US3178502A

US3178502A - Tone modifier

Info

Publication number: US3178502A
Application number: US160968A
Authority: US
Inventors: Jr Melville Clark
Original assignee: Individual
Current assignee: Individual
Priority date: 1961-12-11
Filing date: 1961-12-11
Publication date: 1965-04-13
Anticipated expiration: 1982-04-13

Description

Ill-

\l|[F1Illlllllllllllllllllllll LIII I April 13, 1965 M. CLARK, JR 3,178,502

TONE MODIFIER Original Filed Oct. 51, 1955 s Sheets-Sheet 1 [6- F ii iF -J T g INVENTOR.

May/4456242642 IZ M1 k wi April 13, 1965 M. CLARK, JR 3,178,502

TONE MODIFIER Original Filed Oct. 31, 1955 8 Sheets-Sheet 2 IN VEN TOR. Ma. V/LLE (2424 ./42

W1 MJJM'ZL April 1965 M. CLARK, JR 3,178,502

TONE MODIFIER Original Filed Oct. 31, 1955 8 Sheets-Sheet .5

7 /II I 8 2 Q L INVENTOR.

M54 WM: (249%, 42

April 1965 M. CLARK, JR 3,178,502

TONE MODIFIER Original Filed Oct. 31, 1955 8 Sheets-Sheet 5 IHIIH 111mm 1 rm! IIIHH HUI] L: k m

||||||||||||||ll HIE E 1 l u F I \9 3 Q R Q \E v v HUI] ; IHIH P N i m k M. CLARK, JR 3,178,502

TONE MODIFIER a SheetgLSheet s INVENTOR. fl/flV/Mi (2424 .42 24 m III [III milFUIl |l1lLl| I ElLllll II lllllHHl HI April 13, 1965 Original Filed 001;. 31, 1955 M. CLARK, JR

TONE MODIFIER April 13, 1965 8 Sheets-Sheet '7 Original Filed Oct. 31, 1955 INVENTOR. Mar/u: (2424 A2 M M W April 13, 1965 M. CLARK, JR

TONE MODIFIER 8 Sheets-Sheet 8 Original Filed Oct. 31, 1955 United States Patent 0 3,178,502 TONE MODIFIER Melville Clark, Jr., 8 Richard Road, Cochituate, Mass. Continuation of application Ser. No. 543,874, Oct. 31, 1955. This application Dec. 11, 1961, Ser. No. 160,968 21 Claims. (Cl. 841.24)

This invention relates to tone modifiers for use with electrical musical instruments and the like, and in particular to apparatus for modifying the frequency spectrum of an audio-frequency electric signal having the frequency components of a musical tone to produce vibrato, tremolo, and choral effects.

This case is a continuation of application No. 543,874, filed October 31, 1955 and now abandoned.

In many musical instruments, there is provided by various means an audio-frequency electric signal having the frequency components of a musical tone, and this signal is supplied to one or more loud-speakers, which convert the electric signal into sound waves. Well-known examples of such electrical musical instruments include electric organs, electric guitars, phonographs, tape recorders, and radios. In fact, any musical instrument can be made into an electrical musical instrument by providing a microphone or other pickup, an amplifier to convert sound waves into electric signals, and a loud-speaker system to convert the electric signals back into sound waves. The present invention is useful with any of these musical instruments, and operates by modifying the frequency spectrum of the audio-frequency electric signal to produce corresponding modifications of the musical tone produced by the loud-speaker system.

Various types of tone modification may be desirable at certain times to provide a variety of musical effects. For example, modification of a solo tone may be desired for the production of a choral effect or thick tone, that is, the musical effect produced by a plurality of similar musical instruments playing the same note simultaneously. Such tone modification is especially useful in the case of electric organs that are designed or adjusted to simulate the tone of some other musical instrument. As electric organs are commonly built, a given note of a particular clavier can simulate but one instrument of a specified type playing that note. The musical capabilities of such an organ would be increased if a given note played on it could simulate many musical instruments of one type playing the same note simultaneously. As another illustration, a small orchestra or band has only a limited number of instruments of each type. The musical capabilities of such an orchestra or band would be increased if it could produce the effect of many more instruments of each type playing substantially in unison. Accordingly, an object of this invention is to provide a tone modifier for modifying solo tones to provide choral effects.

Other musical effects are sometimes desired that may not be within the capabilities of a particular musical instrument. For example, a musician may desire to produce a tremolo, that is, a cyclic amplitude variation of the musical tone. Or the musician may desire to produce a vibrato, that is, a cyclic pitch variation of the musical tone. Accordingly, another object of this invention is to provide improved tone-modifying apparatus for producing tremolo and vibrato effects.

Another object of this invention is to provide tone modifiers for producing timbre scintillations and other musical effects. As herein used, the term timbre scintillation refers to a temporal variation of the timbre of a musical tone about its average character. Still further objects and advantages of the invention will appear as the description proceeds.

Briefly stated, and in accordance with one aspect of this invention, novel tone modifiers are provided that modulate an audio-frequency electric signal with a plurality of different subaudio frequencies to provide various musical effects, including the thick tone or choral effect of a plurality of similar instruments playing the same note simultaneously. Subaudio frequencies suitable for this purpose generally lie within the range of 0.5 to 20 cycles per second. All frequencies up to 20 cycles per second are considered subaudio frequencies for purposes of this patent application, although subaudio frequencies in the range of 2 to 7 cycles per second are preferred and are generally employed. In one embodiment of the in vention, hereinafter more fully described, modulation or modification of the tone is accomplished by means of magnetic recording apparatus having a plurality of oscillating heads by means of which the signal is pitch modulated with a plurality of different subaudio frequencies. In other embodiments hereinafter described, modulation or modification of the tone is provided by electrical networks having cyclically varying transmittance characteristics. In still another embodiment of the invention, modulation or modification of the tone is effected by the use of novel frequency translation or carrier techniques. In general, the modulation of an audio-frequency tone or signal with subaudio frequencies provides sidebands within narrow frequency spectra about each audiofrequency component of the original or unmodified tone or signal.

The invention will be better understood from the following description taken in connection with the accompanying drawings, and its scope will be pointed out in the appended claims. In the drawings,

FIG. 1 is a plan view of novel magnetic recording apparatus that is part of a first embodiment of this invention;

FIG. 2 is a section taken along the line 2-2 of FIG. 1:

FIG. 3 is a section taken along the line 33 of FIG. 1;

FIG. 4 is a section taken along the line 4-4 of FIG. 1;

FIG. 5 is a circuit diagram of the first embodiment;

FIG. 6 is an elevation of a switch used in the first embodiment;

FIG. 7 is a schematic and circuit diagram of a second embodiment;

FIG. 8 is an elevation of a switch used in the second embodiment;

FIG. 9 is a schematic and circuit diagram of a third embodiment;

FIG. 10 is a block diagram of a fourth embodiment;

FIG. 11 is a schematic and block diagram of a fifth embodiment.

Referring now to FIGS. 1 through 4 of the drawing, a novel magnetic recording apparatus includes a magnetic drum 1 attached to a shaft 2 that is continuously rotated at a constant speed by suitable means such as an electric motor 3 connected to shaft 2 by a driving belt 4. Adjacent to the periphery of drum 1 there are a plurality of conventional magnetic recording heads 5, 6, and 7, a conventional reproducing head 3, and a conventional erasing head 9. Recording heads 5, 6, and 7 are narrow heads spaced across the width of drum 1, as is best shown in FIG. 1, to provide three parallel magnetic recordings on the periphery of the magnetic drum. Reproducing head 8 is sufficiently Wide, at least in the portion thereof adjacent to drum 1, to receive all three of the magnetically recorded signals, and provides a reproduced electric signal having frequency components that are the sum of the frequency components of the three magnetic recordings. Alternatively, three narrow side-by-side reproducing heads may be used and connected together electrically to provide the same reproduced signal. Erasing head 9 is sufficiently Wide to erase all three of the magnetic recordmgs.

In the embodiment illustrated, the reproducing head 3 and the erasing head 9 are stationary, and these two heads are supported by stationary supporting members 10 and 11 attached to the frame or housing (not shown) of the recording apparatus. Recording head 5 is attached to a U-shaped support 12 that is rotative on a pair of stationary sleeves 13 and 14, held in fixed position by any suitable means (not shown), which also support the shaft 2. Consequently, recording head 5 is movable in a circumferential direction with respect to drum 1. In a similar way, recording head 6 is attached to a U-shaped support 15 that is rotative on sleeves 13 and f4, and recording head 7 is attached to a U-shaped support 16 that is rotative on sleeves l3 and 14. Consequently, the recording heads 5, 6, and 7 are independently movable in a circumferential direction about drum 1.

A plurality of

cams

17, 18, and 19 are attached to a shaft 20 that is rotated at constant speed by suitable means such as a driving belt 21 connecting drums 2-2 and 23 carried by

shafts

2 and 2%, respectively. Cam 17 operates a cam follower 24 pivoted on a stationary shaft 25 and connected by a link 26 to the rotative support 12 that supports recording head 5. As cam 1'7 rotates, recording head 5 is oscillated at a subaudio frequency in the direction of motion of magnetic drum 1.

In a similar way, cam 18 operates a cam follower 27 pivoted on shaft 25 and connected by link 28 to the rotative support 15 that supports recording head 6, so that as cam 18 rotates recording head 6 is oscillated at a different subaudio frequency in the direction of motion of magnetic drum 1. Cam 19 operates a cam follower 29 pivoted on shaft 25 and connected by a link 39 to the rotative support 16 that supports recording head 7, so that as cam 19 rotates recording head i is oscillated at still another subaudio frequency in the direction of motion of drum 1.

Cam follower 29 is held in contact with cam 19 by the force supplied by a spring 31 connected between U- shaped member 16 and a stationary support 32, as is best shown in FIG. 2. Similar springs, not shown, are connected between stationary support 32 and the U-shaped members 12 and 15, so that

cam followers

24 and 27 ac held in contact with cams 17 and 18 respectively.

Drum It may be a conventional magnetic recording drum having on its peripheral surface any suitable magnetic recording medium, such as a coating of a magnetic oxide. To provide a high-quality recording system, drum 1 preferably has a peripheral speed of about inches per second. For example, drum 1 may be about 10 inches in diameter and may be rotated in a clockwise direction, as indicated in the drawing by arrow 33: (FIG. 2) at a constant speed of 120 revolutions per minute.

Shaft 20 and each of the earns 17, 18 and 1% are rotated in a clockwise direction at a constant speed of 12 revolutions per minute, or 1 revolution each 5 seconds. The periphery of each cam has a contour corresponding to a different plurality of substantially sinusoidal waveforms, so that each of the recording heads 5, 6, and '7 is oscillated in a substantially sinusoidal manner at a different subaudio frequency. For example, the contour of cam 19 includes 11 complete sine-wave cycles, and rotation of the cams at 12 rpm. oscillates reproducing head 7 at a subaudio frequency of 2.2 cycles per second. The contour of cam 18 includes 17 complete sine-wave cycles, as is shown in FIG. 3, and rotation of the cams oscillates recording head 6 at a subaudio frequency of 3.4 cycles per second. The contour of cam 17 includes 25 complete sine-wave cycles, as is shown in FIG. 4, and rotation of the cams oscillates recording head 5 at a subaudio frequency of 5 cycles per second.

The oscillation amplitude of each recording head is determined by the waveform amplitude of the associated cam and by the design or adjustment of the cam-follower linkage. Preferably each recording head has an oscillai tion amplitude that is inversely proportional to its oscillation frequency. For example, recording head 5 may be oscillated through the total distance of about 40 mils along the periphery of drum 1, while head 6 is oscillated through a distance of about 60 mils and head 7 is oscillated through a distance of about mils.

As is hereinafter more fully explained in connection with PEG. 5, the same audio-frequency signal may be supplied simultaneously to each of the recording heads 5, 6, and 7. Assume, for example, that a 4G0 cycle per second signal is supplied to recording head 7. The recording head produces a magnetic signal on the surface of drum 1, which reproducing head 8 converts into a reproduced electric signal. The wavelength of the recorded magnetic signal is proportional to the velocity of the drum periphery relative to recording head '7, and is inversely proportional to the frequency of the electric signal supplied to recording head 7. The frequency of the reproduced electric signal is inversely proportional to the wavelength of the recorded magnetic signal. Accordingly, if recording head '7 and reproducing head 8 were both stationary, while drum El rotated at constant speed, the frequency of the reproduced signal would be the same as the frequency of the signal supplied to the recording head, and under the assumed conditions would be 400 cycles per second.

However, as the recording head oscillates, the velocity of the drum relative to the recording head varies at a subaudio rate, and consequently the wavelength of the recorded magnetic signal also varies at a subaudio rate. Since the reproducing head 8 is stationary, the frequency of the reproduced signal is modulated with a subaudio frequency equal to the oscillation frequency of head 7, which in the case under consideration is 2.2 cycles per second. The frequency deviation of the reproduced signal is proportional to the oscillation amplitude and oscillation frequency of the recording head and, with the design parameters herein given, is approximately one percent of the center frequency. That is, when a 400 cycle per second signal is supplied to recording head 7, the reproduced signal provided by reproducing head i has a center or average frequency of 400 cycles per second, while the instantaneous frequency of the reproduced signal varies cyclically between limits of approximately 396 cycles per second and 404 cycles per second. In other words, the reproduced signal corresponds to a musical tone having an average pitch of 400 cycles per second and having a vibrato with a frequency range of about 8 cycles per second at a subaudio rate of about 2.2 cycles per second.

Now assume that an audio frequency of 800 cycles per second is supplied to recording head 7. This audio frequency is likewise modulated with a subaudio frequency of 2.2 cycles per second, and its frequency deviation is about 1 percent of the center frequency, or 8 cycles per second to each side of the center frequency making a total frequency range of 16 cycles per second. When the audio-frequency signal supplied to recording head '7 has the frequency components of a complex musical tone, a fundamental component of 460 cycles per second and an overtone or a second harmonic component of 800 cycles per second, for example, the frequency of each component is modulated with a sub-audio frequency of 2.2 cycles per second, and the frequency deviation of each component is substantially one percent of the center frequency of that component. When the instantaneous frequency of the fundamental component in the reproduced signal is 396 cycles per second, the instantaneous frequency of the second harmonic component of the reproduced signal is 792 cycles per second.

It is thus apparent that the original harmonic relation of an overtone to the fundamental is maintained, and that the frequency deviation of an overtone is greater than that of the fundamental by the same ratio that exists between the center frequency of the overtone and the center frequency of the fundamental. Consequently, a true pitch.

modulation is provided in which the pitch of the tone is cyclically varied at a subaudio frequency wi hout changing the harmonic structure or timbre of the tone.

Thus pitch modulation is somewhat different from true frequency modulation, since in frequency modulation the frequency deviation is not a function of the carrier frequency. If a signal having 400 cycle per second and 800 cycle per second frequency components were frequency modulated, the frequency deviation of both components would be the same, and in general the instantaneous frequencies would not have the same harmonic relation as the original or center frequencies of the same components.

In a similar manner, the audio-frequency signal supplied to recording head 6 is pitch modulated with a subaudio frequency of 3.4 cycles per second, and the audio-frequency signal supplied to recording head 5 is pitch modulated with a subaudio frequency of 5 cycles per second. The reproduced signal is a combination of these three pitch-modulated signals, and con equently represents a signal having a complex modulation pattern corresponding to simultaneous pitch modulation of the original signal with the three different subaudio frequencie In other words, the reproduced signal has a complex vibrato pattern such as would be produced by playing the same note on three similar musical instruments simultaneously, with three different vibrato rates. The musical eff ct of such complex vibrato patterns is a thick tone or choral effect, so that a sound Wave produced from the reproduced electric signal gives the musical impression of several similar musical instruments playing the same notes simultaneously.

Theoretically, an ideal situation for the production of pleasing choral effects would be achieved by making the three subaudio modulating frequencies asynchronous, so there would be no periodic repetition of the modulation pattern. In practice, a satisfactory musical effect can be obtained by making the period of the complex modulation pattern relatively long, 5 seconds for example. in the embodiment illustrated in FIGS. 1 through 4, this relatively long period of the complex modulation pattern is achieved by making the number of complete sinusoidal waveforms in respective ones of the cam contours large integers having no common factor, such as ll, 17, and 25, and by rotating the cams at a relatively slow speed, such as 12 rpm. With this arrangement there is no repetition in the modulation pattern during one compiete rotation of the cams, and the modulation pattern repeats itself only at the relatively slow frequency of once each 5 seconds. Minor imperfections in the cam contours or in other parts of the modulating apparatus are generally not harmful, but merely add to the asynchronism or randomness of the modulation pattern. Conse quently, high precision cams are not required and manufacture of the magnetic recorder modulating apparatus can be relatively inexpensive.

When each recording head has an oscillation ampli ude inversely proportional to its oscillation frequency, he frequency deviation of the modulated signal produced by each of the modulation frequencies is substantially the same, preferably about 1 percent. Here also the manufacturing tolerances are not severe, since some variations in the degree of modulation tend to enhance rather than to detract from the pleasing musical effect.

The erasing head 9 is supplied with a relatively highfrequency current, about 150 lzilocycles per second for example. This produces a high-frequency magnetic field across the gap of the erasing head that erases the magnetically recorded si nals on drum 1, as is Well l: own by those skilled in the magnetic recording art, and provides at all times a supply of unmagnetized magnetic medium to the recording heads.

Preferably, the recording heads 5, 6, and 7 are spaced apart along the periphery of drum 1 at about 3 inch intervals. The signal recorded by recording head 7 is reproduced by reproducing head 8 after a time delay dependin upon the rotational speed of drum 1 and the distance between head 7 and head 8. The signal recorded by head 6 is reproduced after a greater time delay proportional to the distance between head 6 and head 3, and the signal recorded by head 5 is reproduced after a still greater time delay proportional to the distan e between head 5 and head 8. Consequently, when a signal representing a musical tone is supplied to heads 5, 6, and 7, simultaneously, a portion of this signal recorded by head 7 is reproduced first, then after a time delay the signal recorded by head 6 is reproduced, and after a still further time delay the signal recorded by head 5 is reproduced. This is musically desirable, since when the same note is played upon three separate musical instruments, the three instruments do not begin to play the note at precisely the same time, nor do they stop playing the note at precisely the same time. When the reproduced signal provided by head 8 is converted into sound waves, a similar musical effect is obtained because of the different time delays involved in reproducing the signals recorded by the three heads.

Various modifications in the magnetic recording and modulating apparatus are possible without departing from the inventive principles involved. For example, numerous variations can be made in the mechanical structure, such as the use of gear trains in place of belts to interconnect the rotativc parts, the use of cranks in place of cams to produce oscillatory motion, and the use of a magnetic tape, disc, or the like, in place of the magnetic drum. Furthermore, instead of using a plurality of oscillatory recording heads with a single stationary reproducing head, a single stationary recording head may be used with a plurality of oscillatory reproducing heads, or the recording and reproducing heads may both be oscillatory.

instead of three parallel heads each oscillated sinusoidally at a difierent subaudio frequency, a single recording system may be used having only one recording head and one reproducing head. Mutliple modulation of the signal may then be provided by oscillating either the recording head or the reproducing head, or both, in a complex manner corresponding to the sum of several sinusoidal oscillations. For example, mechanical linkages may be provided to add all or selected ones of the motions of cam followers 2 27, and 29, and the combined motion so provided may be used to oscillate a single recording head. Because of the time delay between the recording and reproducing operations, modulation of the signal may occur even if the recording head and the reproducing head are oscillated in synchronism with each other, and are separated at all times by a constant distance. Alternatively, both the recording head and the reproducing head may be stationary, and the drum speed may be modulated to pitch-modulate the reproduced signal.

in a broad sense, the magnetic recording drum is simply a delay or memory device for transmitting and storing signals representing a musical tone, while the recording and reproducing heads are scanning devices for scanning the signals transmitted by the delay device. Accordingly, other delay or memory devices, such as electrical delay lines, may be employed, in conjunction with scanning capacitors or the like for periodically scannin the signals being transmitted by the delay line.

Instead of recording the audio-frequency signal directly, the audio-frequency signal may first be amplitude-modulated upon a carrier-frequency signal, and the modulated carried may be recorded upon the magnetic drum. In the reproduced signal, the carrier signal and its sidebands are pitch-modulated with the subaudio modulation frequencies and upon subsequent demodulation of the carrier to recover the audio-frequency components, it will be found that the audio-frequency signal is pitch-modulated in the same manner as if the audio-frequency signal had been directly recorded. With some magnetic recording 7 media, the use of such a carrier system improves the signal-to noise ratio.

Reference is now made to FIG. of the drawing, which illustrates the electrical circuit of a tone modifier embodying principles of this invention. An electrical musical instrument 34 provides an audio-frequency electri signal having the frequency components ot a musical tone. lnstrument 34- may be an electric organ, an electric guitar, or any other device providing an electric signal that may be converted into sound waves for the production of musical tones. The conversion of electric signals into sound waves is accomplished by one or more lou speakers 35. The tone modifier herein described modifies the musical tone produced by loud-speaker by modifying the electric signal provided by musical instrument 3d.

The complete tone modifier preferably includes a plurality of the magnetic recorders illustrated in FlG. l4-, connected in tandem in a manner to be described. For example, in the tone modifier illustrated, there are three of the magnetic recorders, which may all be similar eXcept that their sub-audio modulation frequencies preferably are different. For this purpose, one of the magnetic rccorders may have three cams with contours of ll, 17, and 25 complete sine-wave cycles, respectively, while a second one of the three recorders has three cams with contours of 12, 19, and 29 complete sine-wave cycles, and the third one of the three recorders has three cams with contours of 13, 23, and 31 complete sine-wave cycles. It should be noted that no pair of these cam numbers have any common factor. Consequently, nine different subaudio modulation frequencies are provided, which for practical purposes are asynchronous.

The three magnetic recorders may be completely separate units except for their electrical interconnections, or they may have certain mechanical parts in common. For example, all three recorders may use the same mag netic drum, with three sets of recording and reproducing heads at different positions along its length.

.lternatively, the three magnetic recorders may have identical sets of camsthat is, the cam numbers may be the same in all three recorderslll, l7 and 25, for example. Asynchronism between the three sets of modulation frequencies can be achieved by rotating the three sets of cams at slightly ditlerent speeds, by making the three driving pulleys for the cams of slightly difierent diameters, for example, or by operating the driving motors of the three recording systems asynchronously.

The audio-frequency signal supplied by musical instrument 34 is transmitted by a lead as to four conventional vacuum-tube amplifiers 3?, 38, 39, and 4 9. Amplifier 37 supplies this signal to a lead at connected to the three recording heads 5, 6, and '7 of one magnetic recorder. Switches 4-2, 3, and 44 are connected between respective ones of these three recording heads and the ground lead 45', as shown, so that the electrical circuit through each recording head is completed only when the associated switch is closed. The reproduced electric signal provided by reproducing head 8 is amplified by a conventional vacuum-tube amplifier id and supplied to a lead 4'7 that also receives through amplifier 3 5 the original signal provided by musical instrument 3%. Consequently, lead 47 receives both the original signal and the reproduced signal provided by reproducing head 3.

The second magnetic recorder includes a continuously rotated magnetic drum 48, three oscillatory recording heads 49, 5t), and 51, a reproducing head 52, and an erasing head 53. The signal provided at lead 47 may be supplied to each of the recording heads 49, 5t and 51 through a plurality of switches 54, 55, and 56, connected between respective ones of these recording heads and the ground lead :5, as shown. The reproduced electric signal provided by reproducing head 52 is amplified by a conventional vacuum-tube amplifer 57 and is supplied to a lead 53 that also receives the original signal from the musical instrument 34 through amplifier 39.

The third magnetic recorder comprises a continuously rotated magnetic drum 59, a plurality of oscillatory recording heads so, til, and 62, a reproducing head 63, and an erasing head 64. The signal provided at lead 58 may be supplied to each of the recording heads 69, (,1, and 62 through a plurality of switches 65, 66, and 67 connected between respective ones of these recording heads and the ground lead 45, as shown. The reproduced electric signal provided by reproducing head 63 is amplified by .a conventional vacuum-tube amplifier 68 and is supplied to a lead that also receives the original signal from musical instrument El -l through amplifier 49. Lead as is connected to the loud-speaker 35 for converting the electric signals into sound waves.

The three erasing heads 9, 53, and 64 are connected to erase oscillator it that supplies a kilocycle per second erase signal to the three erasing heads. Cons quently, each of the recording drums 1, 48, and 59 is ,letely erased at each revolution. A normally closed switch "1 is connected between amplifier 4t and lead 69,

- to a normally open switch '72 connected in ll or" the switches are in the positions n in FIG. 5. S itches d2, 43, 44, 54, 55, 56, 65, 66, and '72 are open, w ile switch 71 is closed. The electric circuits through all nine recording heads are now open,

duced electric signal that is transmitted through amplifier and lead to loudspeaker 35. This reproduced signal is pitchanodulated at a sub-audio frequency, as hereinbefore explained, simulating a musical tone having a vibrato. Loudspeaker 35 now receives two signals, unmodified signal through amplifier 4t and a modulated signal through amplifier 68, and the musical tone produced by the loudspeaker corresponds to a mixture of these two signals. The instantaneous frequencies of the two gnals are different, and as a result a musical choral fleet is produced similar to that of two similar musical nstruments playing the same note simultaneously.

Now assume that switches 66 and 67 are both closed.

reproduced signal supplied through lead 69 to lou 3-5 is now pitch-modulated wih two diilerent submusical instruments playing the same notes simultaneously with different vibrato rates. This signal is mixed with the unmodified signal supplied th ough amplifier to produce a choral effect similar to that of the three musical instruments playing simultaneously. Similarly, when switches 65, 66, and 67 are all closed, a choral effect similar to at of four musical instruments playing simultaneously is produced.

Now assume that switches 56, 65, 66, and 67 are all closed. The unmodified audio-frequency signal provided by instrument 3% is supplied through amplifier 38 to recording head 51, which records an audio-frequency signal on magnetic drum 48. Reproducing head 52 now supplies a re: roduced signal through amplifier 57 to lead 58 that is pitch-modulated at a subaudio frequency. This modulated signal is mixed with the unmodified signal supplied to lead 58 by amplifier 39, and the two signals are simultaneously supplied to each of the recording heads '51, and

As a result, the three signals recorded upon drum 59 are comple. each representing a combination of pitchmodulated and unmodulated signals. Oscillation of the

heads

69, 61, and 62 further pitch-modulates this complex signal, so that a signal is supplied to lead 69 through amplifier 68 that has a relatively complex modulation pattern containing the three modulation frequencies introduced by oscillation of heads 60, 61, and 62 plus combinations of each of these three modulation trenquencies with the modulation frequency introduced by oscillation of recording head 51. Consequently, six different modulaion frequencies are present in the signal transmitted by amplifier 68, and when this signal is mixed with the unmodified signal transmitted by amplifier 49, the musical tone produced by loud-speaker 35 has a choral effect similar to that of seven similar musical instruments playing the same notes simultaneously. The vibrato pattern is now so complex that it is generally not recognizable as such, and the total effect produced is one of thickness of tone.

Now assume that switches 55, 56, 66 and 67 are all closed. The signals supplied to lead 58 represent th ee instruments playing simultaneously, and each of these signals is modulated by the three modulation frequencies provided by oscillation of

heads

69, 61, and 62, so that amplifier 68 supplies to lead 69 a signal modulated with nine different subaudio frequencies. To this signal there is added the unmodified signal supplied through amplifier 4f), so that loud-speaker 35 now provides a tone that simulates ten similar musical instruments playing the same notes simultaneously.

As successively greater numbers of the switches are closed, the choral effect produced is that of successively larger numbers of similar musical instruments playing the same notes simultaneously. For example, closing switch 5d increases the number or" instruments represented in the choral effect to 13, closing switch 44 increases the number of instruments represented to 22, closing switch 43 increases the number of instruments represented to 31, and closing switch 42 increases the number of instruments represented to 40. It will be noted that the tandem arrangement of the magnetic recorders multiplies the tone thickness produced by each, so that choral effects representing a very large number of musical instruments are produccd with relatively few modulators. By adding additional magnetic recording modulators in a similar manner, choral effects representing any number of similar musical instruments playing the same notes simultaneously can be obtained. The choral effect of thousands of instruments playing simultaneously can be produced with relatively small, compact apparatus operated by a single musician.

A still better understanding of the multiplicative effect of the tandem-connected modulators can be obtained by con dering their operation from the standpoint of modulation theory. The electrical musical instrument 3% supplies an electric signal havin a plurality of frequency components corresponding to the frequency components of a complex musical tone. Assume, for example, that there is a fundamental component having a frequency of 400 cycles per second and a second harmonic component having a frequency of 800 cycles per second. In practice there will generally be many other frequency components, but a consideration of two components is sufficient for illustrative purposes. Each of these frequency components is pitch-modulated with three different subaudio frequencies by the oscillation of recording heads 5, 6, and '7, so that the signal provided at lead 47 has, in addition to components at the initial frequencies of 400 and 800 cycles per second, a plurality of sidebands symmetrically spaced in the frequency spectrum about each of the initial frequency components and separated therefrom and from each other by frequency differences equal to each of the subaudio modulation frequencies plus multiples and combinations thereof.

In other words, modulation of the 400 cycle per second signal component by oscillation of head '7 at a subaudio frequency of 2.2 cycles per second produces sidebands spaced in frequency on each side of the 400 cycle coni pouent at intervals of 2.2 cycles per second and multiples thereof. Similarly, sidebands are produced on each side of the 800 cycle per second component. Oscillation of recording head 6 at 3.4 cycles per second produces on each side of each frequency component sidebands spaced at intervals or" 3.4 cycles per second, and oscillation of head 5 at 5 cycles per second produces sidebands on each side of each frequency component spaced at intervals of 5 cycles per second.

in addition, modulation with several frequencies simultaneously produces sidebands spaced at intervals corresponding to combinations of the modulation frequencies, according to frequency-modulation theory. Thus, there is produced for each frequency component a complex spectrum of frequencies centered on the original frequency component, and having a width substantially equal to twice the sum of the maximum frequency deviation and the highest modulating frequency. Assuming a maximum frequency deviation of 1 percent and a maximum modulation frequency of 5 cycles per second, the frequency spectrum centered on the 400 cycle per second component Will be about 18 cycles per second wide, while that centered on the 800 cycle per second will be about 26 cycles per second wide.

The signal at lead 47 has a much larger number of frequency components than the original signal provided by musical instrument 34, but since these components are grouped about the original frequency components, the pitch and timbre of the musical tone are not materially changed. The principal effect produced by the tone modification is a thickening of the tone or a choral effect, similar to that produced by many similar musical instru ments playing the same notes simultaneously.

Each of the frequency components present at lead 47 is modulated with three different subaudio frequencies by oscillation of recording heads 49, 5t and 51, to produce an even larger number of sidebands centered about each of the original frequency components. In a similar way, even more sidebands are added by the oscillation of heads 64., 61, and 62, so that the signal supplied to loud-speaker has an exceedingly large number of sidebands grouped about each of the original frequency components and spaced therefrom and from each other by different subaudio frequency differences. The result is an exceedingly thick tone, giving a choral effect similar to that of a very large number of similar musical instruments playing the same notes simultaneously.

Now assume that switches 42, 43, 44, 54, 5'5, 56, 65, 65, and 67 are all open, as shown in FIG. 5, but that switch 71 is open while switch 72 is closed. The audio-frequency signal supplied by musical instrument 34 is now pitchmodulatea at a single subaudio frequency by oscillation of head 52, and this single pitch-modulated signal is supplied to loud-speaker 35. Under these conditions the pitch of the musical tone produced varies periodically at a subaudio frequency, 6.2 cycles per second for example, and the musical eifect is that of a single instrument played with a 6.2 cycle per second vibrato. Consequently, the same tone modifier can be operated to produce either choral or vibrato effects, or both, selectively, depending upon the switch positions. Switches 71 and 72 may be operated by the musician to provide vibrato effects whenever such effects are desired.

Reference is now made to FIG. 6 which illustrates a convenient switching arrangement for controlling the pro duction of choral effects. Switches 42, 43, 44, 54, 55, 56, 65, 66, and 67 are formed from a stack of metal spring leaves arranged as shown and insulated from one another by strips of insulation at one end of the stack. Leaf 42 is connected to recording head 5 by lead 5, leaf 43 is connected to recording head 6 by lead 6', and so forth, the lead to each recording head being identified by the same reference numeral with a prime added. At the bottom of the stack, there is a metal spring leaf 45 which ears is connected to the ground lead 45. The stack of spring leaves is held in position by suitable means such as rivets 73 and 74 passing through the insulating strips between the spring leaves into a stationary supporting member 75.

A choral pedal 76 is pivotally supported by a stationary shaft '77 and is biased to a rest position, as shown, by spring '78. When the left-hand end of pedal 76 is depressed, the right-hand end of the :edal moves spring leaf 45 upward into contact with spring leaf 67, thereby connecting recording head 62 to the groun lead 45 and permitting an electric signal to pass through recording head 62. Further depression of pedal it; moves spring leaf 67 into contact with spring leaf 66, whereupon recording heads 61 and as are both connected to lead 45' and receive electric signals from lead 53. Further depression of pedal 7e connects successive ones of the recording heads to lead 45, and thereby provides successively greater thickness of the musical tone. in this way the musician can easily control the thickness of the tone to provide choral efiects representing different numbers of similar musical instruments playing the same notes simultaneously.

Reference is now made to FIG. 7, which illustrates a somewhat different tone modifier embodying principles of this invention. Electrical musical instrument 79 provides at lead fill an audio-frequency electric signal having the frequency components of a musical tone. This signal is transmitted through a resistor 31 to a lea $55, which is connected to ground lead 83 by a resistor 34. The resulting voltage across resistor is amplified by a conventional vacuum-tube amplifier 85, which supplies an electric signal to a l ad 855. T is signal is transmitted through a resistor 07 and a lead to a resistor S9, which is also connected to ground lead The voltage across resistor 89 is amplified by a conventional vacuum-tube amplifier 99, which supplies an electric signal to one or more loud-speakers $1. Loud-speaker 9i converts the electric signal into sound waves to produce a musical tone.

A variable transmittance electrical network comprises a rotary scanning switch 92 having a plurality of commutator segments connected as shown to taps of a voltage-divider resistance network consisting of resistors 93, 94, $5, 96, 9'7, 58, 99, and 1%, connected as shown between lead and lead 83. The contact arm of scanning switch 92 is continuously rotated at low speed, 2.2 revolutions per second, for example, by a train of gears 1M and 1% connected to a shaft 1% that is continuously rotated at constant speed by suitable means such as an electric motor llld. Consequently, the electrical transmittance of the network is modulated or varied at a subaudio frequency of 2.2 cycles per second. The contact arm of scanning switch 92 is connected to lead 82 through a resistor 105 and a normally open switch res. When switch 1&5 is closed, scanning switch @2 supplies across resistor 84 a signal that is amplitude modulated with the subaudio frequency of 2.2 cycles per second, which signal is in addition to the unmodulated signal transmitted through resistor 31. Consequently, whenever switch res is closed, two electric signals are transmitted simultaneously to loud-speaker 3'31, one of which is amplitude modulated at a subaudio frequency while the other is unmodulated. if desired, the parallel volt age dividers 93% and FF-lu l may be combined into one voltage divider to reduce the number of resistors by a factor of two.

A similar variable transmittance network includes a' second scanning switch M37 that is similar to switch $2 except that it is rotated at a difierent low speed, 3.4 cycles per second for example, by a train of gears @3465: connected to shaft 193. Consequently, the transmittance of this second network is modulated at a subaudio frequency of 3.4 cycles per second. Switch 107 is connected to lead by a resistor 11s: and a normally open switch ill. When switches 1'36 and 111 are both closed,

CHI

l2 three signals a e transmitted simultaneously to loudspeaker 9, a first one of wh ch is amplitude modulated with a subaudio frequency of 2.2 cycles per second, a second one of which is amplitude modulated at a subaudio frequency of 3.4 cycles per second, and the third one of which is unmodulated.

Another similar variable transmittance network includes a third selector switch 112 that is similar to selector switch 92 except that it is rotated at still another low speed, 5 cycles per second, for exam le, by a train of gears 11., 14 connected to shaft 363. The contact arm of switch 112 is connected to lead through a resistor 125.5 and a normally open switch Elle. Fiiien switch 116 is closed, there is transmitted to loud-speaker 91 an additional signal which is amplitude modulated with a subauuio frequency of 5 cycles per second. If desired, the voltage dividers connected to the commutator segments of scanning switches 92, 197 and 112 may be combined into a single voltage-divider network for reducing the number of resistors.

Other similar variable-transmittance networks include scanning switches 12?, H3, and 113 which are rotated at still other low speeds corresponding to different su audio frequencies by gear trains connected to shaft 1%, as shown. These scanning switches are connected to lead through resistors 11%, 12.1 and 322, in serie with noramlly

open switches

123, 124 and 12-5, respectively. The resistance networks associated with scanning svitches 1 3, and 319 are connected between lead as and lead 33, so that these scanning switches amplitude modulate all of the electric signals or signal components supplied to lead This provides a multiplication of the modulation frequencies, in the n anner hereinoe ore described in connection with 5, so that a very thick tone is produced by loudspeaker 91 when all of the switches rss, 111,

li

123, 124 one 125 are closed. Additional scanning switches may be provided to further multiply the thickness of the tone to desired extent.

Assume that the audio-frequency signal supplied to lead frequency components of 400 cycles per second and 809 cycles per second. These frequency components are transmitted through amplifiers 35 and 9% to loud-speaker $3, so that when all of the switches are open the tone produced by the loud-speaker has the same frequency compone ts as the electric signal supplied by the elecrical musical instrument. When switch ass is closed, each frequency com onent of the original signal is amplitude modulated by the rotation of scanning switch thereby producing for each of the original frequency components two side bands spaced on either side thereof in the frequency spectrum at a frequency diference of 2.2 cycles per second. Such sidebands represent simple amplitude modulation, and the tone produced under such circumstances by loud-speaker 91 is like that of a single musical instrument played with the tremolo of 2.2 cycles per second.

When switches res and Ill are both closed, there are produced two sidebands on each side of each frequency component, spaced therefrom at frequency differences of 2.2 and 3.4 cycles per second, respectively. Under such circumstances the tone produced by loud-speaker 9i is amplitude modulated with two subaudio frequencies simultaneously, and the musical erlect is produced of two similar instruments playing the same notes simultaneously with different tremolo rates. Similarly, when switches 1%, 111, and are all closed, the tone is modulated with three difierent subaudio frequencies, and a choral effect representing three similar musical instruments playing the same notes simultaneously is provided.

When all three of the switches rss, 111, and 116 are closed, the signal transmitted by amplifier 85 to lead as consists of the original frequency components provided by musical instrument 7?, plus six sidebands for each of the original frequency components, so that the signal at lead has seven times as many frequency components as the signal at lead 89. When switch 123 is closed, all of the frequency components present at lead 86 are amplitude modulated with a subaudio frequency by the network including scanning switch 117, thereby pro viding a pair of sidebands for each frequency component supplied to lead 86. Now there are 21 frequency components for each of the original frequency components supplied to lead 80. When switch 124 is closed, another pair of sidebands is provided for each of the frequency components provided at lead 86, and when switch 125 is closed still another pair of sidebands is provided for each of the frequency components at lead 86. The multiplication of frequencies thus produced provides a choral effect that represents a large number of similar musical instruments playing the same notes simultaneously.

When all of the switches 1'56, 111, 116, 12 3, 124, and 125 are closed, there is provided for each of the original frequency components a total of 48 sidebands spaced at subaudio frequency intervals within a small frequency spectrum centered on the original frequency component. The total width of each such spectrum is equal to twice the sum of the highest modulation frequency provided by scanning switches 92, 1117, and 112 and the highest modulation frequency provided by the scanning switches 117, 118, and 119. If the modulation frequency provided. by switch 112 is five cycles per second, and is higher than that provided by switches 92 and 167, and the scanning frequency provided by switch 119 is 6.2 cycles per second, and is higher than that provided by switches 117 and 118, the total width of the frequency spectrum of sidebands about each modulation component is 22.4 cycles per second.

FIG. 8 illustrates a choral effects control that may be used in the apparatus shown at FIG. 7. Switches 1%, 111, 116, 123, 124 and 125 consist of a stack of metal spring leaves arranged as shown and insulated from one another by strips of insulation. An additional metal spring leaf S2 is connected to lead 82, and still another metal leaf 88' is connected to lead 88 as shown. A choral pedal 126 is pivotally supported by a stationary shaft 127. When the eft-hand end of pedal 126 is depressed, the right-hand end of the pedal presses upward against an insulating strip 12% attached to the bottom of leaf 82 and forces leaf 82' into contact with leaf 166, thereby connecting resistor 165 to lead 8-2. Further depression of the choral pedal moves leaf 195 into contact with lead 111, and thereby connects resistor 110 to lead 82. Still further depression of the pedal moves lead 111 into contact with leaf 116 and con nects resistor 115 to lead 82. An insulating strip 129 insulates leaf 116 from leaf 88' at all times. Accordingly, still further depression of the choral pedal moves leaf 88' into contact with leaf 123 and connects resistor 120 to lead 88. Still further depressions of the choral pedal connect resistors 121 and 122 to lead 88.

Reference is now made to FIG. 9 of the drawing, which illustrates still another tone modifier embodying principles of this invention. Electrical musical instrument 135) supplies to lead 131 an audio-frequency electric signal having the frequency components of a musical tone. This signal is transmitted through resistor 132 and lead 133 to a resitor 134, so that a voltage appears across resistor 134 that has all of the frequency components present in the signal supplied to lead 131. This signal is amplified by an amplifier 135 that supplies an electric signal through lead 135, resistor 137 and lead 138 to a resistor 139 across which a voltage appears that has all of the original frequency components. his signal is amplified by a conventional vacuum-tube amplifier- 14i) and is supplied to one or more loudspeakers 141, which produce sound waves corresponding to the electric signal.

The signal at lead 131 is supplied to a plurality of variable transmittance formant circuits having transmittances that are a function of frequency. One such formant circuit comprises resistors 14 2 and 14-3, a switch 144 a capacitor 145, and a variable inductor 146, connected as shown. Capacitor 145 and inductor form a tuned circuit that is parallel-resonant at an audio frequency. Consequently, the parallel impedance of capacitor 145 and inductor 146 is different for different audio frequencies, and the signal supplied to resistor 143 is a modified signal having fre quency components with relative amplitudes different from the relative amplitudes of the corresponding frequency components in the original signal supplied to lead 131. It will be appreciated that formant circuits of other types may also be used, including formant circuits that do not include resonant selections but nevertheless have transmittances that are functions of frequency.

The inductance of inductor 146, and therefore the resonant frequency of the tuned circuit, is cyclically varied at a subaudio frequency by adjustment means connected through a train of gears 147448 to a shaft 149 that is con tinuously rotated at constant speed by suitable means, such as an electric motor 150. F or example, inductor 146 may comprise a coil on a magnetic core havin an air gap within which a non-circular or eccentric magnetic rotor is continuously rotated for cylically varying the reluctance of the magnetic circuit. As a result, this formant circuit modulates the amplitude and phase of each frequency component in the original audio-frequency signal with a subaudio frequency, which maybe 2.2 cycles per second, for example. Components near the resonant frequency may be amplitude modulated at twice this frequency. The amplitude and phase of this. modulation is different for different frequency components in the original signal, where by the timbre of the tone is modulated or scintillated. Consequently, the signal provided at resistor 143 has a varying or scintillating timbre. Alternatively, instead of varying the inductance of inductor 146, the capacitance of capacitor 145 or the resistance of resistor 142 may be varied cyclically at a subaudio frequency.

When considered in terms of modulation theory, the effect of formant circuit 142-146 is to produce sidebands about each frequency component of the original signal separated therefrom and from each other by subaudio frequencydifferences. The sidebands of each frequency component differ in their relative amplitude and phase relations from the sidebands of other frequency components, thereby providing a modulation or scintillation of timbre. When switch 14 is closed, the modulated signal is combined with the original unmodulated signal to produce a choral effect.

A second formant circuit comprises resistors 151 and 152, a normally open switch 153, a capacitor 154, and a variable inductor connected as shown. Capacitor 154 and inductor 155 constitutes a second tuned circuit that is parallel resonant at an audio frequency different from the resonant frequency of the circuit comprising capacitor 145 and inductor 145. The inductance of inductor 155, and therefore the resonant frequency of circuit 154-155, is cyclically varied by gear train 15-34157, connected to shaft 149, at a subaudio frequency different from the modulation frequency of illdtlCLOI' 146. F or example, the inductance of inductor 155 may be modulated with a subaudio frequency of 3.4 cycles per second. Thus, when switch 153 is closed there is supplied to lead 133 still another signal having a timbre that is modulated or scintillated at a nifferent frequency. Accordingly, additional sidebands are added to the signal provided at lead 133, to provide a choral effect with a somewhat thicker tone than was pro vided when only the switch 144 was closed.

Still another formant circuit comprises capacitor 158 and a variable inductor 159 tuned to still another audio frequency, with the inductance of inductor 159 modulated with still another subaudio frequency. Consequently, still another group of sidebands is added about each frequency component of the original signal when switch 161 is closed. When switches 144, 153, and 169 are all closed, a signal is provided at lead 136 that includes the original frequency components plus a considerable number of sidebands.

A further variable formant circuit comprises a capacitor 161 and a variable inductor 162 having an inductance that is modulated with still another subaudio frequency. Consequently, when switch 163 is closed, additional sidebands are provided for each frequency component at lead 136, which includes the frequency components of the original signal plus the sidebands provided by the first three variable formant circuits. Still another variable formant circuit comprises a capacitor 164, a variable inductor res having its inductance modulated with still other sidebands about each frequency component supplied to lead rse. A sixth variable formant circuit comprises a capacitor 167, a variable inductor res ha ing an inductance modulated with a sixth subaudio frequency, and a switch 169 that may be closed to provide even more sidebands about each frequency component. Any numer of additional variable formant circuits may be added as desired, to provide any desired number of sidebands about each of the original frequency components.

When all of the switches M4, 153, 16%, 53, 166, and 169 are closed, the signal supplied to loud-speaker 141 has a very large number of sidebands about each of the original frequency components, so that the choral effect of a very thicc tone is produced similar to that of a large number of similar musical instruments playing the same notes simultaneously. Switches 144, s53, 16%, 163, 166, and 169 may be operated by a choral pedal similar to that shown in PEG. 8.

The embodiments hereinbefore described illustrate the production of choral effects, vibrates, tremolos and timbre scintillations by direct modulation of an audio-frequency signal. Choral effects and other musical effects can also be obtained by indirect modulation of the audio-frequency signal effected by carrier-current techniques, as is illustrated in two embodiments that will now be described.

in FlG. 10 of the drawing, an electrical musical instrument 357i provides at lead 1?]. an audio-frequency electric signal having the frequency components of a musical tone. An oscillator 172 supplies to lead 173 an el ctric signal having a carrier frequency, 100 kilocycles per second, for example. For purposes of this patent application, the term carrier frequency" refers to any frequency that is higher than the highest audio-frequency component of the musical tones that are produced. This carrier-frequency signal is transmitted by a buffer ampliher 174, a lead 1'75 and a normally closed contact of a two-position switch 176 to a balanced modulator 177. The audio-frequency signal at lead l'il is also sup, d to balanced modulator 177, which heterodynes the carrierfrequency and audio-frequency signals to provide at lead 178 a plurality of sidebands having frequencies equal to the sums and differences of the carrier frequency and each frequency component of the audio-frequency signal.

These sidebands are transmitted by lead 1 8 to a mixer 179 that alsoreceives the carrier-frequency signal from lead 173 through a buffer amplifier Mixer 1'79 eterodynes the sideband frequencies with the carrier frequency and supplies to an audio amplifier 181 an electric signal having the same frequency components as those in the original audio-frequency signal supplied to lead 171 by musical instrument 17%. Consequently, the apparatus thus far described is similar to a conventional car- 'rier-current communications system.

A similar carrier-current system comprises a balanced modulator 182 that receives an audio-frequency signal from amplifier 181 through lead 1% and receives the carrier-frequency signal from lead 173 through a buffer amplifier 134 and a lead 185. Sidebands corresponding to the sum and difference frequencies are provided at lead 186, and are supplied to a mixer 187 that also receives the carrier-frequency signal from lead 173 through a buffer amplifier 183. Mixer 187 supplies an audio-frequency signal trrough lead 18% to an audio amplifier 1% and a loud-speaker system 191 that produces sound waves corresponding to the audio-frequency electric signal.

Assume, for example, that musical instrument 17% provides an electric signal corresponding to a simple musical tone having a frequency of 490 cycles per second. Oscillator 172 supplies a carrier-frequency signal at 100 kilocycles per second. Modulator 177 heterodynes the audiofrequency and the carrier-frequency signals, and supplies through lead 178 two sidebands having frequencies of 99,668 and 190,400 cycles per second, respectively. Mixer l7? heterodynes the two sidebands with the carrier frequency of 109 he, vand provides a 460 cycle per second audiofrequency signal that is transmitted by audio ampli'lier and lead 133 to balanced modulator r32.

Modulator 182 heterodynes the 46-0 cycle per second audio-frequency signal with the 106 kc. carrier frequency signal, and supplies to lead 18% two sidebands having irequencies of 99,600 and 100,400 cycles per second, respectively. dixer i237 heterodynes these sidebands with the 100 lac. carrier-frequency signal, and supplies through lead 185* and audio-amplifier 19% a 400 cycle per second electric signal that is converted by loudspeaker 191 into a musical tone having a frequency of 466 cycles per second. Consequently, in this mode of operation of the PEG. l0 pparatus, the signal provided by musical instrument 17% 's transmitted to loud-speaker 191 without any modificaicn of tone.

An oscill tor res supplies to lead 192 a second carrierfrequency si nal that differs from the carrienfrequency supplied to lead 173 by a subaudio frequency. For example, if oscillator 172 provides a signal having a frequency of 100 he, oscillator 13 2. may provide a signal having a frequency of 100,603 cycl s per second. This second carrier-frequency signal is transmitted through a buffer amplifier 193 to a normally open contact of switch 176.

Now assume that switch 17o is operated to open its normally closed contact while simultaneously closing its normally open contact. Now the carrier-frequency signal supplied to modulator 177 has a frequency of 160,003 cycles per second, and this signal is heterodyned with the 400 cycle per second audio-frequency signal supplied by musical instrument 17b to produce through lead lid a pair of sidebands having frequencies of 99,693 and 100,493 cycles per second, respectively. These two sidebands are heterodyned with the 100 lrc. carrier-frequency signal by mixer 179, whereupon mixer 11 9 supplies to audio amplifier 181 a signal having audio-frequency components representing the difference frequencies between the two sidebands and the 106' kc. carrier frequency, In other words, the audio-frequency signal supplied to amplifier now has two frequency components, one of which has a frequency of 397 cycles per second while the other has a frequency of 403 cycles per second.

These two audio-frequency components are transmitted by the second carrier system to loud-speaker 191, which thereupon produces a complex musical tone having a 397 cycle per second component and a 463 cycle per sec ond component. The result is a choral effect, similar to that produced by two similar musical instruments playing the same note simultaneously with slightly different frequencies. In other words, the carrier system has modulated the original audio-frequency signal to provide two sidebands each differing from the original audio-frequency signal by a subaudio frequency difference of 3 cycles per second. In thi case the original signal is suppressed, which corresponds to suppressed-carrier amplitude modulation. If the audio-frequency signal supplied by musical instrument 17% has a plurality of frequency components corresponding to a complex musical tone, a pair of sidebands is produced in a similar manner for each of the ori inal frequency components.

A switch 194 is connected in parallel with the normally closed contact of switch 176, as shown. Assume that the normally open contact of switch 1% is closed and that switch 194 is also closed. Now two carrier-frequency signals are supplied simultaneously to balanced modulator 177, one of which has a frequency of 100,000 cycles per second while the other has a frequency of 100,003 cycles per second. When musical instrument 170 supplies a 400 cycle per second audio-frequency signal to balanced modulator 177 through lead 171, four sidebands are provided at lead 178, having frequencies of 99,600; 99,603; 100,- 400; and 100,403 cycles per second, respectively.

Mixer 179 heterodynes these four sidebands with the 100 kc. car-rier frequency signal and supplies to audio amplifier 181 an audio-frequency signal having components equal to each of the three audio-frequency differences; namely 397, 400 and 403 cycles per second. These three audio-frequency components are transmitted by the second carrier system to loud-speaker 191, which thereupon produces a musical tone having the three frequency components. It will be noted that these three frequency components consist of the original frequency, 400 cycles per second, and a sideband on either side thereof in the frequency spectrum and spaced therefrom by a subaudio frequency difference of 3 cycles per second. This corresponds to a simple amplitude-modulated signal, and pro vides a tremolo effect.

Still another oscillator 195' supplies a third carrier-frequency signal that differs from the first and second carrierfrequency signals by subaudio frequency differences. For example, oscillator 195 may provide a signal having a frequency of 99,995 cycles per second. This oscillator i connected to balanced modulator 177 through a buffer amplifier 196 and a normally open switch 197. Now assume that switch 194 is open and that the normally open contacts of switches .176 and 197 are closed. Also assume that the musical instrument 170 supplies a 400 cycle per second audio-frequency signal to balanced modulator 177. Modulator 177 heterodynes the audio-frequency signal with the two carrier-frequency signals that are supplied to it under these conditions, and provides at lead 178 four sidebands having frequencies of 99,595; 99,603; 100,395 and 100,403 cycles per second, respectively. Mixer 179 heterodynes these four sidebands with the 100 kc. carrierfrequency signal and supplies to audio amplifier 181 an audio frequency signal having frequency components of 395, 397, 403, and 405 cycles per second. These four audio frequency components are transmitted to loudspeaker 19 1, which thereupon produces a musical tone having four frequency components separated by a plurality of subaudio frequency differences. The musical effect is that of a thick tone or choral effect.

If switch 194 is also closed, the 100 kc. carrier-frequency is supplied to modulator 177 along with the two carrier frequencies supplied by

oscillators

192 and 195.

In this case the same four audio-frequency sidebands are transmitted to loudspeaker 191, and in addition the original 400 cycle per second audio frequency signal is supplied to the loudspeaker. The resulting signal, comprising the original signal and two pairs of sidebands, is identical to a signal produced by amplitude modulation of the original signal with two different subaudio frequencies simultaneously, so that the musical tone produced by loudspeaker 191 contains a double tremolo similar to that produced by two similar musical instruments playing the same notes simultaneously with different tremolo rates.

Still other oscillators 1'98 and 199 provide two additional .carrienfrequency signals that differ from all the other carrier-frequency signals by subaudio frequency differences. Oscillator 198 is connected to lead 185 through a buffer amplifier 200 and a normally open switch 201. Oscillator 199 is connected to lead 185 through a buffer amplifier 202 and a normally open switch 203. When switch 201 is closed, two different carrier frequencies are simultaneously supplied to the balanced modulator 182,

whereupon the second carrier system provides two sidebands for each frequency component in the audio-frequency signal supplied to modulator 182. through lead 183. When switch 203 is closed a third carrier-frequency signal is supplied to balanced modulator 182, whereupon another is if? pair of sidebands is provided for each frequency component of the audio-frequency signal supplied through lead 183. Consequently, when all of the switches are closed a large number of sidebands are provided for each frequency component of the original audiofrequency signal supplied by the electrical musical instrument 170, and a very thick tone is produced by loud-speaker 191 to p oduce a choral effect similar to many musical instruments of the same type playing the same notes simultaneously.

The side-bands of each frequency component occupy a small frequency spectrum centered on the original frequency of that component and having a width equal to twice the sum of the larger of the two subaudio frequency differences between the frequency of oscillator 172 and the frequency of

oscillators

192 and 195 and the larger of the two subaudio frequency differences between the frequency of oscillator 17?. and the frequencies of oscillators 198 and 199.

Oscillators 1'72, 192, 195, 198, and 199 may be oscillators of any type providing reasonably stable electric signals of appropriate frequency. The five oscillators are tuned to substantially the same frequency, but operate asynchronously and therefore inevitably operate with slight frequency differences. If these inherent frequency differences are not as large as desired, the oscillators may be slightly de-tuned with respect to one another. The principal purpose of the buffer amplifiers is to isolate each oscillator from the others so that interaction between the oscillators will not inadvertently lock them into synchronism. If electron-coupled oscillators or the like are ernployed some or all of the buffer amplifiers may be omitted. Alternatively, passive networks may be used in place of buffer amplifiers to achieve signal isolation.

The amplifiers, modulators, and mixer-s may be of conventional types well known in the communications art. The mixers may be any devices that produce a heterodyning action or the equivalent between two sets of input signals, and may comprise multigrid mixer tubes of the type commonly used in communications work, or they may be demodulators or detractors of any suitable type, many of which are very well known. Since many suitable types of oscillators, amplifiers, modulators and mixers are well known, any further description thereof would be superfluous.

Various modifications of the circuit shown in FIG. 10 are possible without departing from the invention in its broader aspects. For example, a single carrier-frequency may be supplied at all times to the modulators, while a plurality of carrier-frequency signals are supplied to the mixers for producing audio-frequency sidebands. If desired, a plurality of carrier-frequency signals can be supplied to the modulators and the same or a different plurality of carrier-frequency signals may be supplied to the mixers, provided reasonable precautions that will be apparent to competent engineers are observed to prevent the suppression of some difference frequencies by others. If unbalanced modulators are used instead of balanced modulators, the carrier-frequency signals will not be suppressed, and accordingly the carrier frequencies as well as the sidebands will be present in leads 178 and 186. Con sequently, subaudio difference frequencies will also be produced by the mixers, but these are not necessarily objectionable since they can easily be eliminated, if so desired, by appropriate design of the audio amplifiers. In any event, subaudio components will seldom be reproduced to any considerable extent by conventional loud-speakers, and even if they are reproduced, they are inaudible.

Reference is now made to FIG. 11 of the drawing, which illustrates an embodiment wherein multiple carrier frequencies are produced by modulating a single carrierfrequency signal. Electrical musical instrument 204 supplies to lead 205 an audio-frequency signal having the frequency components of a musical tone. An oscillator 206 supplies to lead 207 an electric signal having a carrier frequency, kilocycles per second, for example. The

215 or its circuit equivalent.

, that convert the electric signal into sound waves. apparatus thus far described is a single-sideband carrier audio-frequency and carrier-frequency signals are supplied to a balanced modulator 208 that provides at lead 20? a .pair of carrier-frequency sidebands for each frequency component of the audio-frequency signal. For example, if the audio-frequency signal has a component at 400 cycles per second, a pair of sidebands are provided through lead 209 having frequencies of 99,600 and 100,400 cycles per second, respectively.

A single-sideband filter 210 attenuates one sideband of each pair relative to the other sideband, so that there is provided at lead 211 one carrier-frequency sideband for each frequency component of the audio-frequency signal.

The sideband at 211 may be either the sum or the difference frequency of the audio-frequency and carrierfrequency signals. Assume, for example, that the signal .at lead 211 has a sideband at 100,400 cycles per second for a 400 cycle per second audio-frequency component, a sideband at 100,800 cycles per second for an 800 cycle per second audio-frequency component, and so forth. This part of the apparatus is similar to the transmitting portion of a conventional single-sideband communications system, and any further description thereof would b superfluous.

The same carrier-frequency signal is also supplied to a center-tapped inductor 212. The center tap of inductor 212 is connectedto a lead 213, while its end terminals are respectively connected to a lead 214 and to a ground lead Inductor 212 supplies carrier-frequency signals to leads 213 and 214, the carrier frequency voltage between leads 214 and 215 being substantially twice as large as the carrier-frequency voltage between leads 213 and 215.

A portion of the carrier-frequency signal between leads 213 and 215 is transmitted through a resistor 21 5, a normally closed contact of a two-position switch 217, and a lead 218 to the primary 219 of a transformer having a center tapped secondary 220. Primary 219 is connected between leads 218 and 213. The center tap of secondary 220 is connected to a lead 221, while its end terminals are respectively connected to a lead 222 and to lead 215, as

. shown. Secondary 220 supplies carrier-frequency signals to

leads

221 and 222, the carrier-frequency voltage between leads 222 and 215 being substantially twice as large as the carrier-frequency voltage between leads 221 and 215.

A portion of the carrier-frequency signal between leads 221 and 215 is transmitted through a resistor 223 and a lead 224 to a transformer primary 225 connected between lead 224 and lead 221, as shown. A transformer secondary 226 that is inductively coupled to primary 225 .transmits this signal through a conventional vacuum tube amplifier 227 to a mixer 228 that heterodynes the carrierfrequency signal'with the sidebands transmitted through lead 211. The difference frequencies are supplied through an audio amplifier 229 to one or more loud-speakers 230 The system that transmits audio-frequency signals from electrical musical instrument 204 to loudspeaker 230 without substantial modification, so that loud-speaker 2312 produces musical tones corresponding to the electric signals supplied to lead 205.

Connected in series between lead 214 and ground lead 215 there is a variable phase-shifting circuit comprising a resistor 231, an inductor 232, and a variable capacitor 233. A normally open switch 234 and a resistor 235 are connected in series between lead218 and the circuit junction 23:? of resistor 231 and inductor 232. Switch 234 is also connected to a normally open contact of switch 217, as shown.

Capacitor 233 may be a standard, commercially available type of variable capacitor having one or more stationary plates and one or more movable plates that may be continuously rotated to vary its capacitance cyclically. The movable plates of capacitor 233 are continuously rosupplied through amplifier 229 to loud-speaker 230.

tated at a constant speed, 2 revolutions per second for example, to vary the capacitance cyclically at a subaudio frequency. Suitable means for rotating the movable capacitor plates may comprise a drum 237 rotated at constant speed by any suitable means, such as a driving belt 233 driven by an electric motor 239 and a driving pulley 240. To keep belt 233 taut and in good frictional engagement with drum 237, there is provided an idler pulley 241 rotatively mounted on an arm 242 pivotally supported by a stationary shaft 243. Ann 242 and pulley 241 are urged downward by a spring 244 to keep belt 238 under constant tension.

At or near its mid-capacitance value, capacitor 233 preferably is series-resonant with inductor 232 at the frequency of the carrier signal provided by oscillator 206. As the movable plates of capacitor 233 are rotated to vary its capacitance at a subaudio frequency, the impedance vof resonant circuit 232-233 is varied or modulated in amplitude and phase at subaudio frequencies. Consequently, the voltage across the resonant circuit is likewise modulated in amplitude and phase. The voltage between circuit junction 236 and lead 213 is modulated in phase with a subaudio frequency, but has relatively little amplitude modulation. For present purposes, with one exception hereinafter discussed, there is no material difference between phase modulation and frequency modula- Accordingly, the circuit including capacitor 233 rier frequency with a subaudio frequency.

Now assume that the normally closed contact of switch 217 is opened, while the normally open contact to the same switch is closed. The carrier-frequency signal transmitted through line 213, transformer 219420, line 224 and transformer 225426 to amplifier 227 and mixer 228, is frequency-modulated with a subaudio frequency, and therefore has sidebands located in the frequency spectrum upon each side of the carrier frequency and differing therefrom by subaudio frequency difierences. Accordingly, three components of signal (or more, depending upon the modulation index) at substantially the carrier frequency are transmitted simultaneously to mixer 228, and each of these three components is heterodyned with the sideband frequencies transmitted to the mixer through the .lead 211. Consequently, sidebands are added to the audio-frequency components supplied through audio amplifier 22? to loud-speaker 230. A mathematical analysis ..will show that these sidebands are the same as those that would be produced if the original audio-frequency signal were frequency-modulated with the same subaudio frequency as that used to modulate the capacitance of capacitor 233. Consequently, the musical tone produced by loud-speaker 230 is frequency-modulated with the subaudio frequency, thereby producing a vibrato like efiect.

A better understanding of the modulation action may be had by considering the instantaneous frequencies of the carrier-frequency signal supplied to lead 218 and thus to --mixer 228. When the frequency-modulated carrier is heterodyned with the sidebands transmitted through 211, the difference frequencies cyclically vary in frequency, and consequentl a frequency-modulated audio signal is If in addition to the frequency modulation of the carrier, there is also some amplitude modulation of the carrierfrequency signal supplied to mixer 223, there .will be a corresponding amount of amplitude modulation of the audiosignal supplied to the loud-speaker. In general this is out undesirable, since vibratos produced by conventional musical instruments contain a certain amount of tremolo or amplitude modulation. However, the amplitude modulation of the carrier can be suppressed if desired, by providing in amplifier 227 conventional automatic-volume control or amplitude-limiting means for degenerating or clipping the amplitude modulation of the carrier.

Now assume that switch 217 is in the position shown in the drawing, while switch 234 is closed. Two carrierfrcquency signals are now supplied to mixer 228 simultaneously, one of which is unmodulated while the other is frequency modulated as hereinhefore explained. Con sequently, the audio-frequency signal supplied to the loudspeaker has unmodulated components and frequency modulated components simultaneously present, and a thick musical tone is produced that gives a choral effect similar to that of a plurality of similar musical instruments playing the same notes simultaneously.

Another phase or frequency modulator for the carrierfrequency signal comprises

resistors

245 and 246, inductor 247, variable capacitor 243, and a normally open switch 249, connected as shown. The capacitance of capacitor 248 is modulated with a subaudio frequency, different from the modulation frequency of capacitor 233, by rotating the movable plates of capacitor 248 at a different speed. The movable plates of capacitor 248 may be rotated by a drum 248 having a diameter different from that of drum 237 that is also in frictional engagement with the driving belt 238.

When switches 234 and 249 are both closed, the carrier frequency is phase or frequency-modulated with two different subaudio frequencies simultaneously, thereby providing an additional number of carrier-frequency sidebands each of which produces a beat or difference frequency when hereterodyned with the sidebands transmitted through lead 211. Consequently, when both of the

switches

234 and 249 are closed, the musical tone produced by loud-speaker 230 is thicker than when only one of these switches is closed, and the choral etfect produced represents a larger number of similar musical instruments playing the same notes simultaneously.

Still another phase or frequency modulator for the carrier frequency comprises a variable capacitor 250 and a normally open switch 251. The capacitance of capacitor 25% is varied at still another subaudio frequency by rotation of the movable plates of capacitor 250 by a drum 252 driven by belt 238. When all three of the

switches

234, 249, and 251 are closed, the carrier-frequency signal is phase or frequency modulated with three different subaudio frequencies, thereby producing a correspondingly larger number of sidebands that result in a still thicker musical tone.

Still another phase or frequency modulator comprises a variable phase-shifting circuit including a resistor 253. and inductor 254, and a variable capacitor 255 connected in series between lead 222 and the ground lead 215. A normally open switch 256 and a resistor 257 are connected in series between lead 224 and the circuit junction 258 of resistor 253 and inductor 254. The capacitance of variable capacitor 255 is modulated with a subaudio frequency by rotation of the movable plates of the capacitor through a drum 259 driven by belt 233. As hereinbefore explained, when switches 234, 249, and 251 are all closed, the carrier frequency is modulated with three different subaudio frequencies simultaneously, and lead 218 carries a signal containing the original carrier frequency and a number of sidebands. Each of these frequency components is transmitted by the transformer 219- 226 to the

leads

221 and 222. Consequently, each of these frequency components is itself frequency-modulated by the phase-shifting network including variable capacitor 255 to multiply the number of sidehands present. Accordingly, when switch 256 is closed, the number of carrienfrequency sidebands transmitted to mixer 22% is multiplied to a large number, and the audio-frequency signal transmitted to loudspeaker 239 has a correspondingly large number of sidebands. Further additions to the number of sidebands can be provided by closing a switch 260 associated with still another phase-shifting circuit including a variable capacitor 261, and by closing a switch 262 associated with a phase-shifting circuit including a variable capacitor 263. The movable plates of

capacitors

261 and 263 are rotated at different speeds by

drums

264 and 265 that are in frictional engagement with driving belt 233. The six drums driven by belt 238 all have different diameters, so that the six

capacitoirs

233, 243, 25% 255, 261, and 263 have their capacitances varied cyclically at a different subaudio frequency. By closing successively increasing numbers of the

switches

234, 249, 251, 256, 260, and 262, an increasingly complex spectrum of sidebands can be provided for each frequency component of the audiofrequency signal provided by musical instrument 26-4, thereby producing increasingly thick musical tones that represent increasingly greater numbers of musical instruments playing simultaneously.

In the description so far it has been assumed that frequency modulation and phase modulation are the same. According to modulation theory this is true, except that in frequency modulation the modulation index is a function of the modulating frequency, while in phase modulation it is not. If all six of the modulators in the FIG. 11 apparatus are made identical except for the speed at which the moving plates of the capacitors are rotated, phase modulation of the carrier will result and the audiofrequency signal will likewise be phase modulated. In this case the frequency deviation will be proportional to the modulating frequency. in other words, if phase modulation of a 2 cycle per second modulating frequency produces a frequency deviation of 10 cycles per second, then a 6 cycle per second modulating frequency of the same amplitude will produce a frequency deviation of 30 cycles per second.

For example, assume that the movable plates of capacitor 248 are rotated three times as fast as the movable plates of capacitor 233 to provide modulating frequencies of 6 cycles per second and 2 cycles per second, respectively. If these two modulators are made identical except for their modulating frequencies, their modulation indices will be identical and the carrier-frequency signal supplied to lead 218 will be phase modulated with two subaudio frequencies when both of the

switches

234 and 249 are closed, and the frequency spectrum occupied by sidebands produced by the 6 cycle per second modulation will be three times as wide as that occupied by the sidebands produced by the 2 cycle per second modulation.

To produce frequency modulation rather than phase modulation, is only necessary to make the modulation indices of the different modulators inversely proportional to the modulating frequencies. This can easily be done by an appropriate design of the modulators. For example, if the resistance of resistor 245 is made three times as large as that of resistor 231, the modulation indices of the two modulators will be substantially inversely proportional to the modulation frequencies, and the frequency spectrum occupied by sidebands produced by the 6 cycle per second modulation will be substantially the same Width as the frequency spectrum occupied by sidebands produced by the 2 cycle per second modulation. In this case the carrier frequency is frequencymodulated with the two subaudio frequencies when both of the switches 2-34 and 249 are closed. Generally, frequency modulation is preferred to phase modulation, but since some mixture of phase as well as amplitude modulation with the frequency modulation is not generally objectionable in a musical tone, the modulation indices need not be as carefully adjusted as would be required in a high-quality frequency-modulation communications system.

The apparatus shown in FIG. 11 can be modified in various ways without departing from the broader inven tive principles involved. For example, multiple carriers can be produced from a single carrier by amplitude modulation of the carrier in place of frequency or phase modulation. Instead of supplying an unmodulated carrier to the modulator 2G8 and supplying a modulated carrier to mixer 22 3, the modulated carrier may be supplied to modulator While an unmodulated carrier is supplied to mixer 228. Many different types of oscillators, modu- 23 lators, filters, amplifiers, and mixers may be employed, there being many suitable types that are Well known in the communications art.

The several forms of tone modifiers here described are compatible and may, if desired, be used together. For example, the pitch modulator illustrated in FIGS. 1 through 6 may be connected in tandem with the amplitude modulator illustrated in FIG. 7 and the timbre modulator illustrated in PEG. 9. 'With such arrangement a great variety of musical effects can be created at will by the musician. Other tandem and parallel combinations of various modulators can be employed, including combinations of the carrier types with the direct modulation types.

Instead of using discrete subaudio modulating frequencies, spectra of subaudio modulating frequencies may be employed. For example, if a random noise signal is transmitted through a low-pass filter having a cut-oil frequency of approximately seven cycles per second, an infinite number of subaudio frequencies all within the range of zero to seven cycles per second is obtained, and the original audio-frequency signal maybe modulated, either directly or indirectly, by this spectrum of subaudio frequencies to produce a continuous narrow spectrum of sidebands about each audio-frequency component. Such modulation by subaudio noise is equivalent to modulation by an infinite number of discrete subaudio frequencies.

It should be understood that this invention in its broader aspects is not limited to specific embodiments herein illustrated and described, and that the following claims are intended to cover all changes and modifications that do not depart from the true spirit and scope of the invention.

What is claimed is: V

1. Apparatus for providing an audible signal comprising,

sources of at least three periodic signals separated by at least three different subaudio frequency differences, at least one of said signals having a frequency in the audible range,

and means for modulating said one periodic signal with n I the remaining ones of said signals to provide an output signal having adjacent spectral components separated by subaudio frequencies in a band embracing said audible range frequency and separated by at least said three different subaudio frequency differences to provide a choral tone signal characterized by an aurally untrackable modulation pattern.

2. Apparatus in accordance with claim 1 and further comprising,

transducing means for converting an electrical signal to an acoustical signal,

and means for coupling said modulating means output signal to said transducing means to provide an audible choral tone. 7 3. Apparatus for providing an audible signal comprising,

a source of musical tone signal, sources of periodic modulating signals establishing at least three different subaudio modulating frequencies, modulating means, means for coupling said musical tone signal source and said sources of said modulating signals to said modulating means, said modulating means being responsive to said modulating signals for modulating said musical tone signal to provide a choral tone audio output signal having a spectrum embracing the frequency of said musical tone signal with sidebands separated by said subaudio frequencies, said output signal having a complex modulation pattern with a period which is long compared to that of a musical tone.

4. Apparatus in accordance with claim 3 wherein said modulating means comprises,

a magnetic recording medium, a

means for recording a signal related to said musical tone signal upon said recording medium,

means responsive to said recorded signal on said magnetic recording medium for providing said output signal,

and means for establishing relative cyclical movement at at least said three different subaudio frequencies between said means recording said signal on said medium and said means responsive to said recorded signal.

5. Apparatus in accordance with claim 4 wherein said means for recording and said means for providing an output signal comprise a plurality of electromagnetic transducing heads,

means for establishing relative movement between said heads and said recording medium to exchange signals therebetween,

and means for establishing relative cyclical motion between at least one of said recording heads and at least one of said reproducing heads at at least said three different subaudio frequencies.

6. Apparatus in accordance with claim 3 and further comprising,

ransducing means responsive to a final output signal for providing a corresponding acoustical signal simulating a choral tone,

and a plurality of said modulating means jointly energized by said musical tone signal and each providing a said output signal,

means for combining the last-mentioned output signals to provide said final output signal,

and means for coupling said final output signal to said transducin g means.

7. Apparatus in accordance with claim 6 and further comprising,

means for selectively controlling the number of said dii erent suoaudio modulating frequencies.

8. Apparatus in accordance with claim 3 wherein said modulating means comprises,

at least three parallel variable attenuators energized by said musical tone signal,

' and means for cyclically varying the attenuation imparted to said musical tone signal by each attenuating means at a respective one of said subaudio frequencies.

9. Apparatus in accordance with claim 8 wherein each of said attenuating means comprises a variable reactance.

1G. Apparatus in accordance with claim 8 wherein each of said attenuators comprises a variable inductance.

11. Apparatus in accordance with claim 3 wherein said sources of modulating signals comprises sources of at least three superaudio signals separated from one another by said subaudio frequencies.

12. Apparatus in accordance with claim ll wherein said modulating means comprises a first modulator energized by said musical tone signal and at least three of said superaudio frequency signals to provide a signal having spectral components in the superaudio frequency range,

an a mixer energized by the latter signal and one of said superaudio frequency signals to provide said output signal.

13. Apparatus in accordance with claim l2 and further comprising,

a source of at least two superaudio f equency signals separated by a subaudio dif erence frequency,

a second modulator energized by said last-mentioned output signal and at least the latter two superaudio frequency signals to provide an intermediate output signal having spectral components in the superaudio frequ ncy range separated by all said subaudio'frequency differences,

Claims

1. APPARATUS FOR PROVIDING AN AUDIBLE SIGNAL COMPRISING, SOURCES OF AT LEAST THREE PERIODIC SIGNALS SEPARATED BY AT LEAST THREE DIFFERENT SUBAUDIO FREQUENCY DIFFERENCES, AT LEAST ONE OF SAID SIGNALS HAVING A FREQUENCY IN THE AUDIBLE RANGE, AND MEANS FOR MODULATING SAID ONE PERIODIC SIGNAL WITH THE REMAINING ONES OF SAID SIGNALS TO PROVIDE AN OUTPUT SIGNAL HAVING ADJACENT SPECTRAL COMPONENTS SEPARATED BY SUBAUDIO FREQUENCIES IN A BAND EMBRACING SAID AUDIBLE RANGE FREQUENCY AND SEPARATED BY A LEAST SAID THREE DIFFERENT SUBAUDIO FREQUENCY DIFFERENCES TO PROVIDE A CHORAL TONE SIGNAL CHARACTERIZED BY AN AURALLY UNTRACKABLE MODULATION PATTERN.