US2122191A

US2122191A - Sound translation apparatus

Info

Publication number: US2122191A
Application number: US54347A
Authority: US
Inventors: Ballantine Stuart
Original assignee: Individual
Current assignee: Individual
Priority date: 1935-12-13
Filing date: 1935-12-13
Publication date: 1938-06-28
Anticipated expiration: 1955-06-28

Description

June 28, 1938. s, BALLANTlNE 2,122,191

SOUND TRANSLATION APPARATUS FiledDec. 15, 1955 2 Sheets-Sheet 1 firom" Aw'a/ 20w Hague/icy Frequency June 28, 1938. s. BALLANTINE SOUND TRANSLATION APPARATUS Filed Dec. 13, 1935 2 Sheets-Sheet 2 Frequenc y Patented June 28, 1938 UNITED STATES PATENT OFFICE scum) 'razmsm'rron mans-rue stun Ballantine, Mountain 1s. 1. Application December 1:, ms, serial No. 54,341

' 19 Claims. (01. 119-1) Apparatus and methods for this general purpose have been described in my copending applixcations Ber. Nos.'6 ,245 and 6,246, filed Feb. 12, 10 1935, and Ber. No. 8,392, filedFeb. 26, 1935.

An object of thepresent invention is to provide methods of and apparatus for the further improvement of the quality of the reproduced sound in order that it may more closely resemble the 7 natural quality of the voice as perceived by aerial sound transmission.

A further object is to improve the articulation.

of speech sounds picked up by the throat micro- Dhone. In my copending application Ser. No. 6,245, I

have disclosed the'results' of my investigations which have established that the mechanicalenergy in the vibrations ofthe throat due to the voice, available for the actuation of a throat mi-' crophone, varies with frequency in a manner which differs from the variation with frequency of the aerial sounds. so that if these vibrations are reproduced by a system devoid of distortion '0 (frequency discrimination) the sounds do not resemble the natural sound of the voice. I have in that application shown how these discoveries may be utilized. to achieve natural reproduction and highly articulate speech with a'throat microphone. The general method comprises, in general, providing an overall transmission which rises as the frequency is increased. This com pensates for, the gross and more important differences in frequency distribution between the n throat .and voice-operated methods of sound pick-up. Application of these principles results in a vast improvement'over throat microphone systems ofthe prior art. v

Further experimental studies under various 5 conditions and with a variety of subjects have revealed a residual fuzziness" in the reproduction ofcertain sounds. This, I .find to be due to a concentration of mechanical energy at certain frequencies in the region. of the throat which is In in excess of theenergy present in the aerial sound, as revealed by the spectral analysis of oscillograms of the sounds as received through the air and through the throat microphone system. Such concentrations are probably caused by resonances inthe larynx, associated cavities and structures involved in the voice production. since the frequency response characteristic of the microphone itself with its associated amplifier is smooth. Wave analysis reveals the presence of several of these resonances. Although natu- 5 rally these vary in frequency and intensity from individual to individual, there is sufilcient clustering' to permit a definite identiilcationin terms of frequency. For example, one of the most prominent of these resonances lies in the neighborhood 10 of 900 cycles and can readily be demonstrated in the semi-vowel sound of "L" as in the word "hello", in the sound of "F'" as in "for" or in the sound of fM". 'Another concentration occurs in the region 2500-3500 cycles, and several of less importance between 1000.and 2000 cycles.

a Reference is made to the accompanying drawings forming a part of this specification, in which:

Fig. 1 is a graphic comparison of vibratory energy available at the throat and in sound waves in air; I Fig. 2 is a frequency-transmission curve for an amplifier or network which will com, ensate for an excessive energy inputat 00 cycles;

Fig. 8 is a circuit diagram of a translating system. including a network having a characteristic suchas shown in Fig. 2: V

Fig.4 is a fragmentary circuit diagram of a compensating amplifier system; f 80 Fig. 5 is a frequency-output curve illustrative of the operaticnof the Fig. 4 amplifier; and Fig. 6 is a fragmentary. circuit diagram of a sound translating system embcdyingthe invention, the microphone being shown incentral sec- The vertical lines of Fig. 1 show'the location and relative amplitude of the more important components in the sound L asperceived by means of the throat microphone system of my 40 copending application Ser. No. 6,245, and as perceived by means of a distortionless microphone actuated by the aerial sound waves. The 900- 'cycle component is seen to be in the ratio of about 5 to 1 in the case of the throat microphone, over that of the aerial microphone. 'I find that a readily recognizable improvement in the quality of the reproduced sound can be obtained by compensating'for these resonances in the frequency characteristic of the throat microphone system. This can beeffected either in the microphone or in the associated circuits. While it is desirable to compensate for all of the throat resonances, I find as a practical matter that attention can be profitably confined to a few of the more prominent ones, particularly to those centered at 900 and 3000 cycles. In the design of certain practical apparatus, I have confined the compensation to the 900-cycle group, employing a characteristic of the type shown in Fig. 2. This effects a considerable improvement, particularly in the case of reception by ordinary telephone receivers which themselves contribute an augmented response in this frequency region due to diaphragm resonance. It is to be understood, of course, that this new characteristic is to be superposed on the characteristic described in application Ser. No. 6,245. The latter can be regarded as a general trend compensation, and the present characteristic as a further refinement.

An embodiment of the invention as applied to e a carbon type of throat microphone is shown in Fig. 3. The apparatus to the left of the dotted line A is identical with the apparatus shown at the left of line 0-1), Fig. 2 of my application Ser.

No. 6,245, and includes a throat microphone M,

and a network I'of the constant resistance type which compensates for the general decrease in vibratory energy towards the higher speech frequencies and also for the decreased response of the carbon granule type of microphone athigher frequencies. The network II between the dotted lines A and B is designed to introduce a loss characteristic at the 900-cycle range similar to that shown in Fig. 2, and the network III between lines B and C provides a similar dip in the transmission at another larynx resonance frequency, say at 3000 cycles. The equalizers are shown for convenience as of the constant resistance type and may be so designed that all apparatus can be con nected on a constant image impedance basis, thereby permitting the substitution of the apparatus to the left of line C for an equivalent carbon microphone, designed for actuation in the usual manner by aerial sound waves, of the same resistance.

A second embodiment is shown in Fig. 4, in which the dip characteristic for the 900-cycle resonance is incorporated in the pre-amplifier. The triode amplifier A and following pentode A preferably have a. common cathode and are incorporated in a single glass envelope but for clearness of illustration, are shown as separate tubes. M designates the throat microphone, which may be of any convenient type. For the purposes of explanation, it is assumed that it is of the electromagnetic type and preferably of a construction to be described in a forthcoming application, having an output voltage which is proportional to. the mechanical actuating velocity over the audio range of frequencies. The output of microphone M is amplified by the triode section A which is coupled to the succeeding stage A by means of the resistance R2 shunted by the series resonant circuit L1, Ci, R1. This circuit is tuned to approximately 900 cycles and the triode stage thus has a transmission characteristic of the type shown in Fig. 2. Suitable values of the constants are Li=3 henries, Ci=.01 mfd., R1: 3000 ohms, Rz=100,000 ohms, R3 is a gain control of resistance of 500,000 ohms. The pentode section A is coupledto a succeeding tube T2 by means of a composite coupling comprising the resistance R4 and the resonant transformer L2L3. This transformer is resonant at the upper end of the frequency range, giving the desired rising frequency characteristic, as described in my copending application, and shown by the curve ABD in Fig. 5. The output circuit of T2 comprises the resistance R5 and transformer L4-L5. It is usually convenient to design the output circuit so that the impedance viewed from the output terminals is a pure resistance thus facilitating the connection of the equipment to a line or other electrical communication equipments. The function of the remaining apparatus in Fig. 4 will be readily understood by one skilled in the art. The. overall frequency characteristic of theequipment shown in Fig. 4 is illustra Fig. 5, curve ACD, which represents t" voltage to frequency relationship icrophone M is actuated at constant herange of frequencies.

ment shown in Fig. 6, the dip in the frequency characteristic is obtained mechanically in the microphone itself. Themicrophone shown is of an electromagnetic type to be more particularly described in a forthcoming application. Briefly, the diaphragm I, in the form'of a strip of electrical steel, is caused to vibrate by the button B in contact with the throat. The ends of the strip I are supported on permanent magnets 2 which are polarized as indicated. A back member 3 carries a pole piece 4 encircled by coil 5 that is connected acrossan amplifier 6. A small air gap of a few thousandths of an inch separates the .pole piece 4 and the diaphragm I. When the diaphragm is caused to vibrate an electrical voltage is generated in-the coil which is amplifiedby the amplifier 6 and actuates the telephone receivers I. The dip characteristic is obtained in this arrangement by means of the springs 8 which are attached to the diaphragm at one end, and which carry the weights 9 at their other ends. The dynamic impedance of such a combination to forces applied at the spring end tends to a very high value at the frequency of resonance between the compliance of the spring and the mass of the weight 9. This high impedance tends to suppress the motion of the diaphragm at this resonance frequency. The springs 8 and weights 9 are so chosen that resonance occurs at the frequency it is desired to attenuate. i

Two springs and masses are shown in Fig. 6, disposed on either side of the pole piece. Both may be tuned to the same frequency, or one may be tuned to one'frequency it is desired to suppress, and the second to another frequency.

It will be understood that the particular circuit arrangements herein described are illustrative of apparatus that may be employed to carry out the novel methods of this invention. Other circuit arrangements which may be used to obtain the desired frequency variation of transmission to compensate for larynx resonances will be apparent to those familiar with the design of trans mission systems.

I claim:

1. Sound translating apparatus comprising the combination of a mechano-electric transducer substantially non-responsive to sound waves in air and actuated by the vibrations of the body due to the voice, and mechanical means operative in the frequency range of from about 200 to 4000 cycles for selectively reducing the current output at a frequency of mechanical resonance of the body.

2. Sound translating apparatus comprising the combination of a mechano-electric transducer substantially non-responsive to sound waves in air and actuated by the vibrations of the larynx the vibrations of said transducer means by the due to the voice, an electrical network working out of said transducer, said transducer and network having an over-all rising frequency transmission characteristic in the range of from about 200 to ,4000 cycles to compensate for the general decrease in vibratory energy of the larynx towards the higher frequencies, and means for reducing the current output at a frequency of mechanical resonance of the larynx.

3. Sound translating apparatus of the type comprising a throat microphone, and a transmission network into which said microphone works, characterized by the fact that the overall frequency response curve of said microphone and network has a region of relatively low transmission efficiency within the frequency range of from about 200 to 4000 cycles corresponding to a mechanical resonance frequency of the larynx.

4. Sound translating apparatus comprising a throat microphone, and an electrical network for said microphone including a section having a rising efficiency with increasing frequency in the range of from about 200 to 4000 cycles and a second section having a marked decrease in transmission efficiency at a frequency of mechanical resonance of the larynx within the said frequency range.

5. Sound translating apparatus comprising a throat microphone responsive to mechanical vibrations of the larynx in the frequency range of from about 200 to 4000 cycles, and an amplifier working out of said microphone, said amplifier having a coupling circuit shunted by a path resonant at a frequency of mechanical resonance of the larynx.

6. Sound translating apparatus comprising a throat microphone responsive to mechanical vibrations of the larynx in the frequency range of from about 200 to 4000 cycles, and a. plurality of cascaded amplifiers working out of said microphone, one of said amplifiers'having a coupling circuit shunted by a path resonant at a frequency of mechanical resonance of the larynx, and one of said amplifiers having an output circuit resonant at a relatively high audio frequency.

'7. Sound translating apparatus comprising a throat microphone, including a vibratory member actuated by the mechanical vibration of the throat, and mechanical means for selectively damping said member at a frequency of mechanical resonance of the throat.

8. In the process of converting mechanical vibrations of the body due to voice into intelligible sounds, the steps which comprise converting the mechanical vibrations of the larynx in the range of from about 200 to 4000 cycles into electrical currents, and selectively reducing the currentoutput at a frequency of mechanical resonance of the larynx of the order of from 500 to 2000 cycles.

9. In the process of converting mechanical vibrations of the body due to voice into intelligible sounds, the step of converting mechanical vibra tions within the voice frequency range of from about 200 to 4000 cycles into electrical currents, and the step of transmitting the electrical currents, the efficiency of one of said steps varying with frequency and being less at about 900 cycles than at adjacent frequencies.

10. Sound translating apparatus comprising the combination with mechano-electrical transducer means substantially non-responsive to sound waves in air and actuated by the vibrations of the body due to voice within the frequency range of from about 200 to 4000 cycles, and means for transmitting electrical currents resulting from body, of means for reducing, the efficiency of at least one of said transducer and transmitting means at a preselected frequency of about.900 cycles... I r

11. Sound translating apparatus comprising a throat microphone, and an amplifier working out of said microphone, said amplifier having a coupling circuit shunted by a path resonant at a preselected frequency of about 900 cycles.

12. Sound translating apparatus comprising a mechano-electric transducer substantially nonresponsive to sound waves in air and actuated by vibrations of the larynx due to the voice, means for producing a relation between output and frequency which rises with frequency in the range of from about 200 to 4000 cycles when the transducer is vibrated at constant velocity, and means for selectively reducing the output at a frequency within the said range corresponding to a mechanical resonance point of the larynx, thereby to compensate for abnormally large vibrations of the body at such frequencies.

13. The invention as claimed in claim 12 wherein the said reducing means is operative in the frequency region of the order of 500 to 2000 cycles.

14. The invention as claimed in claim 12 wherein the said reducing means is operative at a frequency of approximately 900 cycles. a

15. Sound translating apparatus comprising a mechano-electric transducer substantially nonresponsive to sound waves in air and actuated by vibrations of the larynx due to the voice, means for producing a relation between output and frequency which rises with frequency in the range of 200 to about 4000 cycles when the transducer is vibrated at constant velocity, and mechanical means for selectively reducing the output at a frequency within the said range corresponding to a mechanical resonance point of the larynx, thereby to compensate for a normally large vibrations of the body at such frequencies.

16. In a sound translating system, the combi nation with a throat microphone of the type having an output characteristic which rises progressively with frequency from 200 to 4000 cycles when vibrated at constant velocity, of means for selectively reducing the output at a frequency corresponding to a mechanical resonance point of the larynx, thereby to compensate for abnormally large vibrations of the body at such frequencies.

17. In the process of converting mechanical vibrations of the larynx due to voice into intelligible sounds, the steps which comprise converting the mechanical vibrations of the larynx into electrical currents, transmitting the electrical currents with an efiiciency which increases with frequency in the range of from about 200 to 4000 cycles, and selectively reducing the current output at a frequency of mechanical resonance of the larynx in said frequency rangefthereby to compensate for abnormally large vibrations of the larynx at such resonance frequency.

18. The process of reproducing speech which comprises producing from mechanical vibrations of the body due to the voice electrical currents of the useful voice frequencies with an efficiency which rises with frequency in the range of about 200 to 4000 cycles, and selectively reducing the electrical current output at a frequency corresponding to a mechanical resonant point of the body, thereby to compensate for abnormally large vibrations of the body at such resonant frequency.

19. The process of reproducing speech which said range correspondinz'to a mechanical resocomprises producing from mechanical vibrations hate point of the body, thereby to compensate for or the body due to the voice electrical currents of abnormally large vibrstions oi the body at such the useful voice frequencies in the'mnge of about resonant frequency.

5 200 to 4000 cycles and selectively reducing the electrical current output at a frequency within STUART BALLAN'I'INE,