US3446905A - Electrophonic musical instrument - Google Patents

Description

May 27, 1969 I ELEGTROPHONIC MUSICAL INSTRUMENT Filed Aug. 13, less Sheet of 5 p I04 7' 6 72 ,0 i Y 704 ai/vmnroes g' I U l l t gap 94p gfp +2P 5MP FUNDAMENTAL UL TPASON/C F'R'OUENC/ES INIPUTS MVE-SAMPLI/VG UNITS I I I @4 U IUD/O FREQUENCY OUTPUT V Qawpz/r w. E. ROBERTS 3,446,905

W. E. ROBERTS ELECTROPHONIC MUSICAL INSTRUMENT May 27, 1969 Filed Aug. 13. 1965 Sheet wmw, Cm,M+

y 1969 w. E. ROBERTS 3,446,905 I ELECTROPHONIC MUSICAL INSTRUMENT Filed Aug. 15, 1965 Sheet 3 of 5 FZyJb.

.. I nhf fl Filed Aug. 13, 1965 Sheet H an b m, 8.4, M* -Eu.

y 27, 1969 w. E. ROBERTS 3,446,905

ELECTROPHONIC MUSICAL INSTRUMENT Filed Aug. 13. 1965 I Sheet 5 of 5 Wow.- wing.) EM {M walla 4:41.. W m

all M4 4) United States Patent 3,446,905 ELECTROPHONIC MUSICAL INSTRUMENT William Elwyn Roberts, Arfon, Uley, Dursley, Gloucestershire, England Filed Aug. 13, 1965, Ser. No. 479,576 Claims priority, application Great Britain, Aug. 25, 1964, 34,781/ 64 Int. Cl. ClOh 1/06, /00 U.S. Cl. 841.22 3 Claims This invention relates to electrophonic musical instruments, of the kind (hereinafter referred to as electrophonic instruments of the kind described) which are arranged to produce musical notes the harmonic content of which may be varied so as to simulate the tone of other musical instruments.

It is well known that a note produced by a musical instrument contains, as well as a fundamental audio frequency which determines the pitch of the note, harmonics thereof whose amplitudes relative to the amplitude of the fundamental determine the tonal quality of the note which is being played. It is therefore necessary, when simulating the tone of a musical instrument, that each fundamental frequency corresponding to a note should contain the necessary harmonic frequency content of that characteristic waveform of the note produced by the simulated instrument.

Two common methods of producing suitable tone waveforms are harmonic synthesis (in which all the waveforms generated may be sinusoidal and are combined in the correct proportions when the note is played) and the formant principle (in which harmonically rich waveforms are generated at all the fundamental pitches required and passed through tone forming circuits which may simulate resonance systems of orchestral instruments). Both methods, particularly the former, have previously demanded an inordinately large number of waveform generators. The present invention provides an electrophonic instrument utilizing comparatively few generators and is applicable to both these methods and others.

According to this invention, in an electrophonic instrument of the kind described, there are provided means for producing a waveform of ultrasonic fundamental frequency and sampling means arranged to sample said waveform, the sampling frequency or a multiple of that frequency differing from said ultrasonic fundamental frequency by the required audio frequency of the waveform. By ultrasonic frequency is meant, in this specification any frequency above the normal range of audio frequencies. Preferably the sampling means are gates controlled by pulses at the required frequency.

By the above arrangement, a waveform of ultrasonic fundamental frequency may be transformed to a waveform of similar shape, and therefore harmonic content, of audio fundamental frequency. The pulses sampling the ultrasonic waveform would be amplitude modulated to an amplitude equal or proportional to that of the ultrasonic waveform at the instant of sampling. If the fundamental frequency of the ultrasonic waveform is L, cycles per second and the pulse repetition frequency is 1, cycles per second, then if f =nf where n is an integer, all the pulses would be equally modulated. If however,

where f,, is less than half f,, the relative phase progression of the sampling pulses with respect to the ultrasonic waveform causes the sampling pulses to be modulated so as to describe a waveform at frequency f,, of similar shape to the ultrasonic waveform.

It would of course be usual to generate the ultrasonic waveform using an electrical signal generator and it may be necessary to provide a detecting circuit, typically a 3,446,905 Patented May 27, 1969 diode rectifier, to demodulate the sampling waveform, the audio frequency signal or signals thereby obtained may then be amplified and applied to a loudspeaker to produce the desired audible note. It is also possible to generate the ultrasonic waveform using an organ pipe of suitable dimensions, and to convert the ultrasonic vibration of the pipe into an electrical signal by a suitable transducer. Alternatively, a vibrating string or other mechanical oscillator may be used to generate an ultrasonic Waveform.

With the above described apparatus, it is possible to use one ultrasonic waveform for a number of different notes having the same harmonic content. It is often possible to simulate the tone of a musical instrument whose tone varies with the pitch of the note by regarding the harmonic content of a note as varying from octave to octave, but to be substantially the same for notes within a single octave. The above described arrangement permits the use of the same ultrasonic waveform for all the notes in a particular octave, and facilitates the reduction in the number of waveform generators used in the instrument.

Alternatively, the ultrasonic waveform may be a sinusoid at a single frequency. The same sinusoid could then be converted to a number of different audio frequencies by sampling a number of outputs of the waveform generator by pulse sampling units of different pulse repetition frequency. The derived audio waveform would then be combined in the required proportions to produce notes of desired harmonic content. Preferably however, means are provided for combining the harmonically related ultrasonic sinusoidal signals in required proportions for an audible note or notes, the resultant waveform or waveforms being sampled to produce the required audio frequencies. Since, in general, a particular audio frequency derived from ultrasonic frequency may serve, not only as a fundamental frequency for one note, but also as a harmonic frequency of other notes at frequencies an integral number of octaves below that fundamental frequency, the same ultrasonic sinusoid may be used to derive harmonics for notes in different octaves, as well as, using sampling pulses of different repetition frequency, for notes of different pitch within an octave. This arrangement likewise reduces the number of signal generators required for an electrophonic instrument. Another alternative would be for the ultrasonic Waveform to be of a particular shape, for example sawtooth or rectangular, which could be converted into waveforms of similar shapes at all the desired audio frequencies. The final tone of the notes would be obtained by passing the waveform through the appropriate filters, or other tone forming elements.

As mentioned above, this invention enables all the notes in a particular octave to be produced from a single waveform of particular harmonic content, and it is therefore possible to simulate approximately the tone of a musical instrument by providing the same tone for all the notes within an octave. According to one aspect of the invention, in an electrophonic instrument of the kind described, there are provided means for generating a waveform of ultrasonic fundamental frequency; a plurality of sampling means, the number of sampling means corresponding to the number of separate notes to be produced within an octave; the sampling frequency of a first sampling means or a multiple of that frequency differing from the fundamental frequency of the ultrasonic waveform by the audio frequency of the corresponding note in the octave, and the difference in sampling frequency of a sampling means differing from the sampling frequency of said first sampling means by the frequency difference of the corresponding notes to be produced by the respective sampling means.

With this arrangement, a single ultrasonic waveform may be used to provide tone waveforms for each note or each of a series of notes within a particular octave. This represents a reduction in the number of waveform generators necessary to produce the tone waveforms of all the notes within an octave.

Preferably, a pulse sampling unit is provided for each semitone within at least one octave of the instrument. This construction enables switches which would connect the pulse sampling units to the output of the ultrasonic waveform generator to be controlled by a keyboard similar to that used in a piano or pipe organ.

The means for producing the ultrasonic waveform and the sampling units may be controlled by crystal oscillators..The human ear is capable of detecting variations in pitch of a few parts in a thousand and it is therefore necessary to maintain the frequencies of the ultrasonic waveform and the pulse repetition frequency with extremely high relative stability.

As hereinbefore mentioned, it is possible to combine ultrasonic sinusoidal waveforms in desired proportions so as to be able to produce notes of different harmonic content at different fundamental frequencies.

According to another aspect of this invention, in an electrophonic instrument of the kind described, there are provided a plurality of wave sampling units, the number of units corresponding to a number of notes within an octave to be produced by the instrument, the pulse repetition frequency of a unit differing from that of a first unit by the difference in audio frequency of the respective notes in the lowest octave to be produced by the instrument; a plurality of generating means each arranged to generate a single ultrasonic frequency differing from a multiple of the pulse repetition frequency of said first sampling unit by a different multiple of the lowest frequency to be produced by the instrument; and a plurality of adding means, each arranged to combine the outputs of selected generator means in selectable ratios so that the output of each adding means contains a different fundamental frequency having a selected harmonic content, each wave sampling means being arranged to be selectively connected to the output of each adding means.

With the above described construction, the lowest note in the range of the instrument may be produced by connecting the output of the first wave sampling means to the adding means whose output contains the lowest ultrasonic frequency. For example, the lowest ultrasonic frequency could be 25,256 c./ s. and the first wave sampling unit could have a pulse repetition frequency of 25 kc./s., so that the resulting audio frequency would be 256 c./s. The 2 th harmonic of this frequency may be obtained by applying to the same pulse sampling unit a frequency of (r.25,000+2 .256) c./s. where r and m are integers. It would be usual to provide ultrasonic frequencies corresponding to a continuous range of values of m so as to produce a continuous range of notes. It will be seen that it is not necessary for all the ultrasonic frequencies produced to be harmonically related, since for example the 2 th harmonic of a 256 c./s. note is produced what-ever the value of r in the above expression. Preferably however, the ultrasonic frequency genera-tors are arranged in sets, each set producing harmonically related outputs. This enables a minimum of crystal controlled oscillators to be provided, only one being used for each set and the other generators in each set being frequency dividers, mixers, or tuned amplifiers.

An adding circuit would conveniently be provided for each octave in the range of the instrument and may conveniently comprise an adding amplifier with resistive feedback having a plurality of input resistors corresponding to the number of harmonics to be present in a note, the values of the input resistors being such that the output of the adding amplifier has harmonic components in the desired proportion for a note in that octave. It would be possible, however, to provide several sets of resistors for each amplifier, the input resistors having different values in the different sets so that a single note may have a number of different tones corresponding, for example, to different stops in a pipe organ. The selection between the various adding circuits for a particular note would conveniently be made by a switch corresponding to a stop on a pipe organ.

As mentioned above, the production of-a note in instruments of the kind described requires that a pulse sampling unit corresponding to a note within an octave be connected to the output of an adding circuit, the frequency components of which determine the pitch and tone of the note. It has been found hitherto that the switching action produces audible keying clicks due to frequencies produced by rapid switching of continuously running generators. These keying clicks have in the past been eliminated by filtering out the frequencies produced by the switching action. However, since a note produced by a musical instrument has characteristic attack and decay time constants, it is desirable to simulate these while eliminating the frequency components produced by rapid switching. For connecting a waveform generator to a sampling unit, there may be provided a switch having a pair of elongated terminals mounted thereon, and a movable contact member of soft insulating material having a thin facing of slightly conductive material. The movable contact member being arranged to meet the terminals simultaneously and at an angle thereto, so that the attack and decay of a note varies according to said angle.

As mentioned above, an audible tone is produced by providing, at the output of the detecting circuits, an amplifier and loudspeaker. It would be possible to provide a preamplifier and a loudness control unit which may be set for various loudness levels and for variable control of the loudness, and other conventional units.

In the above description, the pulse sampling units have been mentioned as producing pulses which are amplitude modulated by an ultrasonic waveform. It is also possible to arrange for the frequency changing of the ultrasonic waveform to be caused by various other methods of modulation, for example phase or frequency modulation.

The following is a description of a specific example of this invention applied to an electronic organ, reference being made to the accompanying drawings in which:

FIGURE 1 is a diagram illustrating the production of a single musical note;

FIGURE 2 is a diagram illustrating the production of seven octaves with the systems shown in FIGURE 1;

FIGURES 3a and 3b illustrate the tone forming circuits associated with one manual of an electronic organ;

FIGURES 3c and 3d illustrate the switching circuits associated with one manual and the wave-sampling circuits for the whole organ;

FIGURE 4 is a diagram illustrating a stop adding circuit; and

FIGURE 5 illustrates a switch for the switching circuits.

As mentioned above, the specific example to be described refers to the production of an electronic organ capable of producing notes resembling those of a pipe organ. It would have been apparent from the foregoing that a number of possible arrangements of the system is possible, depending on the particular tone spectrum of the instrument simulated. It is well known that the harmonic structure of notes from some musical instruments is highly pitch dependent. In such instruments, basic vibrations of a variety of fundamental frequencies are applied to a single fixed resonance system. The same band or hands of frequencies present in the basic vibration will be amplified or attenuated whatever the fundamental pitch. In terms of harmonic number the tone spectrum of this kind of instrument varies greatly over its pitch range. The pipe organ does not belong to this class of instrument as there is a separate resonance system as well as vibration generating system for each pitch of each stop. For this reason, the variation in tone spectrum with pitch is comparatively small. The slight changes that do occur may be mainly due to frequency dependent absorption coefiicients and variations in pipe wall response. For the present instrument, the same tone waveform is used for all pitches within an octave but may be varied from octave to octave. The waveform is adjusted to be as close as possible to the mid-octave waveform of the instrument to be simulated. The waveform could of course be varied every half octave or less to improve the approximation.

Another important factor in synthesizing music realistically is the so called chorus effect. This is due to small low frequency variations in pitch of sounds from a single instrument such as an organ pipe. The frequency or phase of these variations Will generally be different for each instrument and when two or more sound together the interaction enriches the musical chord. The system of the new instrument includes an arrangement for producing this effect.

The organ system is based on the arrangement shown in FIGURE 1 which would be capable of producing one musical note. A waveform at ultrasonic frequency of similar harmonic content to that required for the audio frequency tone waveform is generated by unit T. The ultrasonic frequency waveform is fed to wave sampling unit N where it is sampled with short duration pulses of repetition frequency i the unit N incorporating a pulse generator providing the short duration pulses at that repetition frequency and a sampling gate, the pulse generator being controlled by the oscillator X From the output terminal of unit N the tone waveform is obtained at the audio frequency f and of similar shape to the input waveform. The output tone waveform is then applied to the power amplifier A which feeds the loudspeaker system L. The audio tone waveform frequency is given by the difference between and f or a multiple of f as previously explained. Since frequencies f and i will preferably be above 20 kc./s, small relative variations in them could result in large relative variations in the audio frequency output. Some method of maintaining and 1 with extremely high stability is therefore essential. Accordingly units T and N are controlled by crystal controlled oscillators X and X respectively to ensure that the ultrasonic waveform and the sampling waveform are maintained with a high degree of accuracy.

It would obviously be unreasonable to repeat the system of FIGURE 1 for each pitch required in a full organ or any other key board instrument. The method of obtaining seven octaves or more with the minimum of units is explained with the aid of FIGURE 2.

In FIGURE 2, tone waveform generators T to T run at the ultrasonic frequencies given in the following table.

TABLE 1 Tone generator Frequency T1 fT fT T3 IT3=2E+4P T4 In T5 [T fT T1 fr =F+8P sible to filter out the high frequency components but this may be an unnecessary complication of the system.

(b) The highest ultrasonic tone waveform frequency component that can be reproduced at audio frequency is limited by the sampling period duration. With simple circuitry the minimum sampling period duration feasible is about /2 micro-second, limiting the highest useful harmonic frequency in the ultrasonic waveform to l mc./s. In FIGURE 2 the units N to N and N to N are wave-sampling units. The sampling rate of the wave sampling unit N is E/ 8 cycles per second. If the output of tone generator T be applied to wave-sampling unit N a tone waveform of pitch P will result P may represent the pitch of the note C that is to say bottom C on the instrument, so that the application of the output of tone generator T to wave-sampling unit N will produce the note C which has the pitch 2P. Similarily, the output of tone generator T applied to wave-sampling unit N will give C The pitch of C (8P), is obtained by applying the output of tone generator T to wave-sampling unit N running at a sampling rate Similarly the outputs of tone generators T and T and T, when applied to wave-sampling unit N will yield the pitches of C C and C which are 16P, 32P and 64P respectively.

The sampling rate of wave-sampling unit N is made slightly less than that of sampling unit N so that when the output of tone generator T is applied to it the note C is produced. The sampling rate of wave-sampling unit N may be determined from the equation assuming an even tempered musical scale. The notes C 1? and C 1? are produced by feeding wave-sampling unit N from tone generators T and T respectively. The notes CJ, C 11 C 1? and C 1; are obtained when N is connected to the outputs of tone generators T T T and T respectively. Similarly, all pitches over seven octaves .are achieved with seven ultrasonic tone waveform generators and two sets of twelve wave-sampling units. The sampling frequency of any unit in the series N to N is given by the expression where n is the sampling unit number within the set N to N Similarly, the sampling rate of any unit in the set N to N is given by the equation where n is the sampling unit number within that set.

The above description has been concerned with the generation of the fundamental frequencies of each pitch of the instrument. As previously mentioned, this invention enables a single series of harmonically related tone generators to be used for the production of all the pitches and the required harmonics of those pitches for seven octaves in the complete organ.

FIGURES 3a to 3d illustrate the tone forming and switching circuit associated with one manual and the wave-sampling units which serve the whole organ.

The harmonic generator is shown at H in FIGURE 3a and consists of 20 conventional sine wave generators each producing a single ultrasonic signal on one of the and the seventeen harmonics that provide all the harmonies up to the tenth for these frequencies and up to the fifth for (F+8P). The value chosen for F is 200 k.c./s. while P is the lowest pitch of the organ, that is of the note C which is 32.70 cycles per second.

The units A1 to A7, which will be called stop adding units add vectorially the various harmonics, the amplitudes of which are determined by sets of resistors connected to them via electronic switches. The resistor sets may each contain up to ten resistors and are fed with the appropriate fundamental and all the available harmonies thereof as given hereafter. An adding unit is shown in more detail in FIGURE 4 showing the resistors R R etc. connected to the input of high gain amplifier A via the switch S. Each harmonic is fed to a separate resistor, and the output of the unit may be expressed as etc. where V V V vectorially represent the harmonics and R is the feedback resistor of amplifier A. Numerous resistor sets may be joined to the input terminal of a stop adding circuit so that the output may be a combination of the tone waveforms of many stops. The resistor sets and switches for each adding circuit are represented in FIGURES 3a to d by R and S respectively where m indicates the stop or coupler to which they belong and n the stop adding unit to which they are fed. Signals are allowed through the electronic switches only while a DC. voltage is applied to their control terminals. The manual stop switches 1 to 8 are divided into four groups: 7 and 8 are couplers to the other manual and pedal board, 5 and 6 are 16 foot stops, 3 and 4 are 8 foot stops and 1 and 2 are 4 foot stops, which each permit, when closed, a voltage from DC. voltage source 60 to appear at the control terminals of the respective switches. Tables 2, 3 and 4 show the fundamental ultrasonic frequencies that would be applied to the stop adding circuits for 16 foot, 8 foot and 4 foot stops respectively, the bottom notes C on the key board for the stops 'being 32.70, 65.41 and 130.8 c./s. respectively.

When a number of stops are operative the outputs of the stop adding circuit will be the sum of the individual stop waveforms associated with it. Coupling is effected by applying the outputs of corresponding stop adding units of the other manual and pedal board to A to A via electronic switches similar to the stop switches. The outputs of the stop adding circuits are taken to one side of the keying switches K K etc. as is illustrated in FIGURES 3c and 3d. The switches are controlled by the keyboard as indicated in Table'S. Keyboard octave is the term used for a set of twelve consecutive keys on a conventional keyboard, octave numbers being obtained by counting keyboard octaves from the left hand end of the keyboard. It will be clear from FIGURES 3c and 3d and Table 5 that each key within the second and third keyboard octaves have two keying switches operated by them whereas each key in other octaves controls One keying switch only.

The live side of the keying switches are short circuited to earth by secondary switches k k etc. except when the keys are operated. Switches K and k K and k etc. are mechanically coupled (indicated by dotted lines). It is arranged that each secondary switch opens, thus allowing the signal to the corresponding keying switch, a short time before the latter closes, when a key is operated. When the key is released the keying switch opens, then slightly later the secondary switch returns to its closed state. The ensures that break-through of ultrasonic frequency signals across the open keying switches, which would cause a Weak but audible sound, occurs only for a brief instance before and after the true notes sound and cannot be detected by the listener. The secondary switches are ordinary make and break contacts but the keying switches have been designed especially to control the attack and decay of the notes.

The outputs of the keying switches are connected to the wave-sampling units as shown in FIGURES 3c and 3d. For example, the output terminals of K K and K are joined together and taken to wave-sampling unit N which samples the combined waveforms at the rate of F/8. The outputs of K K K and K are joined together to N sampling at the rate The other keying switches of the manual shown are similarly connected to their respective wave-sampling units. Outputs of the other manual switches and pedal board, which are similar in principle, are connected to the same wave-sampling units.

Each wave-sampling unit in FIGURE 1 comprises a crystal controlled oscillator X controlling a pulse generator s to give /2 micro-second pulses at the specified frequencies, which control the sampling gate in the wave sampling unit, and an adding circuit to which the various input signals are applied.

The audio frequency waveforms obtained from the sampling units are added together in the audio adding circuit M. The output of M is fed to a power amplifier to drive a loudspeaker system LS, via reverberators etc. and other control units.

Two factors which characterise a musical note in addition to pitch and harmonic content are the ways in which the note builds up and dies away, called the attack and decay of the note. For example, it takes a finite time for the vibration in an organ pipe to reach maximum intensity after a key has been pressed and then to decay when the key is released. The attack and decay times of an organ pipe depend on the dimensions of the pipe and are therefore pitch dependent. Not only does the amplitude of a note change during the attack and decay periods,

but also its harmonic structure. Exact simulation of the attack and decay characteristics of a musical instrument is difiicult in an electronic organ having continuously running tone generators. All that is done in many organs of this type is to eliminate keying clicks. This may be done by an electrical filter placed after the switch in the signal line, to filter out the harmonics caused by the switching action. Another method is to ensure that the attack and decay of notes are not too abrupt. This is done by having key switches that make gradually as pressure is being applied. This is the principle used in the switches K K one of which is illustrated diagrammatically in FIGURE 5. The switch K consists of two fixed contacts which are a pair of parallel wires W W held on an insulating plate I. The moving contact is a block B of soft insulating rubber with a coating or plate of conductive material Z fixed to one face. As the key is operated the moving contact is arranged to meet the fixed contacts at an angle 0. The conductive surface will therefore touch one end of the contact wires first. As the pressure is increased the rubber block is compressed non-uniformly due to the angle of contact so that the length of the wires touching the conductive surface gets progressively greater during key travel. Then provided the surface conductivity of the moving contact is chosen correctly, the keying effect may be achieved. The sharpness of attack and decay for a given key movement rate will clearly be affected by the angle 0.

The chorus effect has already been discussed above. This organ system lends itself very well to the inclusion of this refinement. Each of the twenty four wave-sampling units used are controlled by a quartz crystal which determines the sampling rate and hence the pitch obtained. The frequency of a crystal controlled oscillator may be varied to a small extent by including capacitance or inductance in the crystal circuit and also by applying a voltage across the crystal.

Referring back to FIGURE 1 back biased p-n junction 100 is shown included in the circuit of crystal 101 for oscillator X The bias is applied by means of DC. source 103 and the voltage across junction 100 is varied by low frequency generator 104 to give the required effect. Each of the twenty-four oscillators would be fed with randomly chosen signals to give a realistic chorus effect.

It will be understood that although the present system more particularly utilizes harmonic synthesis, it would equally be possible to utilizes the formant principle, suitably shaped waveforms being fed from adding circuits A to A and appropriate tone forming filters being provided where necessary.

Instead of using sets of fixed resistors in the adding amplifiers, sets of variable resistors may be provided, with corresponding resistors in the sets ganged together and controlled by a drawbar.

I claim:

1. In an electrophonic musical instrument, the combination comprising:

a plurality of sources of ultrasonic waveforms, having respective fundamental frequencies of (F/ 8+1), (F/4+2P), (F/2+4P) and (F-l-8P), where F is of the order of 200 kc./s. and P is the frequency of a note in the lowest octave to be produced by the instrument;

a first plurality of sampling means, one of said first plurality of sampling means having a sampling frequency of (F/8) and the others of said sampling means having sampling frequencies differing from F/8 by the respective difference between said note and a respective different note in said octave;

a second plurality of sampling means, one of said second plurality of sampling means having a sampling frequency of (F/8-7P) and the others of said second plurality of sampling means having samplingfrequencies differing from (F/87P) by the difference between SP and the respective pitch of a respective note in the octave that includes the pitch 8F; and

means coupling said sources selectively at will to said sampling means thereby to produce desired audible notes.

2. The combination as set forth in claim 1 further comprising a plurality of generators of sinusoidal waveforms having respective frequencies of at least said fundamental frequencies and at least the frequencies 2(F/8+P), 3(F/8+P) 10(F/8+P), 2(F/4+2P), 3(F/4+2P) 10(F/4+2P), 2(F/2-I-4P), 3(F/2+4P) l0(F/2+4P), 2(F+8P), 3(F+8P) 5(F+8P); and a plurality of adding means coupled to receive one of said fundamental frequencies and harmonics thereof in predetermined proportions, said adding means constituting said sources.

3. The combination set forth in claim 1 in which said first plurality of sampling means have the sampling frequencies F/8+P(1-2 where 11 takes the respective values 1, 2, 3 12 and said second plurality of sampling means have the sampling frequencies F/8+8P(% -2 where it takes the respective values 1, 2, 3 12. l

References Cited UNITED STATES PATENTS 2,142,580 1/1939 Williams 841.22 2,478,973 8/1949 Mahren 84-1.23 2,697,959 12/1954 Kent 84-1.22 2,831,107 4/1958 Raymond 307-229 2,848,920 8/1958 Lester 841.17 3,079,571 2/1963 Elliott 334-15 3,355,539 11/1967 Munch 84-1.23

ARTHUR GAUSS, Primary Examiner. H. DIXON, Assistant Examiner.

US. Cl. X.R. 84-101

Claims (1)

1. IN AN ELECTROPHONIC MUSICAL INSTRUMENT, THE COMBINATION COMPRISING: A PLURALITY OF SOURCES OF ULTRASONIC WAVEFORMS, HAVING RESPECTIVE FUNDAMENTAL FREQUENCIES OF (F/8+P), (F/4+2P), (F/2+4P) AND (F+8P), WHERE F IS OF THE ORDER OF 200 KC./S. AND P IS THE FREQUENCCY OF A NOTE IN THE LOWEST OCTAVE TO BE PRODUCED BY THE INSTRUMENT; A FIRST PLURALITY OF SAMPLING MEANS, ONE OF SAID FIRST PLURALITY OF SAMPLING MEANS HAVING A SAMPLING FREQUENCY OF (D/8) AND THE OTHERS OF SAID SAMPLING MEANS HAVING SAMPLING FREQUENCIES DIFFEREING FROM F/8 BY THE RESPECTIVE DIFFERENCE BETWEEN SAID NOTE AND A RESPECTIVE DIFFERENT NOTE IN SAID OCTAVE; A SECOND PLURALITY OF SAMPLING MEANS ONE OF SAID SECOND PLURALITY OF SAMPLING MEANS HAVING A SAMPLING FREQUENCY OF (F/8-7P) AND THE OTHERS OF SAID SECOND PLURALITY OF SAMPLING MEANS HAVING SAMPLING FREQUENCIES DIFFEREING FROM (F/807P) BY THE DIFFERENCE BETWEEN 8P AND THE RESPECTIVE PITCH OF A RESPECTIVE NOTE IN THE OCTAVE THAT INCLUDES THE PITCH 8P; AND MEANS COUPLING SAID SOURCES SELECTIVELY AT WILL TO SAID SAMPLING MEANS THEREBY TO PRODUCE DESIRED AUDIBLE NOTES.

Images