US1890640A - Method of and apparatus for sound recording and reproduction - Google Patents

Description

N` DEISCH Dec. 13, 1932. METHOD 0F AND APPARATUS FoR soUND RECORDING AND REPRoDUcTIoR 2 sheets-sheet 1 l .In-

11. 2 I@ .F F ,HU AAH A U V x A Y V HU NW jhm K 2 AU HU AU A AU y U V a\ A A U" V m AV A .WJNH nl N06/ Delscj;

Dec. 13, `1932. N.4 DElsCH 1,890,640

METHOD OF AND APPARATUS FOR SOUND RECORDING AND REPRODUCTION Filed May 29. 1950 2 Sheets-Sheet 2 25 24 za zz z| im IlH Noel Descb FIG. 9

Patented Dec. l 13, 1932 UNITED STATES PATENT OFFICE NOEL DEISCH, OF WASHINGTON, DISTRICT OF COLUMBIA -Application ilf' May 29, 1930. Serial No. 457,167.

The present invention relates to the recording and reproduction of sound, and its general object, broadly stated, is to provide improvements in sound recording and reproducing apparatus which make possible agreat intensity ran ge of reproduction without distortion, and whichbesides ensure a reproduction of sounds of small amplitude in their true timbre and Without extraneous noise due to imperfec- 19 tions in the recording matrix.

More particularly the object of the invention is to provide means (a) for analyzing a train of sound waves incident on the receptor of a soundv recording apparatus into two components, one of these components being a quality factor representing primarily the structural characteristic of the soundand comprising its properties of pitch and timbre, but being substantially of constant intensity, and

the other component being an intensity factor representing the group amplitude of the wave-train at points situated continuously along its length, this intensity factor determining the amount by which the amplitude of the quality factor above referred to must be inreased or diminished at corresponding points to produce a wave-train similar to the original wave-train, means (b) for recording the two components as above described as two separate but associated parts of a single sound record, and means (c) for translating the twocomponent record into an accurate counterpart of the original sound Wave-train.

Referring to the drawings:

Fig. 1 is a graph 'showing a wave-train of sine form and of amplitude varying according to a sine law.

Fig. 2 illustrates the manner in which the wave-train shown in Fig. l may be analyzed 43 into a quality component consisting of a simple sine wave of constant amplitude, and

a volume component whose ordinates follow a sine law.

Fig. 3 is a graph of'a wave-train representa ing a complex sound, the amplitude or volume of which varies according to a complex laW. c

Fig. 4 shows the manner in which the Waver train shown in Ffg. 3 may be analyzed into a quality component consisting of a complex Wave of a given amplitude, and an intensity component whose ordinates follow a complex law.

Fig. 5 is a graph of a complex sound wave of low and constant intensity, as recorded 4by ordinary processes.

also the positions occupied by slits used dur- I ing the operations-of recording and reproduction.

Fig. 8 is a diagram showing an illustrative apparatus by which sound may be recorded according to the method disclosed in the present application.

Fig. 9 is a diagram showing an illustrative apparatus by which a sound record made according to the present invention may be used to reproduce a fac-simile of the original sound.

It is well known that the elements which give character and quality to sound, and which determine the more fugitive subtleties of musical harmonies and govern distinctness of enunciationin conversation especially as respects certaln sibilanis and other consonants, are the harmonics or upper partials superposed on the fundamental. These minor elements of the sound are usually of high pitch and of small amplitude, and they Will for this reason in general be more faithfully reproduced the larger the scale on which they are recorded; that is, there-1s some certain value of the recording scale Which represents the minimum requirement for fully satisfactory reproduction. lVith any of the methods commonly used for the recording of sound, the maximum wave amplitude that can be accommodated in the record is limited by the constants of the reproducing apparatus and more particularly by the character of the record itself. Thus, with those recording methods in which the amplitude of the Wave is represented by the extent of the departure from a certain mean axis, as is the case in the ordinary mechanicalphonograph and in that class of photographic sound recorders in which the sound Wave-train is recorded as a sinuous line, the limits of this maximum recordable amplitude is fixed by the maximum traverse of the recording stylus or pencil of light. In a photographically recording system based on a variation-of-densty graph, the Limit is set by the difference between the least density and the greatest density which the photographic recording medium is susceptible of producing. In a magnetically recording system the limits are iixed by the diiierence 1n value between zero saturation and maximum remenant saturation in the magnetized wire.

I f a recording apparatus of one of the usual types is arranged to record as a maximum a sound wave-train of a given amplitude or intensity, sounds of a smaller ampltude will be recorded on a correspondingp ly smaller scale; it is thus obvious that if the apparatus is adapted to record without distortion very loud sounds, sounds of a certain low intensity will be recorded on a very restricted scale indeed, and will tend to become lost in the imperfections of the recording material. In other words, the capacity for exact recording and reproduction of noes of low intensity with present apparatus varies inversely with the permitted intensity range, and consequently a gain in intensity range can be had only by a sacrifice in the perfection with which the smaller component waves are recorded. This relation places a fundamental limitation on the scope and faithfulness of the record and its reproduction as obtained by apparatus at present in use.

In the present invention the recorded amplitude of. sounds above a given intensity level is increased, and the recorded amplitude of sounds below this intensity level is diminished, all sounds of whatever intensit within a very extended range being recor ed substantially to the same scale, this scale being chosen as one which falls well within the recording capacity of the specific recording process employed. For example, in a variation-of-density photographic record, this scale will correspond approximately to the length of the straight-line portion of the sensitometric curve. The finer detail of waves of whatever amplitude will hence be stated with the greatest prominence which the record is susceptible of providing, regardless of the intensity level of the original sound, the smaller wavelets ofthe lowest notes being recorded in distinct modulation. This allows of a much more detailed statement being made in the sound record of the minor details of the wave than is had when the wave-train as a whole is drawn to its truel scale, i. e., with differences of amplitude reptains the sound wave-train.

resented by corresponding linear diiierences of scale.

rllhe principle underlying the means by which this character of record is produced will now be explained. .A

Considering briefly some characteristics of sound wave-trains, it is known that such a train,'corresponding to a pure tone of constant pitch and volume, may be represented by a sound graph of constant amplitude and of uniform distance from crest-to-crest, as shown in Fig. 2 at l. A train corresponding to a pure tone of constant pitch but of variable amplitude 'or intensity may have a shape such as represented in Fig. l, in which the distance from crest-to-crest d, which measures the pitch, remains constant, whereas the perpendicular distance 71. between the summits and bases of successive half-waves, which measures the amplitudes, varies.

The wave-train represented in Fig. l may be readily analyzed into two dimensions or components, the one, whichmay be arbitrarily taken of any given scale, representing the shape of the wave, this component being shown isolated at 1 in Fig. 2, and the second representing the amplitude, or distance from the axis of abscissac tothe extremity of the loop (h, Fig. l) shown isolated at 2 in Fig. 2. These two dimensions or components will hereafter be referred to, respectively, as the quality component and the volume component of. the wave-train.

With an impure tone of variable pitch, and having a number of overtones superposed on it, such as represent the greater number of sounds to be recorded in commercial work, the parent wave may be of very complex form and may vary in value in an irregular manner, as illustrated in Fig. 3. Nevertheless this wave-train may still be analyzed into two components, one incorporating the complexities of form of the wave, and the other representing the maximum prevailing amplitude or group intensity, that is, the amplitude of the envelope 8, Fig. l, which con- Thus, the complex wave-train shown in Fig. 3 may as shown in Fig. 4 be analyzed into its quality component 4 and its volumeA component 5. The curves in Figs. '3 and 4 are of course in-I tended to be illustrative rather than strictly analytic.

Comparing now a sound record which represents a literal transcript of the original sound vibrations, and a record in which this transcript has been analyzed into two components, as above detailed,-it is apparent that whereas in the former case, as was shown above, the elements of intensity range and of sound-quality bear an inverse ratio to cach other and are hence in a sense antagonistic, in the present case these components are independent of each other. For this reason any desirable intensty range may be adopted for system and of the recording medium, with ,the result that the finer components of the.

sound Wave may all be given adequate prominence.Y f

The advantage of this system as respects low intensity notes is plainly shown in Figs.

and 6, where 'a not-e of very low and of constant intensity, shown in Fig. 5, and which K by ordinary processes Would be recorded on such a small scale as to be confounded with the imperfections of the record, With the method here described is recorded on the scale shown at 6 inFig. 6, the intensity factor being recorded as at7 in Fig. 6.

vrllhe method above described may be applied in various systems of sound recording in which the sound is recorded as a trace of various character and in various media, but

its use is here illustrated in connection with the photographic recording process, a sound record made according to this method heilig shown at 17 in Fig. 7, where the record of the quality component is shown as a strip of variable density 16, and the record of the volume component is shown as a strip of variable density on the developed photographic ilm 14, here represented as a sound motion picture ilm of standard dimensions.

Referring to Fig. 8, which illustrates one type of recording apparatus adapted to produce a composite record of the type shown at 17 in Fig. 7, and considering iirst the two recording systems proper, indicated generally at/ 20 and 30, which two recording systems are used respectively to record the volume component and the quality component of the sound wave-train, it will be noted that the system 20 comprises a light source 21,

Aa color filter 22 which isolates a narrow spec- Y tral` regionof light emitted by ,the source 21,

' a collimatmg lens 23, a polarizing nicol'24,

an electrooptic cell indicatedgenerally at 25, a compensator such as a quarter-wave plate 28, an analyzing nicol 29, and a slit 11. The principal sections of the nicols 24 and v 29 are at right angles to each other and make an angle of 45 With the principal plane of the electro-optie cell 25. The slit 11 is arranged to isolate a `narrow portion 15 of the (Width of ,the film strip 14, as shown' more plainly in Fig. 7.

The recording system 30 consists of an arrangement of parts similar to that of the system 20, the slit 13 in this case isolating a second narrow portion 16 of the Width of the film strip 14, (Fig. 7 The ilm 14 is arranged to move at a uniform rate. behind the slits 11 and 13, being passed fromthe reel 40 to the reel 41 by means'not shown.

When the plates 26 of the electro-optic cell 25 arel subjected to a dilerence of potential, the dielectric 27 therein (which may consist of nitrobenzol) becomes doubly refraeting,

causing rotation of the plane of polarization of the plane polarized light entering the cell from the nicol 24, this phenomenon being Well known'as the Kerr effect. Any change of potential of these plates causes a corresponding change of transmission to the slit 11 and to the photosensitive film 14; thus a transcript of theJ state of charge of the elec-v trodes 426 of the cell 25 is recorded 'on the film 14.

Normally the plates 26 of the c'ell 25 are held at a certain difference of potential by the battery 77, Whih acts through the resistances 76 and 78. A corresponding retardation is hence produced ,in the light traversing the cell 25. For the purposes of the `evident that bysuit-ably adjusting the compensator 28, the light entering the analyzer 29 can be given any desired amount of re-l tardation at any given potential active on the plates 26. The compensator 28 is adjusted so that at'the normal biasing potential'- the retardation of the light which enters the analyzer 29 is such that light is fully transmitted to the slit 11 and to the sensitive lm 14.

In the same Way the plates 36 of the cell 35 are held at a certain normal diderence of potential by the biasing battery 98, the sensitivity of' the cell to diierences of potential superposed on this biasing potential being increased by the biasing potential. rI 'he compensator 38 is adjustedso that at this normal biasing potential the retardation of light entering the analyzer 39 is such thatl light is transmitted corresponding to a median val vebetvveen full transmission and complete extinction. The above conditions of transmission as respects 'the analyzers 29 and 3 9, Yrespective'ly, hold when the apparatus is not actively recording a sound Wave-train, that is, When'it is recording silence.

voir' the sound wave-train; the' numeral 57 designates a sound receptor such as va microphone, which translates the sound waves incident thereon into their energy analogue in the transformer are transmitted tothe filament 71 and the grid 72 of the thermionic tube 70, which last is of the rectifyinrr gas filled) type. This tube, when no soun is incident on the microphone 57, normally passes a space current of some certain value. This current is supplied bythe battery 77 through the resistances 76 and 78. The potential of the plates 26 of the electro-optic cell 25, as before explained, are thus normally held at some certain value. But when a sound becomes incident on the microphone 57 the variational potential active between the filament and the grid of the tube 70 causes a rectified current of greater value than the normal space current to fiow in the plate circuit of the tube 70. The difference of potential of the plates 26 of the electro-optic cell 25 is thereby reduced, and, since the compensator 28 has been set for maximum transmission with normal flow of current in the plate circuit of the tube 70, this increase ofy current,

with corresponding decrease of electrode potential of the electro-optic cell, results in a smaller quantity of light fiux being transmitted to the slit 11 and the photosensitive lilm 14. The louder the sound, the smaller is the amount lof light transmittedY to the film 5 hence the strip or recorded area 15, Fig. 7, will after development show variations of density corresponding inversely with the loudness of the sound, and will constitute a record of the volume intensity, or amplitude component of the sound wave-train. A choke 74 is placed in the plate circuit of the tube 70 to prevent the alternating component of the plate current of this tube from appreciably affecting the electro-optic cell 25. y

The, alterations .of potential set up in the secondary of the transformer 53 are transmitted to the filament 81 and the grid 82 of the thermionc tube 80, which is also of the rectifying type. A choke 84 is placed in the plate circuit of this tube to prevent the alternating component ofthe plate current from appreciably affecting the transformer 58.

This tube, when no sound is incident on the microphone 57, normally passes a current'of some certain value, this current being maintained by the battery 77 and passingthrough the intermediate co'il 60 ofthe adjustable permeability transformer 58, magnetically saturating the core 62 thereof to a certain value. The windings of this intermediate coil `are'so disposed, as may be seen in the figure,

that there is no inductive action between it and the other coils 59 and 61 of the transformer 58. When a sound becomes incident on the microphone 57 a greater current fiows in the plate circuit of the thermionic tube 80,l

and the saturation of the core 62 of the transformer 58 is increased. The louder the sound, they higher the value to which the saturation o f the core62 is raised. Q' f n Itis well known that the permeability of iron varies as an inverse function of its satura tion.` Hence the permeability of the core 62 of the transformer 58 will be the less the greater the magnetizing force induced in it y the current flowing in the intermediate coil 60, and for this reason the alternating potential differences corresponding to the sound vibrations induced in the secondary 61 of the transformer 58 will, for a givenvalue of current in the primary coil 59, be the smaller, the greater the current flowing in the coil-60, that is, the louder the sound incident on the microphone 57.

The variations of potential induced in the secondary 61 of the transformer 58 are brought to act on the filament 91 and the grid 92 of the thermionic tube 90, producing corresponding alterations in the plate current of the latter. This variational plate current is sustained by the battery 94, and flows through the primary 97 of the transformer 95, inducing corresponding potential variations in the secondary 96 thereof` which act on thev plates 36 of the electro-optic cell 35 to produce fluctuations of transmission of light to the slit 13 and the photosensitive film L4. These fluctuations represent an analogue of the sound incident on the microphone 57, and their record, (16, Fig. 7) on the film 14 represents a graphic transcript of the sound wave-train incident' on the microphone 57 as regards quality.

From the discussion in the foregoing two paragraphs it will be plain that as the intensity of the sounds incident on the microphone 57 increases, simultaneously the sensitivity orvoltage amplifying factor' of the transformer 58 is depressed by the increased cur- -rent that flows in its intermediate coil 60,

the effect of which is to decrease the effectiveness of the increased variational current flowing in the primary coil 59 to induce varia'- tions in the secondary 61. By a suitable choice of the constants of the circuits these effects can be made to substantially balance each other over a very considerable range, and the quality record produced by the optical registering system 30 be made to be always of substantially constant amplitude.

It may be stated here 'parenthetically that, for the reason that, for a condition of silence, `the volume record corresponds to maximum density in the developed photographic deposit, and since for this same condition the quality record corresponds to a constant median intensity, the ground noise of the silent record orof arecord containing sounds'of low intensity'will be very 'considerably reduced as compared with a photographic record made by one of the usual processes. For it is well known" that the grai'niness of photographic deposits is more marked in the regions of lower density. yMoreover, these high normal densities in a record made according except that the slit 12 in this case isolates the to the present method tend to minimize the action of scratches and other faults.

An apparatus by which a record made according to the above described process may 5 be retransformed into a facsimile ot the original sound is shown -in Fig. 9. In this apparatus, two photoelectric detecting and amplifying systems, indicated generally at 115 and 116, controlled respectively by the volume component 15 and the quality component 16 of the composite record 17, Fig. 7, reacton each other in a manner presently to be described to produce sound of a quality and volume corresponding with that 'originally incident on the microphone 57 of the y receiving system shown in Fig. 8.

The detecting and amplifying systern'115, which system as above stated is controlled by the volume component of .the record, comprises a source 100, a slit 11', which isolates the volume component trace 15 of the composite sound record 17, Fig. 7, and a photoelectric cell 101', which last allows current furnished by the battery 104 to pass in amount depending on the instantaneous value of the light flux that is transmitted by the volume component of the composite sound record. This photoelectric current is amplified in a resistance-repeating amplifying system of conventional design, which comprises the thermionic tubes 102 and 103 and their attached circuits. electric current -passed by the cell 101 varies the electric charge of the grid 105 of the four-elementamplifying tube 102, the plate current of which -in turn causes corresponding potential variations of greater amplitude to act on the grid 106 of the tube 103,`thus changing the resistance of the latter tube and governing the passage ofcurrent in its plate circuit. Minimum transmission of light through the component 15 of the composite record 17 of the photoelectric cell 101, which condition corresponds to a record of silence, puts the tube 103 in the condition to pass the least current and conversely. i The second detecting and amplifying sys- 'tem 116, which is controlled by the-quality component of thev sound record, consists of an arrangement similar to that just/described,

quality component 16 of the sound record 17, Fig. 7. The density variations in the component 16 of the photographic sound record 17 are thus converted into corresponding variations in the resistance of the tube 113, the groupamplitude ot which variations is substantially constant.

The plate circuits of the two tubes 103 and -113 are in series, as shown, plate current for the' two tubesbeing` supplied by the battery 120. The current in the circuit comprising `these two tbes will hence be governed by the combined resistances of the two tubes .103 and 113. The resistance of the tubev 113 1s Specifically, the photo-` governedby the quality component of the record, whereas the resistance of the tube v 'the eii'ect of which is to reconstitute the sound in respect both to quality and volume.

The resultant variational current passes through the primary 122 of the transformer 121, in the secondary 123 of which it induces changes of potential which are transmitted to the filament 126 andthe grid 125 of the thermionic tube 124. Corresponding variations in the plate current sustained by the battery 127 of this tube thus take place, this plate 'current operating a sound-producing device such as a loud speaker, indicated at 128.

`It Will have been noted that the trace 15 is shown as displaced along ,the time axis with respect to the trace 16 ofthe composite sound record 17', Fig. 7, this being made necessary by the arrangement of the optical apparatus shown in Fig. ,Si and Fig. 9. By optical means not shown in the present application, however, these traces can be madeto be quite parallel, and the time displacement here shown is hence merely illustrative. Whatever the linear vdisplacement of the two components of the record may be, it may be advisable to displace slightly the slit used n recording the volume component of the sound with respect to the slit used in reproducing this volume component, to compensate the lag of the volume component on the quality component caused by the inertia of the recording and reproducing apparatus. That is, the

)pick up of the volume component may be given a certain time lead over'the pick up of the quality component. The positions of the slits used in recording and in reproducing, and illustrating this lead,lare shown at 13 and at 12, respectively, in Fig. 7.

While I have described my invention in detail with respect to a preferred form thereof, I do not desire to be limited to the form described, since many changes or modifications may be made in this form without departing from the -sprit and scopo of m invention, and I desire to cover all modlfications and forms coming within the scope of the appended claims.

I claim: y

1. .A sound recording apparatus -:tor reproducing a sound record including a quality component and an intensity component, comprising detecting means to convert said qual'- ity component into its electric current analogue and 'amplifying means to amphfy said electric current analogue, detecting means to convert said intensity component into its electric current analogue and amplifying meas to amplify said electric current analogue, each of said amplifying means including circuits having input land output stages,

`prising detecting means to convert said quality component into its electric current analogue, amplifying means including a circuit having input and output stages to amplify said electric current analogue, said amplifying means including in its output stage a thermionic vacuum tube, detecting means to convert said intensity component into its electric current analogue and amplifying means including a circuit having input and output stages to amplify saidelectric current analogue, said last mentioned amplifying means including in its output stage a' thermionic vacuum tube, sald output stage" amplifyin tubes being placed in series relation Where y the current iowing in said amplifying tubes is the resultant of the resistances of said tubes.

3. In sound recording apparatus comprising an electro-optic cell the electrodes of which are normally held under a substantially constant electrical bias, means for producing polarized light and for passing said polarized light through said cell, means vfor detecting changes of polarization sustained by said light .in traversing said cell, and an optical compensator interposed between said electro-optic cell and said detecting means for changing the retardation in said transmitted polarized light tent least partially compensate a change of retardation caused by said bias of said cell.

4. A sound recording apparatus comprising an electro-optic cell, means for impressing an electrical potential between the electrodes of said cell, and means responsive to the prevailing group amplitude or envelope 'of the sound wave-train being recorded, for changing the value of -said electrical potential to produce a sound record representing the volume component of the recorded sound.

5. A sound recording apparatus for producing a sound record representing the se arated volume component of the sound to e recorded, comprising an electro-optic cell the electrodes of which are charged y a source of electromotive force acting through a resistance, and means for draining away the charge on said electrodes to vary the electrical force active in said cell as the intensity of the sound to be recorded varies.

6. A sound recording apparatus for producing a sound record representing the separated amplitude component of the sound tol be recorded, comprising an electro-optic cell the electrodes of which are charged by a source of electromotive force acting through a resistance, a thermionic tube shunted across the terminals of said cell, and mean's for varying the potential between the filament and the grid of said thermionic tube to vary its conductivity as the intensity of the incident sound varies.

7. A sound recording apparatus for producing a sound record representing the separated frequency component of said sound, comprising recording means the input of `which yis electrical energy, a transformer having a ferromagnetic core, a primary, a secondary, and a supplementary winding to vary the magnetic saturation of said core independently of said primary, said primary being connected to a circuit including a source of electro'motive force and a microphone, said secondary being connected to said recording means, means to supply current to said sulpplementary winding, and means controlled y the intensity of the incident sound to vary the current supplied to said supplementary winding of said transformer as the intensity of the recorded sound varies. v l

' 8. An apparatus for recording sound, comprising means actuable by a varying electrical current for recording the frequency component of said sound, a microphone for converting sound waves into electrlc current variations, means for isolating the intensity component of said electric current variations, a transformer having a ferromagnetic core, a priirlary, a secondary, and a supplementary winding to vary the magnetic saturation of said coreindependently of said primary, the

primary of "said transformer receiving electrical energy representing said frequency component the supplementary winding of said transformer receiving electrical ener representing said intensity component, t e output ofthe secondary of said transformer constituting said actuating current in said means for recording saidffrequency component of said sound.

9. An apparatus for producing a sound record ofv more than'one component comprising a microphone for converting s ound waves into electric current variations, means comprising a transformer having a ferromagnetic core andra supplementary winding, a circuit including a current rectifier, said supplemental winding being included in said circuit and adapted to vary the ma etic saturation of said core independently o variations of saturation caused by other windings of said core'to separate the frequency component of said electric current variations from the intensity component thereof.

10. In sound recording `apparatus including a source of light, a polarizer, an electrooptic cell having electrodes, means to hold said electrodes under a substantially constant electrical bias, and an, analyzer; an optical compensator comprising a quarter-wave plate interposed between said e1ectro-optic cell and said analyzer, whereby the retardation of the polarized light entering said analyzer may be changed to a desired value.

11. A sound recording apparatus for pro-l ducing a sound record representing the separated amplitude component of the sound to be recorded, comprising a microphone, a thermionic tube containing filament, grid, and plate elements, a source of electromotive force interposed between resistances for supplying potential between the filament and the plate of said tube, means cooperating with said microphone for impressing differences of potential between the filament and grid of said tube, said differences of potential corresponding to sound vibrations incident on said vmicrophone, and recording means responsive to variations of potential across the filament and plate'of said tube caused by said Variations of potential between the ilament and grid of said tube. In testimony whereof I ax my signature.

NOEL DEISCH.

Images