JPS58206240A - Analog signal processing device and using method thereof - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to the field of signal processing, and more particularly, the present invention relates to the field of signal processing, and more particularly, the present invention relates to the field of signal processing, and more particularly, to reduce subjective loss (λ More particularly, the present invention relates to the field of compressing data without compromising data.More specifically, the invention relates to the field of compressing the minimum and maximum values of a In other words, it relates to the field of processing signals in a manner that does not encode the reached value, but only the number of occurrences of the extreme value. The present invention provides a unique device for the human perceptual nervous system, the auditory nervous system. It can be applied to word processing devices, bandwidth compression and data compression methods, noise reduction systems and subjective bandwidth ν, and the calculation method.

Conventional signal 121 width technology uses a duplexer and 1. The device that detects Mugi I, Cod J No. can increase the level of strength of the signal obtained from the glaring signal source of Hiron (Ij Maru). By amplifying the solitary signal, the signal can be recorded, transmitted and reproduced. It can also be used for other devices at Suam Temple.
It is used in numerous devices such as audio equipment, video equipment, hearing aids and sensory simulation equipment. In the case of hearing aids and perceptual simulation devices, the human auditory nervous system has a dynamic range of 120 dB or more (it works for N, but the limit of electric city amplifiers is only about 100 dB, so human power is not required). A lack of dynamic range in a signal processing device has serious consequences, i.e. distortion and/or sensing conditions occur.The dynamic range of a signal processing device is generally affected by a on the other hand, and on the other hand, by the supply voltage.Therefore, essentially the dynamic range of xQ PM (
A device capable of handling 0 dB) must be able to process (i4) without being restricted by either of these two small lines.

Even if the device in which the signal is shaken into binary form, regardless of the amplitude value of the analog waveform, the above-mentioned criteria are met, the bandwidth required for signal reproduction in the frequency domain is It has proven difficult to reduce the dynamic range while simultaneously maintaining the dynamic range.

Known methods for reducing the bandwidth required for information transmission include, for example, the Galta method of reducing bandwidth. In this method, the difference between the amplitudes of successive Zanzals is encoded. Other methods include adaptive delta translation, adaptive transformation coding, voice compression, and linear predictive coding. Although these methods can substantially reduce the bandwidth, they are unreliable, and vocal tract modeling and linear predictive coding cannot transmit IM signals other than voice signals. .

A method is known that encodes only the characteristics related to the extreme values of an analog signal. In particular, conventional techniques have been devised to measure the number of occurrences of Maki-e. The number of occurrences of longing is usually output and encoded in a signal processing circuit. This No. 16 processing circuit differentiates the input signal and calculates the minimum value and maximum value.
l! and an infinite threshold for clipping and pinging the signal to detect zero crossings in response to a very short period of time with respect to the bandwidth of the input signal.
It has a

IJ2 Licklider Lo and I
&(Text [Effects of Differentiation, Integration and Infinite Peak Clipping on Blind Candy Intelligibility" by PollIIck), !1.-IL, 11; , 20.42~
51 pages) (1! + 47 years), the study of j (a certain a voice waveform) becomes seven, and through this research, we first differentiate the By performing peak clipping, it was found that it was possible to obtain an understandable smoothness of Vf'jM l, ju, and I.This technique may have solved the problem of human-powered diving ranges, In addition, if you do not understand the ILJ issue after pinging, the information will be lost during the pinging process. 1・-ance (TI+omaa) and snow ni ichirisu (S
Parks) [Study by tJ of Filtered/Clipped Speech Subjectively by Hearing Impaired Persons], Tsuyanaru Opusa Acoustical Nzanidei Op America, 49. pp. 1881-1887 (19
According to 1971), research has been conducted on whether hearing-impaired people can understand speech waveforms that have been filtered with a high-pass filter prior to clipping. Although it has been found that this method improves the RJNIf for some people, this method also has the same problems as the methods of Rittaraigu and Borak.

British Patent Nos. 1,501,874 and 843,607 describe a method for determining the number of occurrences of maximum and minimum values. However, in the devices of these two patents, in addition to measuring the number of occurrences of the signal's extreme values, the amplitude of the waveform at the maximum and minimum points is also sampled and processed.

The present invention is based on the discovery that the human sense organs, particularly the maximum and minimum amplitudes of the ears, are more sensitive to the frequency of these minute occurrences. .

Therefore, misfire #J provides a new model of the human perceptual nervous system, especially the @Kakushin frog system, and according to the present invention, the information between the extreme points contained in the signal is minutes, 1
Not required for 100 rounds transmission. This feature significantly simplifies various signal processing methods related to the human perceptual nervous system, and improves the subjective perception of the human perceptual nervous system. You will be able to increase or decrease the convergence of the imperial castle without any doubts, use the minimum amount of each piece of good news, and process the information that comes out.

Therefore, the object of the present invention is to use the characteristics of the human ear as described in 91' to calculate the frequency by using only the time interval between occurrences of extreme values of an analog waveform. Understandable stop 1
The purpose of the present invention is to provide a device for transmitting a single message.

The purpose of the present invention is to solve the ground noise problem that was occurring in Samurai and Brazil by the differentiation and clipping method of Kreider and Hook. Ritter Rider and Borak are 18+
8 differential and clipping, the focus of the research was on determining the amount of distortion that could be tolerated without impairing intelligibility for the human ear. In contrast, the present invention relies on differentiating and clipping the signal as a method of encoding the extreme values of the signal, based on the premise that maximum information can be conveyed without subjectively compromising quality. . Significantly, the present invention differs from Licklider and Bolak's proposal in that it emphasizes the response time of clicks and hits, a factor that Litterrider and Bolak did not consider.

Licklider and Bolak state that the intervals between spoken words are accompanied by very strong noise signals, and to remove such spurious signals, repeating (2oKHz)
) proposes to introduce ultrasound signals into the device. According to the invention, since we use natural random noise with a Gaussian distribution that accompanies the analog waveform, there is no need for iterative superposition. Natural analog waveforms include
Always the high frequency noise associated with the spectrum of the analog waveform is considered. Such paternization is usually generated as a residual component of the preceding waveform due to Fi random molecular noise, e or residual effects. Therefore, the noise contains information about the particular environment from which the signal was obtained.

Differentiate the jtfr-signal with 3L1 noise, then
When clipped and pinged infinitely at a sufficiently fast slew rate, the output of the Clino-e is a binary waveform with significantly more transitions than would be obtained with a signal without noise. For example, if we integrate the output of a clear ya, if the waveform strength is about zero, we will get a signal with a signal with a strength of about zero, and therefore it is possible to perform a quiet nyu in the transition state. noise between spoken words is minimized. Cry/? 11 has a bandwidth much greater than that of the analog drum, for example, in the case of the 'Ej' voice signal the noise is 50 KHz, preferably 10 (l KHz
If the clipper bandwidth is greater than , the background noise pointed out by Link-Raigou and Borac can be eliminated, while at the same time a commercial quality, undistorted output signal is obtained. Therefore, rather than removing noise from the analog signal (using filter circuits or local noise suppression techniques), it instead removes the noise itself! By using it, you can communicate with more precision using the least amount of proof data than a trap. Therefore, further details of the present invention/? ! The purpose of 1 is to provide a logic circuit that can serially transmit data at a much slower rate than conventional binary data transmission methods such as pulse coding and the method of the aforementioned British patent. For example,
In two British patents, information about the number of occurrences of extreme values and information about the amplitude of the extreme values are processed. Additionally, in the present invention, since only information regarding the number of occurrences of extreme values is processed, the data rate may be much lower. ) and Russ code modulation sample an analog waveform and convert the amplitude into several bits of binary information, but the device of the present invention can convert one binary level t to another binary level. Since we only use transitions to show the generation of value,
Data speed and speed can be significantly reduced.

A further main purpose of the present invention is to function as a hearing aid for auditory nerve disorders, based on the principle that the main value of the information contained in a speech waveform is not the amplitude of the waveform, but the number of occurrences of the middle and f. A circuit that can do fk! I'll give it to you.

Still another object of the present invention is to compress the bandwidth required for communicating with other conventional communication systems to provide intelligibility such as l=1. This compression of bandwidth reduces the analog waveform to a binary signal that only shows the occurrence of extreme values without encoding other information present in the analog waveform, resulting in more understandable and higher quality speech. The discovery that transmission is possible is made on the basis of a 1μ connection.

Furthermore, as a result of the effect of the tune, the present invention does not allow for subjective %f range rwA expansion. Although the bandwidth is reduced for information transmission, the encoded signal is subjectively heard by the listener as an unaltered signal, so the bandwidth is effectively expanded.
That means it's been a long time.

These and other objects of the invention are achieved by a circuit that includes means for converting an analog waveform into a differential difference in the presence of broadband noise. If the analog signal is previously processed to remove or band limit background noise, means are provided for reinserting the noise signal. After the signal is differentiated, it is sent to an infinite clip circuit to provide t1 binary information that transitions when an extremum occurs in the original signal superimposed with naturally occurring or reinserted broadband noise. Clip circuit asynchronous 2
The output power is then filtered, shaped and amplified or transformed and transmitted for communication. It is also possible to further process the output of the Chryso-I by suitable lock synchronization paths in order to perform communication transmissions synchronously.

Good digital fidelity can be obtained by using a sufficiently fast clock, ie, a clock with a repetition rate significantly faster than the maximum extrema generation rate.

In order to recover the original analog signal from the output of the chryso-mother circuit or synchronization circuit, acoustic or electronic integration is required. In some cases, for example in a hearing aid housing for the neurodeficient or in an iris organ device, other 19 Luz shaping circuits may be utilized.

Binary synchronization (g) from the II + 1 analog conversion circuit is further processed by the means V to an appropriate digital/analog (D/A) and the resulting analog signal can be transmitted. Because of the extra information between the extreme values, the bandwidth of the analog signal obtained at the output of the D/A converter is significantly narrower than the bandwidth of the original analog signal. , the objective of analog bandwidth compression is achieved, and the signal properly encoded by the device of the invention is provided by a device of significantly simplified structure compared to the prior art.
It may be transmitted or stored. For example, a binary output device such as a binary seismic device may be used in conjunction with the fA of the present invention. Furthermore, since the 91-meter summer switching device that generates the input signal is not required, the structure of the input variable benefit can be significantly simplified and improved.

Other objects and features of the invention will become apparent from the following description. Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings of Kayi.

The drawings will be explained. FIG. 1 is a block diagram of a basic extreme value encoding/decoding device r1 and method. The signal from the converter 8 is sent to the image dividing circuit 10. converter 8
A microphone or other suitable input signal source is required, and the differentiator circuit 10 is a resistor-capacitor path M4 or an active differential circuit. An approximation of the filter may be used. For best results, the filter should be 6 dB
/octave linear characteristics. °
7. Alternatively, the differential signal may be passed through numbering means or any other method known in the prior art as shown at 22 in FIG. It may also be generated by a converter designed for T'14 Bij. For example, an input signal can be acoustically differentiated using a technique using a suitable synthetic filter. Part (1) of FIG. 2A shows a typical analog signal f (t) at the input end of the differentiating circuit. This signal is superimposed with broadband noise, that is, noise having a bandwidth that is actually wider than the bandwidth of the analog signal. (2) in FIG. 2(A) shows (P4+fτt) appearing at the output terminal of the differentiating circuit. The high-frequency noise components are strengthened through differentiation processing, and all extreme values are shown in Figure 2 (
It is expressed as a zero crossing including a noise extreme value as shown in (2) of A). By superimposing noise on the input signal, the output signal from the differentiator circuit will have more zero crossings than if there was little or no noise. Without noise, it is not possible to recover an accurate and undistorted signal from the output of the device. If noise cannot be obtained from the input end of the transducer 8, for example, if the noise is suppressed by a preceding circuit such as a Dolby circuit, broadband noise having a Gaussian distribution is introduced from an appropriate Gaussian noise source 6. be able to.

The output of the differentiator circuit is sent to an infinite clock or 12 which has a much wider bandwidth than the input signal.For example, if a highly accurate output signal is desired, the
A clipper range exceeding z would be required. The output of the infinite chryso-arm is shown as m(t) in (3) of FIG. 2(A). 'i in infinite chryso/4, 2. J Answer Time: Extremely short labor <, I can sense that the thick microgelt has turned into a hair. Figure 2 (
As shown in (3) of A), its output is a series of binary nogles, and the /arm radius becomes i. The nogle width is determined by the number of times the extreme value of the noise generated by the input signal f (t > K it 'ti occurs.When the average value of the output of the differentiator circuit is larger than zero, It can be seen that the width of the forward-going /9 pulse is extremely narrow, and the width of the positive-going pulse is wide.The average f- of the differential signal is
2) x a y ma is indicated by 2: When it is a sea urchin, it goes positive/the width of the arm and goes negative! t”1 or 74? The width of the russ is equal. When the average value of the output of the differentiator circuit is smaller than zero, the slope towards the negative is wider. Thus, the occurrence of IrM erosion in the noisy original analog signal is converted into a series of binary signals. The noise signal plays an important role in generating this binary waveform and in properly recovering the waveform.
It is clear from the second prisoner CA). From Fig. 2A2 (A), it can be seen that the number of occurrences of the admiration is determined by the binary 1 in (3) of Fig. 2 (A) by the transition of the bow. analog signal f
The information regarding the amplitude of (t) is expressed by the number of occurrences of extreme values in the noise consumed by the fit f(t), ie, the number of disturbances.

Frequency information tt, L82 kl The repetition characteristic of the signal in (3) of (A), even if it is 0, point X in (2) of Fig. 2 (A)
, 7 and zK^·1 are included in the number of occurrences of zero crossings of the corresponding average E.

In order to recover the binary signal, the signal must be filtered with a suitable filter or other tJ? The signal is sent to the pulse shaper 14. /If it is an I-type circuit or an integrator, the signal o in (4) of M2 diagram (A)
(11)'f can be completely recovered by following the characteristics of random noise as shown by tl. The faster the response of the clipper 12, the better the accuracy σ of each output.

It is understood that the amplitude of the drum-shape detected is related to the characteristics of the random noise present in or intervening in the input signal 1 (1). For example, when the noise is strong, the amplitude is compressed to some extent; when the noise is weak, the amplitude is expanded. This particular effect is due to the nonlinear characteristics of the circuit that occur at high and low input levels.

A synchronization circuit 16, as shown in FIG. 1rc, may be utilized to provide a synchronous output. A D-type frizzo flop performs this synchronization function. FIG. 2(B) shows the signal m(t) applied to the input terminal of the synchronization circuit with the time axis greatly extended.

As can be seen from the figure, this signal is an asynchronous binary signal. The clock to which m (t) should be synchronized is denoted as c(t). Every time the clock pulse rises, the signal at the input end is sent to the output end with a slight delay. The signal line 200 synchronous binary output signal g (t) can then be sent to communication channels and other conventional digital devices. Here, as shown in FIG. 2(B), two types of errors may occur, namely, a synchronization error and an interpretation error. However, if the clock is repeated at a sufficient rate, synchronization will not unduly degrade the performance of the device. If the rate of blackout is slower than the maximum rate of occurrence of extreme values, as shown by the interpreter's error in FIG. 2(B), distortion will occur. However, depending on the application, some distortion may not be a problem.If the signal is highly intelligible with the Eboryo Amejoryoku device or hearing aids, the distortion can be tolerated.

The signal g (tltj) can be sent to a further pulse shaper which is routed to the stimulus recovery block 18 of FIG. It is a circuit. The output signal of signal #24 is then routed to a suitable amplifier and/or output converter, such as a loudspeaker or hearing aid output converter.

FIG. M3 (A) shows a circuit that performs the functions shown in FIGS. 1, 2 (A), and 2 (B). The input from converter 8 is sent to a differentiator circuit comprising capacitor 30 and resistor 31. Resistor 31tj: connected in parallel with the human power impedance after the capacitor 30, looking in the input direction of the circuit.

The human power voltage from converter 8 is typically 10 mV. The resistor 32 and the capacitor 33 form a high-pass filter for removing very high frequency components of noise present in the i!
It is coupled to the infinite clipper circuit 12 via a resistor 36 and a resistor 37. Diodes 34 and 35 protect against voltage peaks. The signal coupled through resistor 36 to the inverting terminal of operational amplifier 41 has a magnitude of approximately 1 μV at inverting input terminal 11C. The operational amplifier 41 and its associated circuitry 1 form a high-band preamplifier. Diode 37 in the feedback circuit
and 38 improve the frequency response of the preamplifier.

Preamplifiers exhibit a non-linear response to strong and weak signals. The preamplifier provides high gain when the signal is weak and low gain when the signal is strong to limit the signal coupled to comparator 50. Together with this, it works to support one clipping. The voltage level at the output terminal of operational amplifier 41 is typically ±1 when the input signal is weak.
Varies between mV and ±0.7V when the input signal is strong. If there is not enough random noise at the input of the differentiator circuit, noise can be introduced from the Gaussian noise source 6. The noise source is, for example, the coupling capacitor 55
The noise of the open collector-emitter base junction of the pie-peak transistor 7 coupled to the non-inverting input of the operational amplifier 41 via the Only the components are coupled to the page inputs of comparator 50. Comparator 50 is a high speed comparator with an input sensitivity of less than 1 mV. Resistor 44 and variable resistor 45 serve to vary the reference voltage against which the comparison is made. The output of comparator 50 is an asynchronous signal corresponding to m (t) in FIG. 2(A). This output is pulled up by resistor 51 to a maximum level of +5 volts. The signal may then be sent to a pulse shaping filter 14. This filter is a reduction filter or an integrator as shown. The integrator includes a resistor 56 and a capacitor 57 and substantially recovers the original input signal 1 (1) from the extremum encoded signal.The filter 14 converts the extremum encoded binary asynchronous signal below a certain frequency. It may also be a reduction filter that limits the band. It has been found that band-limiting an extremum-encoded signal in this manner provides a subjective f-bond that is superior to a similarly band-limited analog signal. In practice, the extremum-encoded signal may be band-limited to be substantially lower than the corresponding analog signal, again providing the same subjective sound quality as described above. Ru. However, in many applications there is no need to completely reproduce the original signal.

That is, in hearing aids used by people with hearing loss, intelligibility is lower than sound quality. Since the human perceptual nervous system is more sensitive to extreme values than to information contained between extreme values, it is not necessary to perform reproduction with 100/f-cent accuracy. Advantageously, the information content of the input signal is therefore effectively reduced without significantly altering the Vζ perceived sensation when the extremum encoded signal is applied directly to the human perceptual nervous system.

That is, even though the extremum encoded signal and the original signal appear to be very different based on time and frequency domain analysis, the encoded signal subjectively sounds similar to the original waveform.

The output of the CRE//PA circuit may be sent to a synchronization circuit comprising a D-type flip-flop 16 as previously described with respect to FIG. 2(B). Clock 9rus c (t
Each time l rises, the data at the input of the Fritznofrog is clocked to the output.

After the synchronization with clock gurus c (t) is completed,
The signal g (t) ft can be sent to a digital storage, transmission or reproduction device on signal line 20. The output of the D-type frizzo float 7'16 may then be sent to the stimulation recovery block 18 of FIG. In FIG. 3A, an active integrator comprising an operational amplifier 66 and associated circuitry is used to rotate the original waveform 1 (1) from the extremum encoded signal. Resistor 70 and capacitor 69 form the integrator feedback network. h'' variable resistance gain 65ii sharpens the magnitude of the signal appearing at the inverting input terminal of the arithmetic amplifier 66, and connects the resistor 68 and resistor 6 to ρ''.
7#'i A voltage dividing circuit is formed to apply an appropriate bias to the non-inverting input. The Q output signal of the flip-flop 64 is approximately O? Since the output voltage changes between 0 and 5 Ω, a voltage divider bias circuit is required. Or flip-flo,
The Q output of Pro 4 could also be capacitively coupled to operational amplifier 66 to block the DC component.

The output of the integrator appearing on signal line 24 is the original input signal 1
It is an analog signal corresponding to (1).

Dual monostable multivibrator 66 in Figure 3(B)
and 68 provide a synchronized output.

One multivibrator triggers on the positive end of the nostril, and the other multivibrator triggers on the negative end. This pair of monostable multivibrators is another form of stimulation recovery block 18 of FIG. Figure 3 (
C) shows a timing chart. Each monostable multivibrator may generate short pulses with pulse widths less than the clock period on both the positive-going and negative-going ends of the binary waveform. element,
It is determined by resistor 69 and capacitor 67.

As shown in FIG. 3A, the Q output of the positive end triggered monostable multi-zibrator 66 is coupled through a resistor 70 to the negative end triggered barrel stabilizer 80. The Q output of multivibrator 68 is coupled to signal S*SU via resistor 72. The outputs of the two base stable multipipe lakes are shown in FIG.
(t). Since the two resistors 70 and 72 form a partial voltage "a," the output waveform is approximately 2.5 mm as a single point, as shown by z (t) in Figure 3 (C). Therefore, the output appearing on signal gso is a series of 9 pulses corresponding to the number of occurrences of the extreme values in the original waveform. The circuitry forms a filter that ensures that only one valid extremum occurs during one period of the clock.

Two Haya-Anki Multi/(Il-Taro 6 and 68) and associated circuit elements and specific output blocks shown in Figure 3 (B) L1 are used for hearing aids for people with auditory nerve disorders. This is particularly advantageous when used in people with auditory nerve disorders, where K requires a very strong signal in order to be able to hear sounds. The ear is often damaged or painful. However, the sensory nerves in the inner ear are much better at sensing the extremes of a sound waveform than the amplitude of such strong signals. , a series of short pulses that temporally correspond to the extreme values of the input signal can be transmitted to the sensory nerves without causing permanent damage to the middle ear due to the limited energy of the short tJ? An asynchronous clip signal can also be sent to this output block, as shown by signal line 25 in FIG.

The device of the invention has a variety of uses. FIG. 4 is a block diagram of the synchronous communication device. The device is composed of a differentiator 101, a broadband critter 103, and a D-type flip-flop synchronization circuit 105.

It is particularly advantageous that for a given intelligibility, this device can utilize lower data rates than conventional digital communication systems. Although the data rate may be low, this is because information regarding the amplitude mK of the analog signal other than the number of occurrences of extreme values is removed. This communication method can be applied to both video and audio information. When applied to audio signals, the amount of data required to display audio or speech signals is reduced, so The recognition work is considerably simplified. Furthermore, in order to reduce the number of Fi dowels required in the field of video transmission and recording, the necessary equipment can be extremely easily reduced, allowing only the bandwidth that was previously required for audio signal processing to be reduced. You can use it to play back TV images.

FIG. 5(A) shows a block diagram of the hearing aid. Blocks 101 and 103#'i perform the same functions as described with respect to the other figures. Block 107Fi pulse shaping circuit, for example, tJ43 diagram (B)!
The low-pass filter, integrator, or dual stable multipigrator circuit described in connection with FIG. 3(C) may be used. For example, extreme value encoded signal #:tAM
It can be used as a language with radio sound quality, and can significantly improve intelligibility for hearing-impaired listeners.

FIG. 5(B) shows the reason why the device based on the VC of the present invention provides a signal that is subjectively considered to be more accurate than the conventional device whose band is limited to the same extent, and the actual subjective FIG. 3 is a waveform diagram for explaining the reason for providing a wide bandwidth gain. It has been found that the human perceptual nervous system responds to extreme values contained in analog signals, and that waveform shapes between extreme values are redundant. According to this discovery, (
A waveform like 1) is shown in Figure 5 (after differentiation and clipping).
It appears as a waveform like (2) in B). Low-pass filtering to band B may be performed by acoustic filtering within a filter circuit or the human ear itself.

After this filtering, a waveform similar to the waveform in (3) of FIG. 5(B) is obtained. As shown in (3) of FIG. 5(B), despite K having undergone the reduction filter processing,
The number of occurrences of extreme values has not changed much.

(4) in FIG. 5(B) shows a signal f(α) which is No. 16 t (t) but whose bandwidth is limited by a low-pass filter with a bandwidth B. The number of occurrences of extreme values is The waveform is slightly altered by the filter, so the waveform is perceived as more distorted than the extremum encoded signal.

Therefore, in reality, the extremum encoded signal is subjectively captured in the frequencies of bandwidth B that are lost in conventional filtering techniques.

FIG. 6 1 shows how the device according to the invention is applied to compress the bandwidth required for the transmission of analog signals. Differentiator 101, clipper 103 and D-type flip-flop Jt1. + conversion circuit 105 performs the function described above. The output of the D-type flip-frog is sent to an n-bit shift register 107. n counter 1 so as to generate one f pulse every n clock cycles.
Exclusion 31 due to 09! is driven by a clock. The output of the counter 109 sets the n-bit latch 111. The n inputs 1tiA of the n-bit latch are connected to the n output lines of the n-bit shift register 107. When the ie pulse appears at the output terminal of the counter 109,
The latch 111 transfers data from its n input lines 1 to its n output lines. The n-bit D/A converter 113 has two
Converts binary information to analog form. D/A converter 113
n input glands are connected to the output end of the n-bit latch 111. The output of the D/A converter is then sent to a low pass filter with band@B. Synchronization information is obtained by coupling the counter output to the signal shot 116 and providing synchronization information. Band-'pJh Byop
The output of a square wave oscillator 117 with a frequency a few hertz higher, for example 20 Hz, and the output of a signal shot 116 with a pulse width greater than one cycle of this oscillator are fed into a tiAND gate 119 to form a synchronization pulse. A highly modulated burst pulse is obtained. To ensure sufficient frequency separation, the modulated synchronization signal is then bandpass filtered by bandpass filter 121. The outputs of the low-pass filter 115 and the bandpass filter 121 are summed in an adder 123 to add a signal h(t'+) containing both analog data and equalization information. The output of the channel is multiplied by conventional techniques.
No. 1 (1). Appropriate low-pass filters, bandpass filters, and detection circuits are connected before the A/D converter 131 and the main divider 133. 0 (1 filter 125 is °)' program data is separated by 10if from the synchronization information. . bandpass filter 127
A synchronization signal is generated by a synchronization separation circuit comprising a detector 129 and a detector 129. signal g (ti, therefore signal h
Because the information content of (t) is much less than that of the analog signal f (t), the channel bandwidth required to achieve comparable intelligibility is much narrower than in conventional analog transmission systems.

Figure 7 shows a low-noise analog channel profile.
This is a second diagram. Traditionally, analog channels used for voice transmission are limited in the dynamic performance of the reliability allowed by the channel and transmitter. By encoding the extreme values, the original analog signal 1
(1) is represented by a series of binary levels, the transitions of these binary levels corresponding to extreme values. Since the dynamic range of the two communication signals is minimal, the dynamic range characteristics of the channel are not so important, and the channel configuration can be simplified. Furthermore, since binary signals are relatively immune to noise, noise can be reduced with respect to conventional analog systems. As shown in Figure 7, the basic system includes a differentiator 101 and a wideband click/103.

At the receiver end, the signal is recovered by a suitable signal shaper, such as a low-pass filter or an integrator.

Although the invention has been described in terms of specific embodiments in the foregoing description, it may be modified in various ways without departing from the broader scope and spirit of the invention as set forth in the appended claims.・It is obvious that it can be changed. Accordingly, the specification and drawings are to be regarded as illustrative rather than in a restrictive sense.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a basic device that performs signal processing by encoding extreme values; FIG. A (B) is a diagram showing the clock signal and other signals appearing at the input and output of the synchronization circuit of FIG. 1; FIG. 3 (A) is a diagram showing a circuit for processing the signals according to the invention; , FIG. 3(B) is a diagram showing another form of the post-stimulus block of FIG. 1, FIG. 3(C) is a diagram showing a timing chart of the waveform of FIG. 3(B), and FIG. , a diagram showing a method of synchronous communication at a low data rate, #g5 (A) is a block diagram showing a hearing aid for the hearing impaired in which the present invention is used, and Figure 5 (B) is a block diagram showing a hearing aid in the present invention. Figure 6 illustrates the subjective method of bandwidth expansion; Figure 6 illustrates the use of the system of the invention to perform analog bandwidth compression; Figure 7 illustrates the configuration of a low-noise analog channel. (Symbols in the figure) 6... Noise source, 8... Converter, 10... Differentiating circuit, 12... Broadband clipping circuit, 14... Filter, 16... Synchronization Circuit, 18... Stimulus recovery block, 20... Signal line, 22... Converter, 113
...D/A converter, 131...A/D converter, 13
3...Autumn divider, 1 (1)...Input analog signal,
m(t)...Extreme value encoded signal, o(t)...Random noise, g(t)...Synchronized binary output signal, ANLG
,0. ... Analog output, BNR, 0, ... Binary output, STML, O, ... Stimulus output, ASYNC, BN
R,...Asynchronous binary, CLK...Clock, 5YN
C, BNR, ... synchronous binary, 5YNC, ER, ...
Synchronization error, INTPR, ER, ... translation error, (CH
, MOD)...channel or modulator, EXT, LO
C.E.R. ...extreme position error, muscle IP, TRND, ... increase Il
! II instrument and transducer, 5YNC,S,・
...Synchronization signal, SER, O, ...Series output, CI...
·channel. Patent Applicant Arie Bina Patent Attorney Akira Aoki Patent Attorney Kazuyuki Nishidate Patent Attorney Matsushita Akira Yamaguchi Procedural Amendment (Method) January 2, 1980 B Commissioner of the Patent Office Kazuo Wakasugi 1. Indication of the case 1980 Patent Application No. 080334-2. Name of the invention Representation of analog signal processing and its method of use 3. Person making the amendment Relationship with the case Patent applicant 4, agent (3 others) ) 5. Date of amendment order Amended ill description 12) Drawings (3) Power of attorney (4) Certificate of nationality 7. Contents of amendment (11. Engraving of specification (no change in content) (2) Engraving of drawings (No change in content) +31 +41 As shown in attached sheet 8, list of attached documents (1) Engraved specification 1 copy (2) Engraved drawing 1 copy

Claims (1)

[Scope of Claims] 1. A device for processing an analog signal received as an input i+ with broadband noise superimposed thereon. means for detecting the number of generated mouth edges of (bl
encoders 18 and 12 which are coupled to the FJ detection means, have a bandwidth wider than the bandwidth of the analog signal, and contain practically sufficient information to reproduce the analog signal; An analog signal processing device that provides a means to reduce only the number of occurrences of minimum and maximum values. (b)
2. The apparatus according to claim 1, wherein the l #p right encoding means comprises clipping means for encoding zero crossings of the time axis into binary transitions. 3. The device according to claim 2, wherein the clipping means has a frequency of at least 50K) (z). 4. (a) Means for generating a clock pulse;
and (b) further comprising means coupled to the encoding means for inputting the encoded signal and synchronizing the encoded signal with the clock pulse to generate a synchronous binary signal.
An apparatus according to claim 1 or claim 2. 5. A device for processing an analog 4Th signal received as input with broadband noise superimposed thereon, the device processing the signal in the band of the signal without impairing the subjective perception by the human perceptual nervous system when reproducing the signal. It reduces the width,
The apparatus includes (8) means for detecting the number of occurrences of the minimum value and maximum value of the analog signal including noise, and (
b) 6 detecting means and having a bandwidth wider than the bandwidth of the analog signal and containing substantially sufficient information to reproduce the analog signal 'fc of the minimum value and An analog signal processing device comprising means for encoding only the number of occurrences of a maximum value. 6 in. The detection means detects the minimum and maximum values. 1.11 (b) The encoding means encodes the zero crossings t·lH,=sign @subtraction of the illuminated axis. Apparatus of claim 5, which includes clipping means for converting. 7.Rin ripping''4-&6 at least 50KHz
%#+Claim 6, 6L
No. 11 8 is connected to the encoding means to manually encode the im signal, and the human perceptual nervous system reproduces a signal that is felt exactly like the analog signal (J signal). i8. Apparatus according to claim 5 or 6 of claim 8, further comprising means.
Section 6e Ridge device. Maxo, (a) Clock! (bJ aA) is coupled to a coding means for receiving the coded signal as an incoming signal and synchronizing the coded signal with the code signal to generate a synchronous binary signal; %11; further comprising means for generating a signal;
Apparatus described in section. 11. The device according to claim 10, wherein the internuclear periodization means comprises a D-type frig-flop ffr. 12. further comprising means coupled to the internuclear periodization means for inputting the synchronous binary signal and reproducing a signal that is subjectively felt to be the same as the analog signal to the human sensory nervous system;
Apparatus according to claim MIO d. 13. The mosquito reproduction means consists of a pair of monostable multivibrators, the multivibrators generating a shorter output dulse of the synchronous binary signal, and one of the multivibrators generates a shorter output dulse of the synchronous binary signal. 13. The apparatus according to claim 12, characterized in that: - when the signal goes positive; - the signal is generated and the other is a synchronized binary; - when the signal goes negative; - 9 signals. 14. Broadband noise is heavy 8! An analog signal processing device used in a hearing aid for processing an audio input signal received as a sharp input, the device! '1 (a) means for detecting the minimum and maximum (1&) occurrences of said analog signal containing 1m noise, and (b) a bandwidth coupled to said detection means and wider than the bandwidth of said analog signal. A signal that has enough t# signal to reproduce a signal that can be understood by a hearing-impaired person as the analog signal is encoded as the minimum value and M), and only the number of occurrences of M) is encoded. An analog signal processing device comprising means for converting into 15 coupled to the encoding means for manually inputting the encoded signal;
Claim 14, further comprising means for reproducing No. 1d that can be recognized as the analog signal by a hearing-impaired person.
The analog signal processing device described in Section 1. 16. An analog signal processing device used in a synchronization device with a low data transfer rate that inputs an analog signal on which broadband noise is superimposed and outputs a binary synchronization signal, the device lt&, 1, (a). means for detecting the same number of occurrences of minimum values and maximum values of said analog signal containing noise;
b) as a transition in an asynchronous binary signal coupled to said detection means and having a bandwidth wider than the bandwidth of said analog signal and containing substantially sufficient information to reproduce said analog signal; E! encodes only the number of occurrences of the minimum and maximum values! - collection, and (cl) coupled to said encoding means for inputting a signal into said binary, and converting said two signals into a repetitive signal;
1. An analog signal processing device comprising means for generating a 1-bit binary signal with a period of 0. 17. An analog signal processing device used in a low-noise analog communication device inputting an analog signal with broadband noise multiplied by means for detecting the number of occurrences of the maximum value; and (b) coupled to the detecting means and having a bandwidth greater than the bandwidth of the analog signal and substantially sufficient to reproduce the analog signal. Analog signal processing device 18, comprising: means for generating only the number of occurrences of the minimum value and maximum value as an encoded signal containing information; a differentiator for exchanging the number of occurrences with zero crossings on a time axis; (b) the encoded signal is asynchronous binary signal-by-number; the analog signal processing device receives the asynchronous binary signal as an input; An analog signal processing apparatus according to Kawashiro et al., No. 17, characterized in that it includes a means for band-limiting an asynchronous binary signal, and further comprises means for band-limiting an asynchronous binary signal. 1' An analog signal processing device comprising: a transmitting device for compressing the bandwidth of a signal; and a device for receiving an analog signal; M
/Means for detecting the number of occurrences of the minimum value and maximum value of the first analog signal including noise; (e) means for converting the number of occurrences of the minimum and maximum values into an encoded signal containing substantially sufficient information to reproduce the first analog signal;
means coupled to the encoded signal for synchronizing the r+ encoded signal with the inverse signal to generate a synchronized binary 16 signal; means for converting into a second analog signal and transmitting it to a receiving device, the receiving device comprising: (a) means for converting the second analog signal into an r-notal signal; and (b) An analog signal processing device for reproducing the second analog signal from the r notal (g signal). 20 An analog signal processing device for processing the analog signal received as input with broadband noise superimposed The method includes the steps of: (a) detecting the number of occurrences of minimum and maximum values of the noisy analog signal; and (b) detecting a bandwidth wider than a bandwidth of the analog signal. 2. To reproduce the analog signal by encoding only the minimum value and the maximum value using the code encoding means, 2. to generate an encoded signal containing substantially sufficient information. 21. The method according to paragraph 20, further comprising the step of synchronizing the encoded code 46 with a repetitive signal to generate a synchronous binary signal. 22. A method using an analog signal processing device L''- used to present a 770G signal containing broadband noise to the human sensory nervous system, the method comprising: generating an encoded signal that includes information corresponding only to the number of occurrences of minimum and maximum values of the noise-containing analog signal, such that the encoded signal is subjectively 11 times substantially different from the analog signal; A method that uses an analog signal processing device to provide the human perceptual nervous system with information contained in an analog input signal superimposed with broadband noise. The method includes the steps of: (a) adding information about the amplitude of the analog signal to the encoded signal and actually determining the number of occurrences of the minimum value and the maximum value of the analog signal including the noise; and (b) generating information regarding the frequencies contained in the analog signal by the occurrence of a substantially repetitive feature of the encoded signal, whereby the encoded signal comprises A method that is designed to include sufficient brilliance to enable realistic reproduction of the analog signal. Claim II! i No. 23 Method of offering 6 pieces.