Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS3600524 A
Publication typeGrant
Publication dateAug 17, 1971
Filing dateJul 9, 1968
Priority dateDec 28, 1967
Also published asDE1762492A1, DE1762492B2, DE1762492C3
Publication numberUS 3600524 A, US 3600524A, US-A-3600524, US3600524 A, US3600524A
InventorsEmanuele Biondi, Leonardo Biondi
Original AssigneeEmanuele Biondi, Leonardo Biondi
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Hearing aid using multiple frequency translation
US 3600524 A
Images(2)
Previous page
Next page
Description  (OCR text may contain errors)

United States Patent [72] inventors Emanuele Biondl; 3,385,937 5/1968 Lafon 179/107 's; 333 Milan via Primary Examiner-Kathleen H. Claffiy [21] A I N 2:33: my Assistant ExaminerThomas L. Kundert PP O r [22] Filed J y E68 Attorney Edwin E. Greigg {45] Patented Aug. 17, 1971 [32] Priority Dec. 28, 1967 [331 "2:2 ABSTRACT: A method for making the human voice audible [31] r ti i t t u Se N and comprehensible to deep deaf persons according to the ag i 2 1 app w on multiple frequency translation technique or system, wherein the sound signals are converted into suitably amplified electri' cal signals, wherein a general electrical signal, containing the information which is desired to be comprehensible t0 deaf I 54] HEARNG A") USING MULTIPLE FREQUENCY persons, is filtered, sampled according a sampling frequency f TRANSLATION so that said electrical signal 15 picked off for durations of time 3 Claimsfl Drawing Figs At spaced at intervals from one another for a time T corresponding to the sampling period and such that each duration [52] U.S.Cl 179/107 R of time i sufficiently Sma hat the individual picked ff [51 1 Int. Cl Htl4r 27/02, portions f h i l are id ed to he of an approximately I H041 25/00 constant amplitude, but varying according to the values of said [50] Field of Search 179/ 107 Signal; the individual portions f the Sigmy thus Sampled are Cited then delivered to at least one holding circuit causing them to [56] Re erences continue for a period of time equal to T, and then sub- UNITED STATES PATENTS sequently filtered and converted to sound signals audible by 3,36! ,877 H1968 Kreer et al. l79/l deep deaf persons.

. MICRG SAMPLING HOLD PHONE AMPL. FILTER SWITCH CIRCUIT FILTER AMPL. TRANSDUCER e (l) e (H u (ll PATENTEU we] 7 I973 FIGI IIeII) SHEET 1 MICRO- SAM HOLD PHONE AMPL. FILTER SWI CIRCUIT FILTER AMPL. TRANSDUCER 2w3 4 s 'e 7 8 eIII eIII eIII uIII 4 STORING DEVICE PATENTEI] AUBI 7 19m SHEET 2 BF 2 HOLD CIRCUIT MICRO- PHONE FILTER AMPL. TRANSDUCER SAMPLING AMPL. FILTER SWITCH SWITC FIG 6 SWITCHING UNIT HEARING AII) USING MULTIPLE FREQUENCY TRANSLATION t This is a continuation-in-part of application Ser. No.

718,506, filed Apr. 3, 1968. t

This invention relates to a method for making the human voice audible and comprehensible to deep deaf persons.

It is known that deep deaf persons retain some sensitivity to lower frequencies (generally within the range of 200 and 500+l,000 c.p.s.) while exhibiting more serious lacks at higher frequencies.

Under these conditions, a mere amplification of a speech signal will not make the speech comprehensible since a significant portion of the information is within the frequency spectrum beyond the above values of 500+1000 c.p.s.

A known method of solving this problem is to effect a band translation by known modulation techniques. In accordance with the V-coder principle, apparatus have also been proposed for making several translations through separate narrow-band channels. In all of these apparatus the outputs of the several channels are added and sometimes the original signal i s also added at the output.

These systems we will refer to as multichannel systems since there are a plurality of separate channels connecting inputto output.

In contrastto these systems, systems having a single transmission path between input and output will be referred to as monochannel'systems." In a multichannel system, the output signal is substantially different from the input signal and the speech is altered so as to be difficult to understand and above all, difficult to memorize.

According to the present invention, the frequency translations are obtained by a simpler method than those mentioned above to provide output signals that are more readily comprehensible. Moreover, the method does not require the use of multichannel techniques.

More particularly, the method according to. the invention provides a multiple frequency translation by sampling of an electrical signal derived from a speech signal, filtering the sampled signal and reconverting it into a sound signal of improved comprehensibility.

In order that the invention may be more clearly understood, reference is made to the annexed drawings, which show also various structural forms of apparatus for carrying out the claimed method; further, the illustrations of such apparatus are given only as block diagrams, deeming that the individual components are known and may be readily carried out by those skilled in the art, whereby a detailed disclosure thereof has been considered as unnecessary.

FIG. I diagrammatically shows a general signal which is to be made perceptible to deaf persons.

FIG. 2 shows the signal of FIG. 1 after passing through the sampling device and the hold circuit.

FIGS. 3-6 show different embodiments of apparatus for carrying out the method according to the invention.

FIG. 7 is a circuit diagram of a storage element with its corresponding recording and read-off switches.

For a better understanding of the principles of the invention, the method will now be described with reference to the apparatus of FIG. 3.

The method is based on the sampling of input signals by means of an electronic switch or sampling device at a sampling frequency f In such a system, the input signal is sampled during sampling intervals of duration At occurring once in each time interval T=l/ fl. A! should be small enough for the input signal to be considered approximately constant during the interval.

Referring to FIGS. 13, the audio or speech signal which is to be rendered comprehensible to a deaf person is converted into an electrical signal e(t). The sampling device 4 converts the signal e(t) into a series of pulses e(t), the amplitude of which is approximately constant for the duration At of the pulse but varies from one pulse to another according to the values of the input signal.

The signals e'(t) are then passed to a. hold circuit 5 causing them to continue for the time interval T so as to produce at the output a step signal e"(t), as shown in FIG. 2. The step signal e"(t) is passed through a low-pass filter 6 which, as a firstdegree approximation, can be considered as ideal, having a cutoff or critical frequency j}, andwhich further modifies the signal e"(t) and provides anoutput signal u(t), which is amplified and reconverted into a sound signal which is more comprehensible to deaf persons.

The transformed speech produced according to the abovedescribed method has the following characteristics, as experimentally checked:

1. It is perfectly comprehensible to a normal ear, without giving the heater the sensation of any substantial voice'distortion.

2. It is audible to deaf persons and supplies in a comprehensible and storable form the infonnation contained in high frequency range speech without requiring excessive amplification.

3. The noise can be substantially reduced by means of suitable input filters 3, without invalidating the statements (I) and (2).

Particularly, as to the effect of the input filter, it has to be designed to provide a good compromise according to the following conditions: by increasing the filtering effecton frequencies higher than the sampling frequency, speech will be rendered more readily distinguishable to a normal ear (as readily deducible from the theory of the sampled signal systems) and hence probably even more comprehensible to deaf persons; however, if the increase is too large, the desired effect will be attenuated (and will finally disappear), that is, the effect of signal perceptibility by deaf persons.

The following hypothetical interpretation was given for the above phenomena.

Let it be assumed for simplicity of disclosure that the sampling device is of the ideal type and the hold circuit is comparable with an ideal low pass filter. Then, from the theory of sampled signal systems it is known that a periodical input signal e(t) having a period T,

when sampled, caused to pass through the hold circuit and a low-pass filter (having a cutoff or critical frequency )1, will provide an output signal 14(2):

wherein:

-a-different type from those which would be produced by ordinary modulation.

It is believed that the multiple translations as performed by sampling have provided the good experimental results hitherto obtained owing to one or more of the following reasons:

a. Being of the monochannel type, the output signal u( t) can be readily obtained from the input signal and it can be noted that the shape of said output signal u(t), as a time function, is not substantially different from the input signal e(t) as to its most significant aspects; thus, such an output signal will be readily comprehensible. Moreover, as a result of the aforedescribed frequency translation phenomena, said signal u(t) also has components of lower frequency than the corresponding components of natural speech and accordingly the signal can be more readily understood by deep deaf persons. On the other hand, in multichannel systems the output signal as a time function can be drastically distorted relative to the readily interpreted in terms of frequency (that is, assuming periodical signals) and when the filters are considered as ideal filters. However, what is probably of much greater significance is the signal shape as a function of time. The nature of the output of a multichannel system as a function of time is difficult to foresee, and it may be also very different from the input signal.

b. In a system based on modulation, frequencies which in the input signal are close but belong to different frequency bands may be separated in the output signal and be very part apart; this will tend to reduce comprehensibility. On the other hand, according to the invention there is less separation of frequencies. I

The apparatus of FIG. 3 comprises the following components in'tandem arrangement: microphone 1, input amplifier 2, input filter 3, sampling switch 4, hold circuit 5, output filter 6, output amplifier 7, and transducer 8 converting the electrical signals into sound signals.

' There will now be described some, modifications of the method and apparatus just shown. It has been experimentally I verified that, some advantage could be derived from the modified fonn shown in FIG. 4, in which the signal e(t) appearing at the output of sampling device 4 is delivered to a storing device designated 9. Y

The unit 9 cyclically stores the signal e'(t) at a sampling frequency f, on a set of storing elements, and picks it up from the storing elements cyclically at a read-off frequency f,. Frequencies f, andf are independent of each other, but such that the ratio thereof is kept constant.

FIG. 7 shows in more detail the electric circuit diagram of a storage element 37 with its recording and'read-oif switches 33 and 38, respectively. The element comprises a transistor T acting as an impedance adapter and affording power amplification of the signal applied to terminal 30. The .base of transistor T5 is connected to the common terminal 31 of resistors R1 and R2 forming a voltage divider.

The ends of the voltage divider are connected to ground 32 and to conductor 35, to which is also connected the collector of transistor T5. The emitter of transistor T5 is connected to the base of a transistor T6 of the recording switch 33. The emitter of transistor T6 is connected through conductor 34 to the emitter of a transistor T3 forming in conjunction with transistor T6 a switching current circuit. A resistor R3 is connected between the conductor 34 and ground 32. The collector of transistor T3 is connected to a power supply through resistor R4, conductor 35 and terminal 36. The values of 'resistors R3 and R4 are selected to determine the quiescent point of storage element 37. Through conductor 35 the collector of transistor T6 is connected to the base of a transistor T4 of the read-off switch 38. A capacitor C1, forming the storage element 37, is shunted between conductor 35' and ground 32. A diode D1, connected between conductor 35' and the collector of transistor T3, prevents the storage element from discharging when switch 33 is open. A diode D2, connected to the base of transistor T3 and the positive potential terminal 39 maintains the base voltage at a value not greater than 4.5 volts, and a resistor R5 supplies the base current of transistor T3 and biases diode D2.

Recording switch 33 is controlled by pulses 27 supplied to terminal 39. Transistor T4 of switch 38 has its collector directly connected to conductor 35, whereas the emitter is similarly, input signals are applied from terminal 43 to the other elements via conductors 44. l

Withregard to unit 9 of FIG. 4, the number N of memory or storage is low, such as 2 or 4, and sampling frequency fir can be both greater and lesser than the reading frequencyf Thus, in

' the apparatus of F 4 according to theiinvention, only slight changes in the output signal are desired. Experimentally, good results have been obtained both with f smaller than fs and with f, greater thanfir, depending on the subjects.

The apparatus according to FIG. 4 was provided with two storage elements, zero order hold circuits and means for varying the sampling frequency flas well as the reading frequency f,in the range of500 to 3000 Hz.

Excellent results were obtained with sampling and reading frequencies of 1500 to 3000 Hz, respectively, and vice versa.

A further modification of the invention will be described with reference to FIG. 5. 1

1 The sampled signals from sampling device 4 are switched by an electronic switch .10 into two separate channels, each of which comprises the components 5, 6, 7 and 8 of FIG. 3,and terminating at the right ear D and left ear S, respectively. The switching of the sampled signals can be provided by switch 10 both alternately and by groups.

Where the switching is by groups, the apparatus of FIG. 5 may be modified to include a switching unit 11 as shown in FIG. 6, similar to unit 9 of FIG. 4, except that in this case the number N of storage elements is high and f larger than f,. In this case, unit 11 performs also the function of switch 10 of FIG. 5 for the signal switching by groups to said two ears D and S. Practical tests gave good results with this apparatus. It is also possible tocombine the advantages ofthe apparatus shown in FIGS. 4 and 5 by applying signals fr'om'sarnpling device 4 to'unit 9 of FIG. 4 comprising a low number of storing elements (considering both the casef, is greater than f, and the case wherein f, is less than 1",) from unit 9,- alternately or by groups, to the two ears.

l. A method for making the human voice comprehensible to deep deaf persons according to the multiple frequency translation technique, wherein the sound signal is converted into an amplified electrical signal containing the information which is to be made comprehensible, is filtered and then sampled at a certain sampling frequency, the time interval between the picking off of the signal samples being sufficiently small so the information content of theoriginal speech signal will not be lost, and the time interval, during which the picking off of a signal sample is effected, beingsufficiently small so that the signal will remain approximately constant as to amplitude during the picking-off time, the signal sample being supplied to at least one hold circuit causing it to continue until a subsequent signal sample will enter the hold circuit, the frequency translated signal sample being subsequently filtered and reconverted into an output sound signal of improved comprehensibility for deaf persons.

2. Apparatus for making the human voice comprehensible to deep deaf persons according to the multiple frequency translation technique comprising: means for converting sound into an electrical signal, a first filtering and amplifying circuit, a sampling circuit converting the electrical signal into a series of pulses, the duration of which is considerably smaller than the time between the individual pulses, a hold circuit to translate the frequencies of the individual pulses into lower frequencies and to extend the pulses, so the output of the hold circuit will be substantially continuous, a second filtering am plifying circuit, and means for'applying the output of said second circuit to the ear of a deaf person.

3. Apparatus as claimed in claim 2,comprising a second hold circuit, a third filtering and amplifying circuit and a second means for applying the output of said third circuit to another ear of a deaf person.

EDWARD M.FLETCHER, JR.

(by/6) UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3, 600, 524 Dated August 17, 1971 lnventofls) Emanuele Biondi and Leonardo Biondi It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

Column 2, first formula, delete the last "n" Column 3, line 11, cancel "part" and insert far Column 4, lines 4, 8, 9 and 12, "fs" should be f Signed and sealed this 7th day of March 1972.

(SEAL) Attest:

ROBERT GOTTSCHALK Attesting Officer Commissioner of Patents

Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US3361877 *May 1, 1964Jan 2, 1968Bell Telephone Labor IncNarrow band speech transmission system
US3385937 *Jan 29, 1964May 28, 1968Centre Nat Rech ScientHearing aids
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US4289935 *Feb 27, 1980Sep 15, 1981Siemens AktiengesellschaftMethod for generating acoustical voice signals for persons extremely hard of hearing and a device for implementing this method
US4403118 *Mar 20, 1981Sep 6, 1983Siemens AktiengesellschaftMethod for generating acoustical speech signals which can be understood by persons extremely hard of hearing and a device for the implementation of said method
US4874535 *Apr 28, 1988Oct 17, 1989The Britisch Petroleum Company P.L.C.Scale inhibitors
US8340333 *Feb 29, 2008Dec 25, 2012Sonic Innovations, Inc.Hearing aid noise reduction method, system, and apparatus
US20090220114 *Feb 29, 2008Sep 3, 2009Sonic Innovations, Inc.Hearing aid noise reduction method, system, and apparatus
USRE31031 *Dec 11, 1979Sep 14, 1982 Implantable electronic hearing aid
Classifications
U.S. Classification381/316, 381/312
International ClassificationA61F11/04, H04R25/00
Cooperative ClassificationH04R25/353
European ClassificationH04R25/35B