracted based on the electric audio signal generated in parallel in real time, and gives this control data to any electronic device such as a type. Writers, robots, video cameras, television receivers, computers, telephones,
The present invention can also be applied to the case of controlling an automobile navigation device or the like, or the case of responding to voice information corresponding to the meaning and content of the extracted voice. Further, the present invention is also applicable to a case in which the position and appearance frequency of, for example, a formant of a voice generated by a speaker are measured based on an electric voice signal generated in parallel in real time by the voice electric signal converter 10 to identify the speaker. it can.

[0055]

As described above, according to the present invention, the vibration film is vibrated in the film thickness direction by the sound wave, and the amplitude of each of the maximum amplitude positions for each of the frequency components differently present in the plane of the vibration film is determined. A sound wave detection device and a sound wave processing device capable of converting a sound wave into a parallel electric signal in the frequency domain at a higher speed with a simpler configuration as compared with the related art by separately detecting by a plurality of amplitude detection means. realizable.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of a sound wave detecting device and a sound wave processing device according to the present invention.

FIG. 2 is a schematic diagram for explaining a sound electric signal conversion unit;

FIG. 3 is a block diagram showing a connection state of a plurality of electrodes.

FIG. 4 is a connection diagram illustrating a configuration of a maximum value detection unit.

FIG. 5 is a signal waveform diagram for explaining direct conversion from vibration of time-domain sound to electrical sound signal in a frequency domain.

FIG. 6 is a schematic configuration diagram illustrating a plurality of electrode arrangements according to another embodiment.

FIG. 7 is a connection diagram illustrating a maximum value detection circuit for detecting a vibration state of a vibration film by a plurality of electrodes according to another embodiment.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Voice recognition apparatus, 2 ... Voice, 3 ... Traveling wave type microphone, 4 ... Cable connector, 5 ... Interface cable, 6 ... Computer, 7 ... Display part, 8
...... Display screen, 9 ... Sound collection unit, 10 ... Audio electric signal conversion unit, 11 ... Maximum value detection unit, 12 ... Interface unit, 13 ... Outer box, 14 ... Travel wave type diaphragm , 15 ...
Electrode support plate, 16 ... Traveling wave type diaphragm support plate, 17 ...
... upper floor room, 18 ... lower floor room, 19 ... holes, 20, 21 ...
... opening, 22, 23 ... short side, 25 ... electrode, 26 ...
... A minute capacitor, 27... A constant voltage source, 28... An amplifier, 29, 43... A maximum value detection circuit, 30... A time-domain signal waveform, 31. …… Envelope, 34 …… Stray capacitance, 35, 36
… Buffer, 37… voltage amplifier, 38, 41, 41
... capacitors, 39, 40 ... diodes, 44 ...
Analog-to-digital conversion circuit.

Claims (19)

[Claims] 1. A vibrating membrane which is vibrated in a film thickness direction upon application of a sound wave using a fluid as a medium, and in which the maximum amplitude positions of all frequency components included in the sound wave are different from each other in a plane. It is arranged corresponding to each maximum amplitude position,
A plurality of amplitude detecting means for separately detecting respective amplitudes of the plurality of maximum amplitude positions. 2. The sound wave detecting device according to claim 1, wherein the vibration is generated by a traveling wave propagating in an in-plane direction of the vibrating film based on the sound wave. 3. The sound wave detecting device according to claim 2, wherein the vibrating film is formed to be elongated, and the traveling wave propagates in a longitudinal direction of the vibrating film. 4. The sound wave detecting device according to claim 1, wherein said vibrating film is made of a metal thin film. 5. The sound wave detecting device according to claim 1, wherein the vibrating film has a metal thin film adhered on a film surface made of an arbitrary material. 6. The sound wave detecting device according to claim 1, wherein the vibrating film has a member that is electrostatically charged. 7. The amplitude detecting means is an electrode arranged in proximity to at least one surface of the vibrating membrane, and outputs an electric potential according to a stray capacitance generated between the electrode and the maximum amplitude position. The sound wave detection device according to claim 1, wherein: 8. The sound wave detecting device according to claim 1, wherein the vibrating film vibrates at least in a human audible frequency band. 9. The sound wave detecting device according to claim 1, wherein at least the vibrating film is formed on a substrate. 10. An electric amplitude signal generating means for generating a plurality of electric amplitude signals corresponding to respective detection results of said plurality of amplitude detecting means.
The sound wave detection device according to claim 1. 11. The sound wave according to claim 10, wherein said electric amplitude signal generating means has a sampling circuit which samples said electric amplitude signal at a predetermined timing to generate a sampled amplitude signal. Detection device. 12. The sound wave detecting apparatus according to claim 11, wherein said sampled amplitude signal is a maximum value of said electric amplitude signal every predetermined period. 13. The sound wave detecting apparatus according to claim 11, further comprising an analog-to-digital converter for converting said sampled amplitude signal into digital data. 14. The sound wave detecting apparatus according to claim 11, further comprising an interface means for converting the digital data into data according to an external processing method and outputting the converted data. 15. A sound wave processing apparatus comprising: a sound wave electric signal converting means for converting a sound wave into an electric signal; and a signal processing means for processing an output signal of the sound wave electric signal converting means. The sound wave is given by using the fluid as a medium and vibrates in the film thickness direction, and the maximum amplitude positions of all the frequency components included in the sound wave are different from each other in the plane. Each is arranged correspondingly,
A sound processing apparatus comprising: a plurality of amplitude detecting means for separately detecting respective amplitudes of the plurality of maximum amplitude positions. 16. The sonicator according to claim 15, wherein the vibrating membrane vibrates in a human audible frequency band. 17. The sonication device according to claim 15, wherein at least the vibrating film is formed on a substrate. 18. The signal processing means extracts features of the sound based on a plurality of detection results generated in the same period by a plurality of amplitude detection means, and outputs the sound based on the extraction result. The sonicator according to claim 16, wherein the recognition is performed. 19. The signal processing means generates control data according to the meaning of the recognized voice, and supplies the control data to an arbitrary controlled object to control the controlled object. The sonicator according to claim 18, wherein