US2439392A - Generation of tones in photoelectric musical instruments - Google Patents

Description

E. M. JONES Ani-ril 13, 194s.

GENERATION OF TONES IN PHOTOELECTRIC MUSICAL INSTRUMENTS Filed Aug. 5, 1946 4 Sheets-Sheet 1 Q JUAN.

INVENTOR. BY fan/#fa /W JONES.

AT TO RN EYS.

A Plil 13, 1948 E. M. JONES 2,439,392

GENERATION OF TNES IN PHOTOELECTRIC MUSICAL INSTRUMENTS Filed Aug. 5, 1946 4 Sheets-Sheet 2 o o oooo N V ENTOR. bn/449D /Z JQNES.

Bylew M ATTORNEYS.

E. M. JONES 2,439,392

GENERATION OF TONES IN PHOTOELECTRIC MUSICAL INSTRUMENTS April 13, 1948.

Filed Aug. 5, 1946 4 Sheets-Sheet IN VEN TGR. EnBW/S70 M rfa/ves.

ATTURN EYS.

April 13, 1948. E. M. JONES 2,439,392

GENERATION OF TONES IN PHOTOELECTRIC MUSICAL INSTRUMENTS Filed Aug. 5, 1946 4 Sheets-Sheet 4 M" Humm INVENTOR. En n/Aa M rfa/VES.

lex/MVM ATTORNEYS.

Patented Apr. 13, 194% GENERATION F TONES IN PHOTOELEC- TRIO MUSICAL INSTRUMENTS Edward M. Jones, Cincinnati, Ohio, assigner to The Baldwin Company, Cincinnati, Ohio, a

corporation of Ohio Application August 5, 1946, Serial No. 688,576

15 Claims. (Cl. Z50-41.5)

My invention relates to electrical musical instruments in which tones may be selectively generated and reproduced in accordance with the requirements of a musical score; and in particular it relates to that type of musical instrument wherein one or more beams of light are interrupted or modified in a periodic fashion and then irnpinged upon a photocell whereby the pulsations in the beam or beams of light are converted into electrical pulsations which may be amplied and reproduced.

Electrical musical instruments of this general class have hitherto been suggested. In one such instrument a succession of light beams is produced by causing light from a suitable source to be interrupted by alternate opaque and transparent portions of a rotating disk. The light beams are caused successively to traverse a portion of a photoelectric cellfthe arrangement being such that as soon as one beam leaves the traversed area, another begins its traverse. The light beams are caused also to traverse a wave pattern screen of variable translucency on their way to the photoelectric cell so that the derived electrical pulsations will be characterized by a fundamental frequency and various harmonic frequencies in desired proportions to produce a selected voice or timbre. By changing the wave pattern screens, various voices may be obtained.

i principal Object of this invention is the provision of means for producing tones having valuable characteristics beyond a bare content of fundamental and selected harmonics in selected proportions. I-litherto, all such frequencies have been constant during the continuous and constant-volume sounding of any given tone. The bare characteristics of constant funeamental and harmonic frequencies just set forth, will provide tones having voices or timbres which may be made imitative of the voices of Known musical instruments, or may constitute new voices. But such tones are relatively cold and mechanical in character and lack something oi the effect of the tones of an orchestral rendition, or even of any one instrument played by human agency.

When musical instruments are playing to gether in unison or in octaves, the sound can be accurately represented by a series of wave forms in a graph or chart. Successive wave forms will be nearly the same; but they will not be identical. The change usually occurs at a rate which is comparatively slow compared with the audio rate of the vibrations themselves.

For example, when a group of violins are played in unison, not only are there slight differences in the speciiic frequencies of the fundamentals and harmonics produced by each, but each violin may have its individual vibrato.

It is an object of my invention to provide means and methods whereby photoelectrically generated wave forms may be given similar characteristics. it is an object of my invention to provide means and methods whereby the generated wave forms may be caused to pass through a sequence which repeats continuously every few seconds or a sequence which is i 'ated when desired, as when a key is deor both.

heee and other objects of my invention wtich will be set forth hereinafter or Will be apparent to one skilled in the art upon reading these specifications, l accomplish by that construction and arrangement of parts and in those procedures of which I shall now describe eX- emplary embodiments.

Reference is made to the accompanying draw-- ings wherein:

Figures l to 6 inclusive are representations of wave forms to which I shall refer to illustrate results of the practice of my invention.

' Figure '.7 is a diagrammatic representation of elements of a photoelectric instrument to which my invention may be applied.

Figure 8 is a diagrammatic representation of an interruptor a wave form screen.

Figures 9 and lo and il represent various embodiments of wave form screens.

Figure l2 is a diagrammatic representation of yet another form oi interrupter.

`figures i3, le, l5, i6 and 17 are diagrammatic representations of still ther forms of interrupter and screen mechanisms, Figure 15a showing a were form screen for percussive eiiects.

In Figure 7, I have shown in a purely diagrammatic fashion the operating elements of a photoelectric device such as those to which my invention, as later described, may be applied. Light from a source l is broken into a series of traveling beams by slots 2 in a disc 3 which rotates at a fixed speed. The beams, condensed as may be required by a lens system ll, traverse a photocell A shutter, having a fixed element f5 with a window, and a movable element 1, connected by some means 3 to a playing key 9, serves the dual function of permitting the beams to pass when the particular note is desired, and of conning the beams in such fashion that as soon as one beam completes its traverse another begins to traverse the photocell.

The light source I may be controlled by a stop l@ through a switch H. The photocell will be connected through an amplifier I2 to a loudspeaker IS for the production of sound.

On their way to the photocell, the beams of light pass through a wave-form screen ifi with portions of varying opacity, or transparent portions of varying area, the purpose of which is to cause the beams to change in intensity, during their traverse, in accordance with a predetermined Wave form or pattern. In this Way, the electrical pulsations produced in the phctoceil may be made to have a. Voice comprising desired harmonics in desired intensities, in addition to the fundamental. Since, as pointed out, the beams in continuous sequence traverse the photocell, it will be evident that the wave pattern is constant, i. e., repeated over and over again at a constant frequency, and can be represented on a chart or graph.

Figure '7 is, as indicated, diagrammatic. A .practical instrument would involve one or more of the interrupter discs, each having many circles of slots, aggregating one circle for each of the notes within the range of the instrument. Further multiplication of parts may be made if the instrument has more than one manual. There may be a separate wave form screen for each desired tone quality at each pitch, although it will be evident that wave form screens for a plurality of notes and voices may be mounted on thesame supporting means. There may be a separate light (and photocell means) if desired, for each tone quality, or means may be provided to shift different wave forms screens into position as required. The wave form Screens may be, for example, variable density patterns on a photographic plate, and the sound may be made to build up and decay in accordance with movements of the shutter by special shutter construction.

The specific construction of the instrument in which the invention I am about to describe is embodied, is not limitative upon the invention itself, and may be widely varied. The principles of my invention are applicable generally to devices in which beams of light are caused to impinge upon a photocell in succession at a desired fundamental frequency, and during such im pingement are modified in such fashion as to provide harmonics. In the exemplary embodiment a point source of light is desirable so that sharp shadow images of the slots can be cast on the wave form screens. The ability of such a system to reproduce the higher harmonics of a tone depends upon the narrowness of the beams of light passing across the Wave form, which in turn depends on the width of the slots, the size of the light source, and the distance between the slots and the wave form. If the light source is very concentrated, diffraction effects may pre'domin nate, in which event better results may be cbtained if the spacing between the slots and the wave form 'is increased. If the light source cannot be made concentrated enough, images of the i disc slots can be projected onto the wave form and 300 cycles respectively, the last two of which are harmonics of the first. This may be represented graphically as in Figure 1 as a musical tone i5 of constant timbre. Such a tone would not normally be produced excepting by purely mechanical or mechanico-electrical means. When an instrument is played by human agency there are usually slight variations in timbre which impart variety and interest. Similarly, when a group of violins play in unison, for eX- ample, they are not all exactly in tune, and each has its own pitch-tremolo. Moreover, most instruments have percussive effects having to do with the onset of the tones, variations in volume after the onset, and, frequently, variations of timbre.l

In the practice of invention I am able to .produce tones of varying timbres which can be accurately represented as a series ci wave forms, successive ones being nearly the saine, but exhibiting a change occurring at a slow rate compared with the frequency of the tones. I can cause my Wave forms to pass through sequences which may be classified into various types. In type 1, the timbre variation repeats every few seconds. In type 2, the tone dies away within a few seconds, with or without timbre change. Type 3 is essentially a combination; but it may be characterized by an initial sequence of wave forms followed by repetitive sequences,

Tones of the first type may :be made to imitate a group of instruments playing in unison. Tones of the second type are essentially percussive in nature. Tones of the third type could imitate for example, combinations of organ pipes, each having its own delay, manner of attack, initial timbre, and pitch, as well as steady timbre pitch.

In Figure 2 Ihave illustrated graphically a simple form of musical tone l of varying timbre of type 1 above. In this particular case, the waveform sequence repeats twice a second, so a curve showing what happens in a half-second interval is suiiicient to dene the tone. Around :0 and t=.5 second, the Wave is like the one shown in Fig. 1, but in the middle (at t=.25) the Wave is entirely different. Notice, however, that the curves for successive lAOO sec. intervals do not dier much, which means that one can call the sound essentially a 10D cycle sound. In the equation for the curve, one will notice frequencies of 100, 102, 196, 200, 204, 296, 298, 300, 302, 30a cycles per second. (The Fourier theorem requires that the frequencies be spaced 2 cycles apart because of the fact that the wave repeats twice a second.) These frequencies may be put into 3 groups, those close to 100 cycles, those close to 200 cycles, and those close to 300 cycles.

These three groups of frequencies are plotted separately in Figs. 3, 4 and 5 at i'i, i8 and ld. Notice that the curves in these figures look like sine waves. The slow' changes in phase and amplitude of these sine waves during the half-second interval may be considered to be the results of beats between the various steady sine Wave com ponents. Another way of looking at these varying sine waves is that each cycle of the wave in Fig. 2 Vmay be analyzed by Fourier analysis into lst, 2nd and 3rd harmonics, and since the waveform is slowly changing, the magnitude and phase angle of each harmonic will change from cycle to cycle.

Three basic vphenomena that occur when steady sine waves of nearly the same frequency com bine are illustrated by the curves of the three harmonics. The curve i7 in Fig. 3 is a typical result of combining two sine waves of slightly different frequencies. At the ends of the curve,

the magnitude of the resulting sine wave is largest because the two components are in phase or add. In the middle of the curve, there is minimum amplitude because the two components are out of phase, or subtract from one another. Notice that at t=.2 and t=.3 second, the peaks of the waves do not occur exactly at the same places relative to the 1/100 sec. marks, which means the phase angle of the sine Wave is changing. This little fact prevents the wave from being considered merely as a 100 cycle wave the magnitude of which is changing.

The curve I8 in Fig. 4 is made up of three sine wave components; but it happens that in this case the principal eifect lies in phase changes, instead of magnitude changes in the resulting sine waves. It takes a close examination of the curve to detect the phase changes and to see that the peaks doshift back and forth relative to the to@ second marks. The ear, however, will easily detect the vibrato, or shift in pitch up and down 4 times a second (which is equivalent to the phase varying 4 times a second).

Figure 5 shows in curve i9 the third harmonics which happen to be undergoing changes in ma.,n nitude, but are constant in phase. This is because the component frequencies are equally distributed above and below 300 cycles. Those familiar with amplitude modulation in radio will recognize that the wave is equivalent to a 300 cycle carrier modulated with frequencies of 2 and e cycles, producing side bands of 296, 298, 302, and 304 cycles.

To sum up it may be said that a type l tone as defined above consists of a fundamental and various harmonic sine waves, each of which is periodically undergoing separate changes in magnitude and frequency or phase.

Fig. 6 shows a type 2 or percussion type tone in curve Eil. Unlike the curves in the foregoing figures, it does not repeat after t=.5 second, but instead, the 100 cycle wave that is still present at the end of the curve dies out in another second or so. The equation for the curve shown barely scratches the surface of the many possibilities for tones, but it does show some principal characteristics of percussion sounds. Note that the 100 cycle component takes a few cycles to build up to its maximum value, and then dies out more slowly than the other harmonics. The second and third harmonics predominate at first, as seen by the initially complicated wave-form. Though the second and third harmonics decay at the same rate, and neglecting the fundamental, the waveform would still be changing because the third harmonic, having a frequency of 304.3 cycles, is not a true harmonic. Notice that type 2 tones can contain decaying sine waves of any decimal frequency as long as it is not too different from a true harmonic.

Taking up first my methods and means for securing the type 1 variation, which is the simplest form of variation here involved, I have illustrated in Fig. 8, a rotating disc 2i which, instead of having concentric circles of slots for each pitch is provided with a series of perforations 22 arranged in a spiral configuration. Each of these perforations therefore will scan a slightly different part of an elongated wave form screen 23. It will be evident that the perforations move transversely acrossthe screen, which not only is of variable density transversely to impart the desired harmonies, but also changes progressively in the opposite or longitudinal direction.

The waveform 24 shown in Fig. 9 is an actual representation of the variable density pattern necessary to produce the tone in Fig. 2 (assuming the shaft 25, Fig. 8, runs 2 revolutions per second) The time at which each portion of the wave form is used is shown, and it should be noticed that the waveform extends above and below the portion actually used (between 12:0 and 15:.5 sec.) The extended portions are repeated parts; but they insure that the tone quality will not be affected appreciably if the waveform is slightly out of position with respect to the disk.

Fig. l0 shows a waveform 2li for a single frequency, a 406 cycle sine wave. It can be seen that there would exist a Waveform for any frequency which is a multiple of the frequency of rotation of the disk (which in this case is 2 cycles per second). To reproduce the waveform efiiciently, the holes in the disk must be smaller than the distance between the slanting ridges on the pattern. This means, roughly, that one cycle of the harmonic (the 4th in the above case) plotted in the horizontal direction, must be larger than the hole; and the distance traversed in the vertical direction during one (hypothetical) beat between the frequency in question and the pure harmonic (400 cycles) must be larger than the holes. 1n the example given, these conditions are met, because the holes are smaller than A of the horizontal length of the waveform and smaller than 1A; of the vertical depth of the waveform. It can be seen then that the system is incapable of reproducing well frequencies above certain value and differing more than a certain amount from the true harmonics (multiples of cycles in this case). These limitations may be considered as an advantage, since it means that imperfect waveforms do not produce much dissonance.

Any waveform used in the spiral scan method of Fig. 8 may be considered to be a combination of many waveforms like the one shown in Fig. ld. In making waveforms, it is preferred to synthesize them out of single frequency waveforms, after having emphasized those frequencies that reproduce poorly because of the size of the holes on the disk.

Fig. 11 shows a Waveform Z'l that produces the vibrato tone, the waveform of which is shown in Fig. Ll at i3. This tone is one of the components of the tone of Fig. 2, and a visual comparison between the waveform 2s of Fig. 9 and the one in Fig. 11 shows some resemblance.

In the use of complex waveform screens such as those illustrated at 24 in Fig. 9 and 27 in Fig. 10, the holes in the disc do not need to be limited in diameter to the depth of the representation of any particular cycle on the screen. On the contrary, if some lapping occurs, a smoother variation is produced.

While as indicated above, the principles of this invention may be incorporated in instruments of Widely varying design and characteristics, a consideration oi certain features of an exemplary embodiment contribute to an understanding of the principles.

A practical instrument based on the principle of Fig. 8 may be made as follows: A photo graphic disk rotating 2/3 revolution per second has '73 concentric spirals of holes .002 in diameter. (This means the waveform sequences would repeat once every 11/2 seconds.) The spirals may be spaced .070 apart and the total variation of radius of each spiral would be .050. The inner spiral would have an average radius of 1 with 49 holes for the lowest C in a 16 foot tone. The outer spiral would have an average radius of 6.04" with 314() holes to produce the highest C for 8 or l foot stops. Each waveform, instead of being just 1 cycle long as Fig. 9, would contain from 3 to 30 cycles, which means that an imperfection in one of the holes 'in the disk would cause much less disturbance. The waveforms may be arranged in two radial rows, with the shutter apertures alternating from one side to the other, so that each waveform 'could have awidth of .12 of which .05 was used. The waveform for one stop may he photographed on a strip of film 1 X 6 and despite dimensional instability of the lm, the spirals would stili pass over the waveforms.

Y The shutters are preferably arranged to let light onto the whole area traversed by the spiral gradually as the lrey is depressed, in order to produce a smooth rise and decay of the tone, under the control of the performer.

Arrangements to have combinations of two or more waveforms acting when one key is pressed, of course, add to the versatility oi the instrument and malte it like an organ; but there is also the possibility of producing binaural effects by putting the light from two diilerent wave forms into separate photocells, amplifiers, and speakers, which could give the effect of instruments or pipes located in diilerent places pla'ing in unison (or octavos). The application of this method ci producing binaural effects on type 2 and type 3 sounds discussed below might be still more eiecn tive.

In Figure 12 I have shown a waveform screen 28 in combination with two slotted discs. The disc 29 is provided with a series of slots which, co-acting with the slots Si of the disc 32 give a scanning effect when these discs are rotated in proper speed ratios on their respective shafts and 3s.

The method and means of Fig. l2 produces type l tones, and has the advantage that the pitoLv disk 32 need not run as slowly as the one discussed above to prevent too frequent repetition of the waveform sequence. The same type of wavei'orm is used. If the shaft 33 of the auxiliary disk Qt runs at such a speed to cause one slot 3G to pass the stationary waveform every half second, and the shaft 3d of the pitch disk rotates every half* second, then the wave of Fig, 2 will be reproduced. However, there is no reason why the auxiliary1 disk could not run much slower than this to produce a waveform sequence repeating say every 2 seconds, while the pitch disk Si could run fi times a second, which would give good .high 'frequency response. Furthermore any deviations from the tempered scale resulting from having' to choose an integral number of slots on the pitch disk which might be plus or minus 2 cycles, could be corrected waveform to an accuracy of plus or minus l/a cycle.

In a practical instrument, it might be best to have the scale split up onto two pitch disks, the ones with the high pitches running much raster. There are many ways of arranging the auxiliary disk, and the same circle of slots on the auxiliary disk may be used for many circles on the pitch disk. is no necessity that the disks move across the wave 'forms exactly at right angles.

While I have illustrated a disk it@ as having the slots it will be onderste that there slots may ed be .located on another membr, g. a band, runcrradially of disc 32.

' In Fig. 13, I have shown a pitch disc 35 having a series of slots 35 for a particular frequency.

Abovevit, and mounted rotatably on the same shaft 3l for VAon a sleeve lon that shaft, I have provided another disc 3B, characterized by radial slots '35, vhaving a width which is the same as the circumferential distance between the radial slots 3S in the disc 35. The tone screen 'it carries a sequential representation All of a series of wave forms Aof differing characteristics. Each represented cycle has a length approximately equal to the width of a slot 39 at the same radial position, and the total length of the representation lll is at least equal to the circumferential distance rb'etween slots 3Q at the same radial position,

Fig. 13 shows a method and means by which type l tones can be pro-duced without using a two` dirnensionalwaveform. In this case the pitch disc 35 and the auxiliary disc S3 run concentrically as indicated, and the variable area plot il is made of a Asequence of ten wave forms. In this particular case, whereit was desired to reproduce the tone of Fig. 2, every fth cycle of the curve of Fig. -2 is plotted in succession, and the pitch disc runs twice a second, while the auxiliary disc t8 turns once every 21/2 seconds. This means that during a half second interval, a slot 3g on the auxiliary disc @e passes across the succession of vaveiorms, and since this slot is one cycle wide, there is always one narrow slot exposed on the pitch disc 35. This slot is passing across the waveform at the correct speed; and the wave form is changing in approximately the correct way.

The speeds of the auxiliary and pitch discs can be made to obtain the same advantages as to accuracy of pitch and high frequency response as in the two disc method of Fig. 12, but here it is still more desirable to split the lscale between two pitch discs going at different speeds. Otherwise, the low notes would require waveform-s clear around a circle. The method can be used with variable density waveforms, and by operating the shutter in radial direction, and varying the Idensity pattern in the radial direction, the tone quality can be made to change in any desired manner as the key on the Akeyboard is .slowly depressed. The waveform is therefore made to be essentially a three dimensional pattern.

The successive waveform method of Fig. 13 has some ySerious disadvantages, however. In the rst place, `a large area has to be illuminated very evenly, and 'the dimensional and density stability of the waveforms has to be good to prevent a poor transition as a new slot 39 enters the Waven form. In the second place, the method inherently distorts instead of simply attenuating frequencies that diifer apprcciably from the pure harmonic frequencies. Distortion occurs every time a transition is made from one slot to another on the pitch disc 35,' and the amount of distortion 4depends upon the jump between two successive waveforms. By makingblurred variable density instead of sharp edges on the slots 39 of the auxiliary disc, so that a blurred shadow is cast, the transition from one slot to another on the pitch disc may be made very smooth, and distortion will be much reduced.

Certain of these disadvantages may be obviated by eliminating the disc i3 and placing the Waveform representations il concentrically on a disc which is concentric with the disc 35. The waveform representations may vary progressively or repetitively throughout an entire circle. In this form, the waveform screen rotates; but the window of the shutter member i3 or its equivalent is -used to connue the beams to the proper scanning area. It will be evident that if the full circumference of a disc bearing the wave patterns is used in this fashion, a different pitch disc and appurtenances must be used for each desired voice.

in Fig. le, I have shown a pitch disc l2 rotating on a shaft i3 and having a series of perforations il arranged circularly rather than spirally. The waveform screen or screens l5 may be arranged on a suitably guided member which is moved back and forth radially of the disc l2 by a cam di.

The back forth method of Fig. lll can re produce only a restricted class of type 1 tones; but it has merits of simplicity and reliability that make it worth while. The system uses variable density waveform screen similar to the kind used in the spiral scan method. of Fig. 3. The waveform moves back and forth, maintaining a constant speed during most of each half cycle by means Iof the cam. Since the motion of the waveform screen is back and forth, the sequence of waveforms occurring during half a revolution of the cam must be repeated backwards during the next half revolution.

Mathematically this appears as follows: Let T be the total time for one revolution of the cam and let ,fo be the frequency with which the holes i4 pass across the waveform. The resulting wave will then be composed of continuous sine waves of various frequencies Moin/T, where n and k are integers. The restriction is that for every component of frequency kfn-l-n/T there must exist a complementary component of frequency ftu-n/ T whose amplitude is the same, and which starts out in the same phase when the waveform screen is in one of the extreme positions. This ,restriction eliminates the possibility of producing tones that are better in tune than the frequency of the holes passing across the waveform; and inharmonic components or" tones cannot be produced unless the pitches are distributed equally above and below this frequency. Vibratos and combinations of large numbers of instruments, however, can be imitated successfully.

En an actual instrument the various circles of holes can be spaced rather closely on the pitch disc, and by staggering the shutters into several rows, the waveform screens can still be made fairly large (radially). The waveforms for all the pitches can be moved back and forth by a single cam.

There are Various ways in which percussive tones may be obtained in musical instruments. It has hitherto been suggested that the keys of an instrument be coupled to a control mechanism functioning for example to cause the volume of the reproduced tone to die away, or to cause the volume to follow a desired definite curve. In a photoelectrie instrument, such elfects might be obtained by means controlling the movement of the shutter mechanism, or the brightness of the light source. The application of such means to my instrument has the advantage that percussive effects may be obtained in tones of type l timbre variation, and devoid of the monotony usually characteristic of mechanically or electrically generated tones. However, in most percussive or chestral instruments, a. variation of timbre accompanies percussive variations in volume in a definite relationship. in my instrument this may be accomplished by coupling the wave form screens or the scanning devices or both to the keys or shutters so that a definite waveform sequence which starts from the beginning whenever a key is pressed may be obtained. Percussive variations of volume may be built into the waveform screen or may be attained otherwise, as by controlling the movements of the shutter, the intensity of the light source, and the like.

Figs. l5 and 16 show a method and means for producing type 2 tones. When the key on the keyboard is pressed, a shutter diagrammatically indicated at i8 (Fig. 16) opens. A slot 4S in a member 5d, starts moving across the wave form 5l at a slow rate (from left to right). The slots 52 on the disc de are continually moving across the wave form and they scan that part of the waveform exposed by the slot rihe right hand sido of the waveform is so dense that by the time the slot ld has reached the right hand position, the light variations and the resulting tone have died out completely. When the key on the keyboard is lifted, the shutter closes and the slot 49 returns quickly to the left hand side of the wave form screen, whether o-r not its traverse has been completed.

The waveform 5i (Figure 15a) is representative of the waveform necessary to produce the sound shown in Figure 6. The most remarkable feature of this particular wave form is the appearance of some faint ridges slanting down toward the right, which is the effect of the inharmonic frequency of 304.3 cycles. It will be observed that this wave form is made darker on the right-hand side. In the previous wave forms, the transparency of the waveform at any point was made proportional to a constant plus the instantaneous value found at the corresponding point on the graph of the sound wave. For the percussion case, however, instead of adding a constant to the curve, the exponential function le() eSt was added. Theoretically this hardly affects the tone, but practically it is important to bias the wave form to be very dark at the end, so that imperfections in the wave form will not produce a stray tone at the end.

A mechanical method of making the slot pass slowly across the waveform is shown diagrammatically in Figure 16. When the key connected to the transmission means E4 is pressed, it opens the shutte' 58 and puts tension on a spring 55, which draws upwardly a lever 56, pivyoted at one end 5l, and carrying a roller 53 at the other end. This roller presses a strip 59, which is connected to the member 5d, against a shaft 63 which is rotating-at a slow, constant rate. The strip Ell moves slowly to the right, carrying with it the member 5o which has the slot 49. The strip 59, in moving, stretches a spring El, until a stop 52 is hit, by an abutment 53 on the strip. When the key is released, the shutter it closes and the roller Se drops, no longer pressing against the strip 59. The spring 6I quickly pulls the strip t9 and the slotted member t back to the original position, as determined by a stop E54. Stop means 65, G and 6'! are provided for the shutter.

TheV shutter t3 is arranged to open last and close first; and if the key is pressed down slowly, the loudest part of the waveform may be entirely by-passed by the time the shutter opens. This is the eifect desired to imitate a piano, where the loudness depends on the speed of pressing the key. The effect could be improved by having the strip 59 initially make contact with a faster moving shaft.

The extreme versatility of this method for producing percussion tones can be appreciated when it is considered that any percussion tone can be recorded by pressing the key and modulating the light beam, as in the methods for recording on movie film. The tone will reproduce satisfacaeaaaaa torily providing that the frequency of the percussion tone recorded was approximately synchronized to the frequency of the slots on the disc passing across the Wave form, providing the harmonics were not too high or inharmonic, providing the tone died away soon enough, and providing the wave form or loudness did not change toorapidly.

With different Wave form screens, different percussive effectsv may be obtained; and I am not limited to the particular one described in which the timbre and volume vary as shown in Fig. 6.

Various combinations and modifications may be made within the perview of my teachings. Onesuch variation lies in the production of the type 3V sequence mentioned above. In this sequence, the tones, when initially struck have a different timbre and (if desired) volume from that which they attain after a brief interval; but when the initial tonal effect is achieved, the note continues With a periodic or cyclical variation of timbre.` These eifects may be attained by pro viding a waveform screen with the initial effect atA one end merging into the repetitive effect in the remainder of the screen, and by scanning the whole screen first and then the said remainder repetitively.

In Fig. 17 I have shown a pitch disc 'it having slots 69'. A wave form screen 'lli is providedof the type just described. A member 'il is operated in the same manner as the member 59v of Figs. and 16, as by a rotating shaft i2, but is'shown as moving transverse to the radius of the disc 68, and as provided with a series of diagonal scanning slots. The' first of these slots, .marked 7.3. is a slotof such length as to'be capable cf scanning they entire depth of the wave form screen 1G. The Wave form markings on the screen will, of course, be tilted to match the scanning direction which isy substantially parallel to the slots assuming that these move slowly with respect to the pitch disc. Additional slots 'lil' are provided of lesser length and operating, as Will now be clear, to confine the scanning to the lower portion only of the Wave form screen. The slots 'I3 and 14 are so related as to their ends that as soon as one slot moves away from the Window of the scanning screen. (or such other window as is provided), the next enters it.

Tones produced in this fashion may have an initial.V timbre and/or volume followed by a different timbre and/or volume, the latter .being repetitive.

In: such a system, the playing of the note may be prolonged for as long a time as itis feasible to keep the strip member 1| moving. Various mechanical embodiments of the principle may bel made; but it is readily possible to provide in this way for type. 3" effects with tones sustained as long as may be required for the usual musical composition.

It will be understood that the various teachings herfrinv may be embodied in diiferent mechanical constructions and maybe combined rFhus in an instrument having a. plurality of manuals, thepercussive effects,v may be confined to one of them, allowing sostenuto effects to be obtained on anu other. Again, percussiveeifects may be conned to a particular voice on one manual.

Modifications may be made in my invention Without departing from the spirit of it; but having outlined the principles of the invention in certain exemplary embodiments, what I claim as new and desire to secure by Letters Patent is:

1. In aphotoelectric instrument, interrupting means. for imparting tov a beam of light variations at a desired fundamental frequency and a. wave form pattern for imparting thereto variations corresponding generally to harmonics of the fundamental, said wave form pattern having progressive variations thereon for producing changes in said last mentioned variations corresponding to minor shifts in the frequencies of said harmonics.

2. In a photoelectric instrument, means providing a Wave pattern to impart harmonic frequencies to an interrupted beam, said means comprising a wave form screen containing repre.- sentations of more than one cycle of the Wave, said cycles differing from each other, and means for causing different parts of said waveA form screen to modifyv said beam of light successively to generate a tone of varying timbre.

3. In a photoelectric device, interrupting means for interrupting a beam of light, Wave form means for modifying the interrupted. beam to impart frequency variations which are harmonics of theV frequency of interruption, said last. mentioned means having a plurality of wave form. modifying means in different` parts and differing from each other, means for producing relative movement of the light beam and said modi fying meansin one direction to effect addition of said harmonics, and means for producing relative movement of the beam and the modifying means in another direction to produce variations in the harmonic additions.

e. In a photoelectric device, a wave form screen bearing a succession of wave form representations differing one from another in harmonic content, and scanning means for producing relative movement of a beam of light and said screen in such manner as to cause said beam of light to scan said representations successively.

5. 'Ihe structure claimed in claim 4, wherein said wave form representations constitute a repeatab-le series and wherein said scanning means operates. to scan said representations successively and repetitively.V

6. In a device of the character described, a pitch disc having a series of light openings arranged spirally thereon in combination with a` wave form screen carrying a succession of differing wave form representations which are scanned successively by said spirally arranged light openings.

7. In combination, a pair of discs rotating on diiferent centers, one of said discs having a series of circumferentially arranged light slots,A the other of said discs having a series of circumferentially arranged light slots crossing the path of the rst mentioned series in one radial position, and a wave form screen, said arrangement of said slots causing a beam of light to scan said wave form screen progressively from one end to the other while said beam is movingl transversely across said screen, said screen bearing representations: of successive wave forms of diierent: harmonic content in position to be soscanned;

8; In combination, a disc having a series of' circumferentially arranged slots therein to interrupt a beam of light at a fundamental frequency, a wave form screen carrying a plurality of representations of cycles of the fundamental frequency differing as to harmonic content, and a scanning means interposed in the path of light including said first mentioned means for causing the beam of light to scan different parts of said wave screen successively.

9. In combination, a disc having a series of circumferentially arranged slots therein to interrupt a beam of light at a fundamental frequency, a wave form screen carrying a plurality of representations of cycles of the fundamental frequency differing as to harmonic content, and a scanning means interposed in the path of light including said first mentioned means for causing the beam of light to scan different parts of said Wave screen successively, a playing key shutter, means in connection with said playing key to control said beam of light, driving means in connection with said playing key to initiate movement of said scanning means, and driving means effective upon release of said key for returning said scanning means to starting position.

10. The structure claimed in claim 9 in which said Wave form screen embodies a general variation in density throughout said representations so as to eiect a progressive variation in the intensity of said beam of light during said scanning.

11. The structure claimed in claim 8 wherein said scanning means operates to scan one portion of said wave form screen repetitively, and one portion thereof non-repetitively.

12. In a photoelectric instrument, having a source of light and a photocell, interrupting means for interrupting a beam from said source at a desired fundamental frequency, wave form means for imparting to the interrupted beam variations at frequencies responding to harmonics of said fundamental frequency, said Wave form means having progressive variations therein for shifting the phase and varying the amplitude of said harmonics, and scanning means for causing said beam to scan said variations successively and repetitively, to imitate the beats obtained from instruments played in unison.

13. In a photoelectric instrument having a light source and a photocell, interrupting means for interrupting a beam of light from said source at a desired fundamental frequency, wave form means to impart to the beam variations responding to harmonics of said fundamental, and further wave form means for producing a progressive variation in the harmonic content of said beam.

14. In a photoelectric instrument having a light source and a photocell, interrupting means for interrupting a beam of light from said source at a desired fundamental frequency, Wave form means to impart to the beam variations responding to harmonics of said fundamental, and further progressively varying wave form means for producing a progressive variation in the harmonic content of said beam, and a progressive Variation in its intensity.

15. In a photoelectric instrument having a light source and a photocell, interrupting means for interrupting a beam from said source at a de sired fundamental frequency, wave form means for producing in said beam variations responding to harmonics of said fundamental, and said Wave form means having Variations for progressively changing the nature of said variations both as to the individual amplitude and frequency of said harmonics.

EDWARD M. JONES.

Certificate of Correction Patent No. 2,439,392. April 13, 1948. EDWARD M. JONES It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction es follows: Column 3, line 37, for the Word forms readform; column 13, line 10, for ke5T shutter, read key, shutter; and that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 8th day of June, A. D. 1948.

[smh] THOMAS F. MURPHY,

Assz'sta/nt Uommissioner of Patents.

Certificate of Correction Patent No. 2,439,392. April 13, 1948. EDWARD M. JONES It is hereby certified that errors appear in the printed specification of the above numbered patent requiring correction as follows: Column 3, line 37, for the Word forms reed form; column 13, line 10, for key shutter, read key, shutter; und that the said Letters Patent should be read with these corrections therein that the same may conform to the record of the case in the Patent Office.

Signed and sealed this 8th day of June, A. D. 1948.

[smh] THOMAS F. MURPHY,

#Assistant Uommssioner of Patents.`

Images