JPS583559B2 - Denshigatsuki - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION This invention relates to improving the voice (timbre) of electronic musical instruments, and in particular to providing "chip" and other transitional timbre effects, and reducing the harmonics contained in the generated sound over time during attack and decay. This invention relates to an electronic musical instrument that modulates with a function of and generates a musical tone with a special tone.

In general, there is a sound produced by electronic musical instruments called the percussion voice.

This is a complex tone that is played by combining a sustained tone like an organ and a note with an attack/decay envelope like a piano at the beginning of the sound.

The effect is to include a percussive sound at the beginning of the sound generation.

One object of the present invention is to realize the generation of percussive sounds as described above in an electronic musical instrument.

More generally, it is the realization of many types of percussive-like "passage notes" in electronic musical instruments.

One special passing sound effect is called "chiff".

"Chifu" refers to a sustained sound that contains percussive sounds of a specific foot.

In a pipe organ, a "nuff" occurs during the attack and the pipes produce a sound that is predominant with respect to the 3rd or 5th harmonic.

This dominant harmonic rapidly decreases in relative strength as it reaches regular high notes, and the pipe begins to articulate.

On an electronic organ, a chiff can be imitated by playing a short grace note at the beginning of the note.

Grace notes are generated by manipulating 8 feet, which imparts a chiff effect.

Typically, chifu is used only in dearperson and flute-type tones.

Another object of the present invention is to realize chiffism in an electronic musical instrument by emphasizing a certain Fourier component contained in a musical sound waveform synthesized in real time.

Generally, the number of different tones available with electronic organs is very limited.

This is particularly noticeable in musical instruments in which musical tones are generated by a set of multiple oscillators, and high-order harmonics of the musical tones are emphasized or removed by filters.

Another object of the invention is to transform or introduce a new set of harmonic coefficients in electronic musical instruments, making it possible to provide the player with a very wide range of timbre choices.

The various objects mentioned above are achieved in the electronic musical instrument of the type described in the specification of the aforementioned Japanese Patent Application No. 47-65873 (Japanese Patent Application Laid-Open No. 48-90217).

In this type of electronic musical instrument, musical tones are generated by calculating amplitude values XO(qR) at sequential sample points qR of a musical sound waveform in real time, and converting the amplitude values into sounds while performing this calculation.

Each sample point amplitude value is calculated at regular time intervals tx according to the following equation.

Here, q is a variable that increases with each time interval tx,
The values n=1, 2, 3...W represent the order of the Fourier component to be calculated, and R is a frequency pick-up that establishes the fundamental frequency of the generated musical tone.

Attack and Decay are time-dependent scale factors that describe the amplitude envelope of the generated sound.
S(t).

The timbre of the generated sound is established by a set of harmonic coefficients Cn representing the relative amplitudes of the constituent Fourier components.

For example, the Dear Person sound can be obtained by using harmonic coefficient set "A" as shown in Table 1 below.

A flute sound is produced when the "B" coefficient is used.

A very economical scheme for the generation of transitional tones is based on the observation that transitional tones are usually flute tones.

Referring to Table 1, it is shown that the flute tone has one dominant Fourier component.

In other words, the flute waveform is a substantially sinusoidal waveform.

In set "B" of Table 1, the dominant coefficient is C1, making the flute voice the nominal fundamental frequency of the selected note name.

On the other hand, if the second harmonic coefficient (n=2) C2 is a relatively large value, and all other coefficients C1 and C3~
Assuming that Cta has a value of zero, a four-foot flute tone will be generated at twice the frequency of the selected note name.

From the above, a sound like a flute has one harmonic coefficient Dn
It is clear that it can be generated using only ', where the order n' establishes the number of feet of the generated sound.

According to this invention, a transitional tone, such as a flute, has a homogeneous order (=n) in the set used to generate the desired tone.
') is generated by adding one coefficient Dn' to the coefficient Cn=Cn'.

For example, if the set of coefficients "A" in Table 1 is used to produce a dear person sound, a passing sound such as a two-foot flute will have one coefficient D for the corresponding coefficient C4.
This can be provided by simply adding n'=D4.

Usually this is done only during the attack part of the sound generation.

The desired passing sound is achieved. FIG. 3 shows a suitable circuit for generating a passing tone in the manner just described.

A passing sound is generated and rapidly discontinued, as described below in connection with FIGS. 1A, 1B, and 1C.

On the other hand, the magnitude of the coefficient Dn' can be varied in time, creating a gradual onset and termination of the passing sound.

This is explained in FIGS. 1E and 1F, as described below.

More complex transitional vocal effects can be achieved by using the harmonic coefficient D for transitional tones.
This is achieved by preparing n sets separately and adding these to the corresponding (order) harmonic coefficients Cn for the selected timbre.

Such an implementation is illustrated in Figure 4, where the second
executed according to the formula.

The sound produced by a flute-like transition note as mentioned above is a special case of equation (2), where Dn - Dn' is a substantial value and for all other values of n, D
n=0.

Another aspect of the invention includes modulating the harmonics as a function of time during attack and decay.

This is achieved economically by limiting the Fourier components included in the current waveform amplitude value calculation.

At the beginning of the attack, only higher harmonics are included.

This is achieved, for example, by initially setting all harmonic coefficients C1 to Cm to O.

As time passes during the attack, the value m representing the lowest order Fourier component included in the amplitude calculation becomes smaller.

As a result, low-order Fourier components gradually appear in the generated sound.

Conversely, the value m during decay gradually increases. Thus,
At the beginning of Decay, all Fourier components are included in the calculation, but then only the higher-order Fourier components are included.

The circuit of FIG. 6 performs such harmonic modulation.

As explained in Table 1 above, various tones can be obtained simply by using various sets of harmonic components Cn.

Another aspect of the invention relates to an apparatus for providing an addendable or selectable set of such coefficients for use in an electronic musical instrument.

In the exemplary embodiment of FIG. I, the special audio is obtained via an external data insertion device, such as a card reader, or from storage of an additional set of harmonic coefficients in an auxiliary memory.

The invention will now be described in detail with reference to embodiments of the accompanying drawings.

The timbre improvements disclosed herein operate in conjunction with the basic electronic musical instrument 10 of FIG.

In this device, each time one of the keyboard switches 11 is pressed, a corresponding sound is generated via the sound system 12.

The sound of the vocalization is based on the harmonic coefficient Cn recorded in the memory 13.
set by a set of

In the embodiment of FIG. 2, a sinusoidal or flute-like transition tone is provided during the attack by means of additional circuitry 14.

This circuit 14 uses an adder 15 to add the transition coefficient Dn' to the corresponding harmonic coefficient Cn=Cn'.

Therefore, the apparatus shown in FIG. 2 executes the second equation in a special case in which the values of all coefficients Dn except for order n=n are set to 0.

If the circuit 14 is removed, the basic electronic musical instrument 1
0 (FIG. 2) operates according to the first equation and generates a sound without a passing timbre.

The improvement of each tone disclosed here is based on this basic electronic musical instrument 1.
First, the operation of the electronic musical instrument 10 as described above will be explained.

Sequential waveform sample point amplitude value X.

(qR) is calculated at regular time intervals tx according to the first equation above.

In this example, up to W=16 individual Fourier components are calculated separately in corresponding calculation times tcp1 to tcp16.

This calculation time is established by a pulse generator 16 supplying pulses via line 11 every time interval tcp to a counter 18 modulo W=16.

The contents of counter 18 represent the order n of the Fourier component currently being calculated.

A signal representing the order n is provided on line 19.

The timing pulse of calculation time tx is provided on line 20 by slightly delaying the reset pulse of counter 18 (occurring at time tCp16) in delay circuit 21.

The fundamental frequency of the generated sound is established by the frequency number R read out from the frequency pick-up memory 23 in accordance with the selection of the keyboard switch 11.

At the beginning of each calculation time tx line 24, gate 2
The frequency number R supplied via 5 is added to the previous content of the tone adder 26.

The contents of adder 26 provided via line 27 represent a value (qR) representing the waveform sample point currently being calculated.

Preferably, the tone adder 26 is modulo 2W, where W is the highest harmonic component calculated by the device 10.

In this example, 16 Fourier components are sufficient to synthesize most pipe organ sounds, so W=
It was set at 16.

Each calculation timing pulse tpc is supplied via line 17 to gate 28.

This gate 28 supplies a value qR to a harmonic adder 29 which is cleared at the end of each amplitude calculation time tx.

In this way, the contents of the harmonic adder 29 are calculated at each calculation time tcp1.
Increased by the value qR every ~tcp16, the content of adder 29 represents the quantity (nqR).

This value is obtained on line 30.

Address decoder 31 receives from line 30 and device 32
Read from.

The sine function storage device 32 can be configured as a read-on memory.

Here, D is a memory division constant.

This configuration is fed to line 33.

Next calculation time tcp2 Thus, in general, counter 1
8 is supplied from the chord function memory 32.

As mentioned above, the set of harmonic coefficients Cn is stored in the harmonic coefficient memory 13.

As each sine function value is applied on line 33, its corresponding nth order component harmonic coefficient Cn is read from memory 13 by memory address control circuit 36 which receives value n from line 19.

The read value Cn is fed via line 37, adder 15 and line 37a to a harmonic coefficient scaler 38 where it is multiplied by the amplitude scale factor S(t) of the attack/decay occurring on line 39. Ru.

The multiplication result S(t)Cn provided via line 40 is passed to a harmonic amplitude multiplier 41.

The output of multiplier 41, which corresponds to the value of the Fourier component currently being calculated, is applied via line 42 to accumulator 43.

The separately calculated Fourier components are summed in an accumulator 43.

At the end of each calculation time interval tx, the content of accumulator 43 is the waveform amplitude value X of the current sample point qR.

(qR). When a pulse of time tx occurs on line 20, the contents of accumulator 43 are transferred via gate 44 to digital-to-analog converter 45.

The accumulator 43 is then cleared in preparation for the summation of the Fourier components for the next sample point, ie for the calculation to begin immediately.

Digital-to-analog converter 45 supplies sound system 12 with a voltage corresponding to the now calculated waveform amplitude.

Since these calculations are performed in real time, the analog voltage from the converter 45 then constitutes a musical waveform with a fundamental frequency established by the frequency number R supplied from the memory 23.

The amplitude scale factor S(t) is determined by a suitable control circuit 4 operating in conjunction with attack/decay control logic 49.
The attack/decay scale factor memory 4T is read by the attack/decay scale factor memory 4T.

Details regarding these will be described later in connection with FIG.

Reading of the attack scale factor from memory 47 begins when keyboard switch 11 is closed.

A "key suppression" signal is generated on line 50 to initiate the closing or attack of this switch.

The sequential scale factor S(t) is determined at the end of each period of the selected tone since, depending on the setting of the switch 51, the full (1) or half tone adder 26 of the tone being generated is modulo 2W. The count reaches 32.

At this time, an output is produced on line 52. With switch 51 now in the position shown in FIG. 2, a signal is produced on line 53 at the end of a complete period of sound generation.

In each half period, the tone adder 26 reaches a count of 16 or 32.

The corresponding output is the count value 8, 16, 24, or 3 via the OR gate 54 at the terminal 5526 of the switch 51.
When it reaches 2, it is supplied to the OR game}56.

supplied to each of these.

Thus, the setting of switch 51 determines whether it is supplied and at what intervals the scale factors S(t) are sequentially read from memory 47.

The circuit 14 of FIG. 2 effectively inserts transitional tones into the sounds produced by the device 10.

This is done by adding one harmonic coefficient Dn' for the elapsed sound to the corresponding harmonic coefficient Cn=Cn' for the original sound during the calculation time of the Fourier component of order n=n'.

In effect, this results in the addition of a passing tone, such as a sine wave or a flute, with a frequency determined by the order n', and the original tone.

Conveniently, the coefficient Dn' is a binary value containing a single binary "1" bit.

The position of this one bit sets the relative amplitude of the passing tone.

In the circuit 14, the coefficient Dn' is read into the shift register 59 by the generation of the "key suppression" signal at the beginning of the tone generation.

Switch 60 allows manual selection of the relative amplitude of the elapsed tones and controls the position of a single bit in coefficient Dn'.

For example, with switch 60 set as shown in FIG. 2, a single binary bit is loaded into second shift register location 59-2.

Binary 0s in all remaining positions 59-1 and 59-3
~59-i.

The order n' of the coefficients Dn' of the passing tones is set by the signal supplied on line 61.

As shown in Table 2 below, the value n' establishes the number of feet of the passing note.

This n' signal can be provided on line 61 from a manual switch (not shown) so that the selection of the number of feet of the elapsed note can be made by the performer.

Conversely, the input of line 61 can also be fixed so that a constant value n' is always supplied.

Insertion of the elapsed tone begins as soon as any of the keyboard switches 11 is selected.

The "Key Suppress" signal on line 50 then sets flip-flop 62 to the "L" state and AND gate 63
is ready for operation.

The value n' applied on line 61 is compared with the value n from line 19 by a comparator 64.

When a Fourier component of order n=n' is being computed, comparator 64 provides an output via line 65 and enabled AND gate 63, enabling gate 66.

As a result, the coefficient Dn' for the passing tone is applied from shift register 59 via line 67, gate 66 and line 68 to adder 15 where it is summed with the corresponding harmonic coefficient Cn from line 37.

The mixed coefficients (Dn'+Cn) are supplied to a harmonic coefficient scaler 38 via line 37a.

In this manner, apparatus 10 calculates the waveform sample point amplitude according to the second equation.

The insertion of the transition sound is completed.

The length of time that the elapsed sound is inserted is established in a elapsed sound duration counter 69 which is reset by the "Key Suppression" signal on line 50.

With switch 10 in the position shown in FIG. 2, counter 69 counts timing pulses from elapsed tone duration rate clock generator 71.

When the preset count value is reached, counter 69 provides a signal on line 72 which resets flip-flop 62.

As a result, AND gate 63 is disabled so that the comparison signal from comparator 64 can no longer enable gate 66.

This prevents the coefficient Dn' from reaching the adder 15, so that the sound generation continues only with the harmonic coefficient Cn.

In other words, the insertion of the elapsed sound is finished and the device 10 continues to generate sound according to the first equation.

The duration of the elapsed tone is related to the number of cycles of the original tone generated by the device 10.

For this purpose, the switch 70 is switched to the contact 70a,
The one period, half period or full period signal from line 53 is provided to elapsed tone duration counter 69.

After a predetermined number of such signals have been generated, counter 69 provides a signal on line 72 indicating the end of elapsed tone insertion.

Figures IA, IB, and IC illustrate the transitional insertions described above.

Waveform T4 in FIG. 1A represents the original sound generated by electronic musical instrument 10 using only harmonic coefficients Cn.

Although this waveform is shown as a sine wave for simplicity, it is typically a complex waveform.

The inserted transition note itself is illustrated by waveform 75 in FIG. 1B.

This sound is generated by a coefficient Dn' of order n'=3, and this elapsed sound 75 starts at time To when the keyboard switch 11 is pressed and is established by the elapsed sound duration counter 69. It ends at time T8.

Waveform 76 in FIG. 1C is the actual waveform generated by device 10. Waveform 76 in FIG.

Until time T1, this waveform 76 consists of the original sound 14 and the elapsed sound 7.
This includes combinations of 5 and 5.

The waveform 16 generated following the abrupt end of the elapsed tone at time T1 contains only the original tone 74.

In waveform 76, both the original tone and the transition tone begin abruptly.

However, the attack scale factor S(t) actually supplied from the memory 41 (FIG. 2) is applied to the scaler 38 to scale (weight) the combined harmonic waveform number (Cn + Dn') supplied to the line 372. used in

As a result, the amplitude of the combined original sound and elapsed sound gradually increases during the attack period.

This is illustrated by waveform 77 in Figure 1D.

This waveform 77 is the same as waveform 76, but its amplitude gradually increases as a result of weighting by the attack scale factor S<t>.

It is desirable that the elapsed tones die out slowly, rather than ending abruptly as shown in Figure 1B.

This decay of the elapsed sound can be completed immediately by simply shifting the contents of the shift register 59 to the right (to the lower digits) during the insertion of the elapsed sound.

Recall that this register 59 contains a coefficient Dn', which consists of a binary number with a single 1 bit.

The position of this one bit determines the relative amplitude of the passing tones.

By shifting a single 1-bit position of register 59 to the right, value Dn' is decreased.

Obviously, the value is halved each time it is shifted to the right by one place value.

The relative amplitude of the inserted transition tone correspondingly decreases.

Right shifting of register 59 is accomplished by enabling AND gate 79 by closing elapsed tone decay switch T8.

This allows shift pulses to be provided to register 59 from elapsed time decay rate clock generator 80.

These pulses can also be supplied for the entire duration of the elapsed tone, so that decay begins immediately.

Conversely, supplying an operating signal from the counter 69 to line 81 during the desired portion of the elapsed sound generation enables the AND gate 79 during the slow portion of the elapsed sound generation. You can also do it.

As a result, the gradual decay of the passing sound is as shown in waveform 82 of FIG. 1E.

Here, the insertion of the elapsed sound ends at time T2.

Between times T1 and T2, elapsed tones continue to be inserted with gradually decreasing amplitude.

Looking at FIG. 1E, both the elapsed sound and the original sound increase in amplitude since they are set by the attack scale factor S(t) at the beginning of the sound.

In a converse example, as shown in FIG. 3, when the original note exhibits a normal attack amplitude characteristic, the elapsed note may start abruptly at maximum amplitude.

The resulting waveform 83 is shown in FIG. 1F.

To achieve this, the harmonic coefficients Cn are weighted by the attack/decay scale factor S(t), while the coefficients Dn' for the passing tones are unweighted.

This means that the harmonic coefficient memory 1
3 directly to the scaler 38.

Coefficient Dn' is added to the weighted coefficient S(t)·Cn produced on line 40a from scaler 38.

This is performed in adder 15a, whose output Dn (S(
t).Cn) is supplied to the harmonic amplitude multiplier 41 via line 40b.

As a result, the waveform shown in FIG. 1F is generated.

In the elapsed sound insertion circuit 85 of FIG. 4, a set of harmonic coefficients Dn related to elapsed sounds are held in register positions (stages) 86-1 to 86-16 of the shift register 86.

Each coefficient Dn is added to the harmonic coefficient Cn of the original tone of the corresponding order in an adder 15' during the transition tone insertion.

As a result, the electronic musical instrument 10 equipped with the circuit 85 generates musical tones according to the second equation.

The period for inserting the elapsed sound is gate 8 throughout the entire elapsed sound period.
7 is set by counter 69'.

When enabled, gate 87 transfers the harmonic coefficient Dn held in first shift register location 86-1 to line 8.
8 to an adder 15'.

The coefficient Dn is cycled through register 86 in synchronization with clock pulse tcp, thus starting register position 86-1.
always holds a harmonic coefficient Dn of order n corresponding to the coefficient Cn supplied to line 37 at the same time.

The circulation is accomplished by feeding the output of the first register location 86-1 back through line 88, gate 89 and line 90 to the last register location 86-16.

The right shift operation of register 86 is enabled by clock pulse tcp from line 17.

The circuit described above facilitates the insertion of transition notes of any timbre.

Like a real sound, the elapsed sound contains W Fourier components with unique relative amplitudes established by a set of coefficients Dn held in shift register 86.

The values of some or all of these coefficients may be zero.

For example, the typhus effect is achieved by setting all coefficients Dn to O except for coefficients D3 and D.

As a result, the 3rd and 5th harmonics of the original sound increase like a chip.

Special elapsed sound effects can be achieved in circuit 85 by changing the contents of shift register 86 during elapsed sound generation.

As a result, various sets of harmonic coefficients of elapsed tones are held in each storage device 92,93.

storage device 92 or 9 at a certain time during the generation of the elapsed sound.
The set of coefficients held at 3 is transferred to the position of the previous contents of shift register 86.

For example, after a number of periods of original tones have been generated, elapsed tone duration counter 69' may provide a signal on line 94 that causes harmonic coefficients to be transferred from storage 92 to shift register 86.

The signal on line 94 via OR gate 95 sets flip-flop 96 to a "1" state.

This disables the gate 89 and the previous coefficients of the shift register 86 are not rotated.

The memory read control circuit 97 is activated and the line 19
The coefficient Dn of order n indicated by the signal is read from the storage device 92.

The read coefficients are applied to shift register locations 86-16 via gate 99 and line 90, enabled by signals on lines 98 and 94.

This transfer operation continues until all 16 coefficients Dn have been transferred from storage device 92 to shift register 86.

Counter 100 determines that the transfer is complete.

That is, the AND gate 101 is enabled by the output "1" of the flip-flop 96, and the timing pulse t from the line 17 is activated.

p is made conductive to the counter 100.

When the count reaches W-16 upon transfer of all 16 coefficients to register 86, counter 100 sends a signal on line 102 that resets flipflop 96.

As a result, the transfer is completed and gate 89 is re-enabled, allowing shift register 86 to continue cycling.

During the second half of the elapsed tone period, another output from counter 69' is provided on line 103.

This signal initiates the transfer of a set of harmonic coefficients from second storage 93 to shift register 86 via gate 104.

In this way, the set of harmonic coefficients used for elapsed sound generation is varied in a planned manner as a function of time.

Very special elapsed tonal effects can be achieved.

Details of the attack/decay scale factor memory 47, memory read control circuit 48 and attack/decay control logic circuit 49 are shown in FIG.

During the attack and sustain periods, the scale factor 8(t) is supplied from a memory 47a comprising a plurality of storage units 47-1 to 47-P, each holding a separate value S(t).

These memory scale factors correspond to multiple positions 1
The data are sequentially read out under the control of parallel read shift registers having 06-1 to 106-P.

Only a single of these positions has a binary "l" bit.

The storage section of the memory 47a corresponding to the register location holding this "■" bit supplies the scale factor S(t) to the line 107, and further supplies the gate 108 and the OR gate 10.
9 to line 39.

The above-described read operation of the attack-sustain scale factor memory 47a is started each time the keyboard switch 11 is closed.

For example, if the C7 tone is selected by closing the corresponding switch 110, the one-shot multivibrator 113 is connected via line 111 and OR gate 112.
A signal is added to

This generates a "key press" pulse on line 50 that initiates reading of memory 447a.

This "key suppression" pulse is applied to the "read" control input of shift register 106 and places a binary "1" at location 106-1.
” bit and binary ``0'' bits in all other locations.

This "key suppress" pulse also sets flip-flop 114 to the "l" state and enables AND gate 115.

Accordingly, a quarter or half or full period pulse on line 53 is applied via AND gate 115 to the "shift" input of register 106.

As a result, a single binary "l" held in that register
The bits are advanced from position to position as each pulse on line 53 occurs.

Thus, the sequential scale factor S(t) is read from the memory 47a at a rate commensurate with the generation of the sequential waveform of the selected original sound.

A single "1" bit in register 106 is located at location 106-P.
The attack ends when .

At that time, flip-flop 11 is connected via line 116.
A signal is applied to the reset input of 4.

This resets flip-flop 114 to the "0" state, which disables AND gate 115 and no longer provides shift pulses to register 106.

The last attack scale factor S(t) held in the storage unit 106-P continues to be supplied via the line 39 until the selected keyboard switch 11 is released.

This means that the scale factor in the storage unit 106-P establishes the envelope amplitude of the generated sound during the sustain period.
That's what it means.

Decay begins when the selected keyboard switch 11 is released.

In order to continue sounding throughout the decay period, the frequency pick-up memory 23 is read out in response to the flip-flops 118 each associated with a corresponding keyboard switch 11.

For this reason, switch 1 for C7 note,...D2 note, C2 note
10, . . . 119 and 120 are each flip-flop i
18-i, . . . 118-q and 118-r are connected to the set human power S.

For example, when switch 110 is closed, flip-flop i1a-i is set and a signal is provided via line 121 to store the frequency number R associated with the C7 note in memory 2.
3.

Even if the key switch 110 is released, the flip-flop 1
18-1 is not reset immediately.

Therefore, the signal on line 121 remains high level,
The selected frequency number R is stored in memory 2 during the decay period.
It continues to be read from 3.

However, opening switch 110 causes the output of OR gate 112 to go low.

Therefore, inverter 122 sends out a high level output that triggers one-shot multivibrator 123.

This trigger generates a "Decay Start" signal on line 124.

This signal generates an amplitude scale factor S(t) on line 39 from Decay scale factor memory 47b.

This Decay Start signal on line 124 is applied to the read 1 control input of a parallel read shift register 128 for reading Decay scale factor memory 47b.

Similar to the register 106, the shift register 128 includes a plurality of positions (stages) 128-1 to 128-k corresponding to each storage section 129-1 to 129-k of the memory 47b.

At the beginning of Decay, a single binary ``1'' bit is read into shift register location 128-1.

The storage section 129-1 corresponding to this is the storage section 47-1 of the attack-sustain scale factor memory 47a.
Preferably, it has a scale factor S(t) of the same or similar value to that stored in p.

The output "1" from flip-flop 125 enables AND gate 130 which applies a quarter or half or full period pulse from line 53 to the shift input of register 128.

Therefore, the Decay scale factor S(t) increases every k times a single "1" bit is shifted in register 128.
This data is sequentially read out from the memory 47b.

This results in a reduction in the amplitude of the original sound generated.

Decay ends when a single "1" bit arrives at the last register location 128-k.

At this time, a "decay done" signal is generated on line 131.

This signal resets all flip-flops 118, finishes reading the selected frequency pick-up from memory 23, and stops sounding.

Furthermore, this "day-key end" signal is applied to flip-flop 1.
25 to the "0" state.

This disables gate 127 and enables gate 108 to ensure that the attack scale factor from memory 47a is provided to pin 39 the next time keyboard switch 11 is pressed.

Harmonic modulation as a function of time between attack and decay is realized in electronic musical instrument 10 using circuit 135 of FIG.

At the beginning of the attack, only higher-order Fourier components are included in the waveform amplitude calculation.

As the attack progresses, lower-order Fourier components are added to each calculation until all W Fourier components are included in the calculation.

This point is illustrated by Table 3 below.

In this table, '0' indicates that the corresponding Fourier component is removed from the waveform amplitude calculation, and '1' indicates that the component is included.

Considering the above example, during the first 1/4 period of the generated sound,
Only Fourier components of order n=6 and above are included in the calculation.

The number of Fourier components included in the calculation increases until the 11th quarter period, when all W components are included.

During decay, the highest order harmonics are removed from the amplitude calculation.

With increasing time, a large number of Fourier components are removed as shown in Table 4, illustrated below.

In the contents of Table 4, only Fourier components of order n=5 or lower are included in the amplitude calculation during the first quarter period of the decay.

All higher-order Fourier components have been omitted.

In the latter quarter period of Decay, additional Fourier components are removed until only the fundamental component (n=1) is utilized in the ioth quarter period.

The examples shown in Tables 3 and 4 are implemented by circuit 135 (FIG. 6).

In particular, shift register 136 contains a single 16-bit binary class, each of which is selected regardless of whether the order of the corresponding Fourier component is included in the amplitude calculation.

For example, at the start of the attack, the shift register 136 contains the number (000001111111111l).

This corresponds to the second line of Table 3 above.

Each bit is contained in a respective shift register location 136-1 through 136-16 associated with a respective value of n.

The contents of the shift register 136 are shifted one step to the left (to the upper digit) at each calculation time tcp.

The first register location 136-1 is applied to the gate 140 via line 137, AND gate 138 and line 139.

If register location 136-1 contains a binary 1, then the signal on line 139 is sent to gate 140 to provide the harmonic coefficients on line 37 to harmonic coefficient scaler 38 via line 37'. make it operational.

This component is included in the waveform amplitude calculation.

On the other hand, if register location 136-1 is a binary 0
, no signal is supplied to line 139 and gate 140 does not operate.

This prevents the corresponding harmonic coefficient Cn from reaching the scaler 38, thereby effectively eliminating the corresponding Fourier component from the amplitude calculation.

Shift register 136 is read at the beginning of each computation time when pulse tx occurs on line 20.

The number read into register 136 is determined by a set of flip-flop amplifier circuits 142 in conjunction with attack/decay cycle counter 143.

Counter 143 is reset at the beginning of the attack by the generation of the "Key Suppression" signal on line 50.

This signal, provided via OR gate 144 and line 145, resets both counter 143 and all flip-flop circuits 142.

The first full period, half period or 1/2 period occurring on line 53.
Due to the generation of the four-period pulse, the counter 143 stores a count of 1.

The resulting output on line 146 sets the first flip-flop circuit 142-1 to the "I" state.

This operates gate 147 to shift register location 13.
The binary number l is supplied to all of 6-6 to 136-16.

The state of flip-flops 142-2 to 142-6 is "O".
” remains the same.

Therefore, binary 0 is sent to each of the positions of shift registers 136-1 to 136-5.

That is, the contents of shift register 136 now match the second row of Table 3.

The shifting operation of register 136 and the readout to line 139 proceeds as described above, so that only n=6th order or higher Fourier components are included in the waveform amplitude calculation.

When a quarter period of sound is generated, a corresponding pulse is generated on line 53.

When counter 143 reaches count 3, flip-flop circuit 142-2 is set.

Therefore, when the next pulse tx occurs on line 20, 2
The base number 1 is the shift register position 136-5 to 136-16
added to.

The current contents of register 136 correspond to the contents of the fourth row of Table 3.

Therefore, Fourier components of order n=5 or higher are included in the amplitude calculation.

The operation continues in this manner, and the flip-flop circuit 14
2-3 to 142-6, the counter 143 counts 5, 7, 9.
, 11, they are set respectively.

The binary scale l is then read into all positions of the shift register 136 on each occurrence of pulse tx, and all W Fourier components are included in the amplitude calculation.

Similar operations are performed throughout decay.

A "decay start" pulse applied via line 124, OR gate 144 and line 145 resets all flip-flop circuits 142 and counter 143.

In addition, a "decay start" pulse sets flip-flop circuit 149 to a "1" state.

This disables AND gate 13B and causes other AND gates 150 to operate.

The signal from shift register location 136-1 is then inverted by inverter 151.

The inverted signal is applied to gate 140 via AND gate 150 and line 139.

As a union of inversions, the binary number l in register location 136-1
suppresses G-N40, thereby removing the corresponding Fourier component from the waveform amplitude calculation.

Conversely, if a binary number O is applied to location 136-1, gate 140 is enabled and the corresponding harmonic coefficient Cn is applied to scaler 38.

During decay, shift register 136 is read as accurately as during attack.

Therefore, when the first pulse of the next "decay start" occurs on line 53 and counter 143 counts 1, flip-flop circuit 142-1 is set.

A binary number (0000011111111111) is read into shift register 136.

Now, as mentioned above, since it is an inversion operation, register 1
The above content of 36 will include the first five Fourier components in the waveform calculation.

Thus, all components of order n=6 or higher are removed from the calculation.

That is, the circuit performs the harmonic modulation shown in the second row of Table 4 above.

As the decay progresses, fewer components are included in the calculation, and by the time the ninth pulse is generated on line 53, only the fundamental (n=1) component is used. .

Since the flip-flop 149 is set at the calculated value l1, the AND gate 152 is thereafter inactivated throughout the decay period, and the flip-flop circuit 142-6 is not set.

When the decay ends, a "decay done" signal on line 131 resets flip-flop 149.

This is an AND gate 150 to prepare the next sound generation.
138 and 152 are made operational.

A limitation of the harmonic modulation system just described occurs when the desired sound has few harmonics.

For example, suppose an 8-foot flute is played.

This sound mainly consists of one sine wave that coincides with the fundamental frequency, and therefore only the Fourier component of order n=1 has a substantial value.

Therefore, in the attack example shown in Table 3, the sound is l1
It is not generated until the second quarter period, at which time the fundamental wave generation begins abruptly rather than gradually.

Also, an unpleasant "keying click sound" is produced.

Similarly, if a foot sine function or flute-like note is played, the attack will be gradual but with a steep decay.

This is clear from Table 4, and ■feet signal (n=
(corresponding to the 6th harmonic) ends precariously after the second l/4 period of the decay.

These limitations are overcome through the use of additional circuitry 154, shown in dashed lines in FIG.

A set of switches 155 are associated with flute-like or Class I-like sinusoidal tones.

If the 8 foot sine function stop switch 155a is closed, the counter 143 is preset to a count of 11.

As a result, the attack immediately proceeds from the 11th step listed in Table 3 above.

This therefore causes all 1's to be placed in register 136 and all harmonics are included in the waveform amplitude calculation immediately from the beginning of the attack.

The selected 8 foot sine function tone is mainly or totally (n
Since it consists of a single Fourier component of order = 1, this component is included in the amplitude calculation from the beginning of the attack.

The sound is generated slowly and precisely as desired.

During decay, counter 143 remains inactive no matter what sine function tone is selected.

As a result, all zeros are read into shift register 136 and all harmonic coefficients Cn are added to scaler 38.

If any one of these coefficients Cn has a substantial amplitude, the sound will continue to be generated throughout the decay with the envelope amplitude controlled by the decay school factor S(t) supplied on line 39. .

Decay does not end abruptly. To disable counter 143, a signal is applied to AND gate 158 via OR gate 157.

During the decay period, the output of the flip-flop circuit 149 is “1”.
” enables the AND gate 158 and therefore, when either switch 155 is closed, the signal is on line 15.
9 to the "dead" terminal of counter 143.

Preparation of the special musical tone of the electronic musical instrument 10 is performed using circuit 1 in FIG.
70 can be easily achieved.

Typically, the device 10 is provided with not one but a set of harmonic coefficient memories 13, 13' and associated readout control circuits 36, 36'.

Each of the memories 13, 13' stores a set of harmonic coefficients Cn associated with a respective timbre.

For example, memory 13 stores coefficients of set "A" of Table 1, while memory 13' stores coefficients of set "B" of Table 1.

Therefore, if the stop tab switch 171A is closed, a dear person sound is generated, and if the stop tab switch 171B is closed, a flute sound is generated.

The number of such memories 13, 13' included in the device is of course an economic factor.

Generally, device manufacturers provide sufficient musical tones to satisfy the average user.

However, greater flexibility in timbre is desired by musicians.

For example, the devices on sale contain tones for rather recreational purposes.

However, musicians desire a more accurate pipe organ-like tone.

Musician-selected musical timbres are conveniently achieved by providing apparatus 10 with an external timbre insertion device 172, which is used to add harmonic coefficients Cn to apparatus 10.

For example, device 172 may be a marked card, a tape sensing perforated card reader, a perforated tape reader or an optical reader, a magnetic card, a magnetic tape reader, a diode pegboard, or simply a switch device.

Any musical tone tone is selected by turning the switch 173 to connect it to the terminal 173a.

This is random access memory (scratch pad) 1
74 is connected to the scaler 38 via line 175, switch 173, and line 37b.

As a result, during the waveform amplitude calculation, the appropriate coefficient Cn receives the current value n from line 19 in memory readout control circuit 17.
The signal is added from the memory 174 under the control of the control circuit 176 of No. 6.

The harmonic coefficients of any musical tone timbre are read into random access memory 174 from insertion device 172 when switch 177 is in the position shown in FIG.

For example, device 172 may be a punched card reader.

The manufacturer of the device provides for the user a set of perforated cards, each of which contains in coded form a set of harmonic coefficients for a different musical timbre.

The musician then selects the desired tone and device 172 inserts the card.

The coefficients of the card are transferred to random access memory 174 from which coefficients are read in real time as required for waveform synthesis.

Alternatively, a relatively large read-only memory 178 may be provided containing a large number of sets of harmonic coefficients relating to different musical tones.

Switch 177 is used to select which of these stored special tones are transferred to random access memory 174 for use in the electronic musical instrument.

You can also mix tones more freely.

For example, if switch 173 is set to position 173b, the harmonic coefficients stored in random access memory 174 are combined with the coefficients added from selected memory 13.13'.

The coefficients summed in adder 179 are applied to scaler 38 so that device 10 produces a sound having the combined characteristics (tones) of two or more separate tones.

The embodiments of this invention will be summarized and listed below.

■． Synthesizing a musical sound waveform by calculating the amplitude of sequential sample points of the waveform in real time, converting the waveform amplitude into a musical sound signal while performing the calculation, and individually calculating the composition Fourier components of the musical sound waveform. an electronic musical instrument having a generator configured to calculate each waveform amplitude by summing these Fourier components, and to establish the relative amplitude of each Fourier component by a set of harmonic coefficients; a readout control device for supplying each timbre modification information from said storage device as scheduled in accordance with the calculation of the Fourier component of the corresponding order; a changing device that changes a harmonic coefficient of a corresponding order in accordance with the voice change information, and causes the generator to use the changed harmonic coefficient to set a relative amplitude of a corresponding Fourier component; An electronic musical instrument that dynamically changes the timbre of a musical tone signal.

2. In the electronic musical instrument according to item 1, the storage device includes a register that holds a single harmonic coefficient of a predetermined order related to the transition tone, and the readout control device stores the order of the Fourier component currently being calculated and the transition tone. a comparator for comparing the order of a single coefficient relating to a tone; and a gate device supplying the coefficient relating to the elapsed tone from the register to the changing device when the comparator detects a match of the comparison orders; The electronic musical instrument includes an adder for adding a coefficient related to the elapsed tone to a harmonic coefficient of a corresponding order to obtain a elapsed tone.

3. The electronic musical instrument according to item 2 further includes a transition sound amplitude control device that controls the relative amplitude of the transition sound by changing the position of a single bit held in the register, An electronic musical instrument, wherein a coefficient related to the elapsed note is configured by binary information having a "1" bit, and a relative amplitude of the elapsed note is set depending on the position of the "1" bit in the binary information.

4. 3. The electronic musical instrument according to item 3, wherein the register is a shift register, and the amplitude control device is a shift device that shifts the contents of the shift register at controlled time intervals.

5. In the electronic musical instrument according to item 1, the storage device stores a set of harmonic coefficients of different orders regarding passing tones;
The electronic musical instrument is characterized in that the changing device includes an adder for adding each coefficient of the supplied elapsed tone to a harmonic coefficient of its corresponding order, so as to obtain a elapsed tone.

6. In the electronic musical instrument according to item 5, only a large value among the set of harmonic coefficients related to the elapsed sound is a coefficient of a Fourier component contributing a third or fifth harmonic to the musical sound waveform. An electronic musical instrument designed to produce a chiffon effect.

7. 5. The electronic musical instrument according to item 5, further comprising a device for changing the value of a harmonic coefficient related to the elapsed tones stored in the storage device at controlled time intervals.

8. In the electronic musical instrument according to item 1, the storage device includes:
The value of each bit represents whether or not the corresponding Fourier component is included in the waveform amplitude sum, and the shift register holds a set of those binary bits, each associated with an individual Fourier component order. comprising, said modifying device makes available to said generator a harmonic coefficient associated therewith if the corresponding bit has a binary value l, and a value if the corresponding bit has any other binary value. The electronic musical instrument includes a gate that supplies a harmonic coefficient of O to the generator, and changes harmonics contained in the musical tone signal in accordance with time.

9. In the electronic musical instrument of clause 8, the minimum value of the order of the Fourier components included in the total waveform amplitude during the attack period is gradually increased by loading different sets of binary bits into the shift register at controlled time intervals. The electronic musical instrument further comprises a device configured to gradually increase the minimum value of the orders of the Fourier components included in the total amplitude in the decay period by E.

10. The electronic musical instrument according to item 1, further comprising an external sound insertion device for supplying the set of harmonic coefficients to the electronic musical instrument.

11. a first device for calculating the Fourier component individually at each sample point, the calculation being performed at a clock rate independent of the period of the waveform, and calculating the total Fourier component of a periodic complex waveform sequentially for each sample point; An accumulating device that mathematically adds each component value calculated by the first device to obtain a waveform amplitude at each sample point, and a device that generates a musical tone based on the waveform amplitude obtained by the accumulating device. and the first device includes a tone selection device including a keyboard switch for selecting a frequency pick-up that establishes the interval of sequential sample points, that is, the period of the generated musical tone; A sound adder whose stored content represents the sample point of the waveform amplitude that is being calculated by adding the frequency pick-up to the stored content which is the total value before the operation, and a sound adder whose stored content represents the sample point of the waveform amplitude being calculated; a device for obtaining a trigonometric function value whose variable is the product multiplied by the number of orders of the Fourier component by the sample point representing the sample point; and a set of harmonic coefficient values representing the relative amplitude of each constituent Fourier component of the original sound. a harmonic coefficient memory for storing and each calculated Fourier component being summed in said accumulator to obtain said sample point amplitude by multiplying said each trigonometric function value by a modified harmonic coefficient of its corresponding order; an attack and decay signal device that individually issues an attack start signal and a decay start signal in response to the press and release of any one of the keyboard switches in an electronic musical instrument having a multiplier that obtains respective values;
a changing device for changing the harmonic coefficients read from the harmonic coefficient memory as a function of time during an attack period or a decay period which are respectively started by the attack start signal and the decay start signal; An electronic musical instrument that changes the tone of the generated musical sound depending on the part.

12. In the electronic musical instrument according to Item 11, the changing device includes a storage device that stores harmonic coefficients related to at least one passing tone, and converts each harmonic coefficient related to the passing tone into a harmonic coefficient related to the original tone of the corresponding order. and an adder for obtaining the modified harmonic coefficients which are used in the multiplication device when calculating the Fourier component value of the corresponding order, and the timbre is changed by adding the elapsed tones to the original tones. An electronic musical instrument that has changed.

13. In the electronic musical instrument according to item 12, an elapsed tone duration counter counts the number of cycles or partial cycles of musical tones generated from the time when the attack start signal is generated, according to the content held by the tone adder; an electronic musical instrument, further comprising: a device for changing the value of a harmonic coefficient regarding the elapsed sound in the storage device according to a count value of the elapsed sound duration counter;

14. In the electronic musical instrument of clause 13, the storage device is configured to store a unit that is added to the harmonic coefficients for the original note of the corresponding order for a time established by the elapsed note duration counter to produce a sinusoidal elapsed note. An electronic musical instrument characterized in that the electronic musical instrument is a register that stores harmonic coefficients related to one elapsed voice.

15. In the electronic musical instrument of item 14 above, a single binary “
A binary number having a “1” bit constitutes a harmonic coefficient for said passing tone, and further a single “1” of said binary number
An electronic musical instrument characterized in that it has a clock device that shifts the position of bits to vary the relative amplitude of elapsed notes as a function of time.

16. In the electronic musical instrument according to the above item 14, there is provided an attack scale factor memory having a set of attack scale actors that set the amplitude envelope of the original generated sound during the attack period; an attack memory read control circuit for sequentially reading out attack scale factors from said scale factor memory according to a frequency pick-up of a period; and a scale currently read by said control circuit before said modified harmonic coefficients are utilized in said multiplier. and a scaler for scaling (weighting) the modified harmonic coefficients by the value of a factor.

17. In the electronic musical instrument according to item 16, the total harmonic coefficients of the elapsed tones and original sounds summed by the adder are scaled by the scaler, and the combined original tones and elapsed sounds are both combined with the stored attack. An electronic musical instrument characterized by having an envelope established by a scale factor.

l8. In the electronic musical instrument according to item 16, only the harmonic coefficients related to the original sound are scaled by the scaler before being added to the harmonic coefficients related to the elapsed sound in the adder, so that only the original sound, not the elapsed sound, is scaled. has an envelope set by the stored scale factor.

19. In the electronic musical instrument of Item 11 above, the changing device includes a storage device containing a single binary information having one binary digit respectively associated with each Fourier component included in each waveform amplitude calculation; operating to read out a harmonic coefficient from said harmonic coefficient memory, and if one binary digit associated therewith is a binary value "1", causing said harmonic coefficient to be utilized in said multiplier; If the related binary digit is any other binary value, the gate sets the value of its harmonic coefficient to 0 and supplies it to the multiplier, and the gate starts operating according to the attack start or decay start signal, and the attack and decay start signals are activated. or a time control device for selectively varying said single binary information at predetermined time intervals during decay, selectively scheduling predetermined Fourier components from each waveform amplitude calculation during attack and decay. An electronic musical instrument that changes the tone by erasing it.

20. In the electronic musical instrument of Clause 19, the storage device includes a shift register whose contents in one position are shifted in accordance with the reading of the harmonic coefficient memory and whose contents are used to control the operation of the gate; The control device is reset by the attack start signal and the decay start signal, and counts the number of periods or partial periods of the original sound generated from the start of the attack or the start of the decay, depending on the contents stored in the sound adder. An electronic musical instrument, characterized in that it includes a counter and a read circuit for feeding further binary information into the shift register when the counter reaches a predetermined preset count value.

21. In the electronic musical instrument as set forth in Section 20, the reading circuit is arranged so that only Fourier components of an order larger than the gradually lower minimum order are included in each waveform amplitude calculation in the attack, and An electronic musical instrument characterized in that information is fed into said shift register so that only Fourier components of a smaller order are included in each waveform amplitude calculation in Decay.

22. 22. The electronic musical instrument according to item 21, further comprising a circuit for suppressing a change in the compositional Fourier component when a sound with a sine waveform original like that of a flute is generated.

23. Synthesizing a musical sound waveform by calculating the amplitude of sequential sample points of the waveform in real time, converting the waveform amplitude into a musical sound signal while performing the calculation, and individually calculating the composition Fourier components of the musical sound waveform. and a generator that calculates each waveform amplitude by summing these Fourier components, and sets the relative amplitude of each Fourier component by a set of harmonic coefficients. Adding a single binary information consisting of a 1 bit and all other binary 0 bits to one of said harmonic coefficients, the relative amplitude established by the position of the single 1 bit in said binary information. An electronic musical instrument equipped with a device that generates a sine waveform transition sound.

24. Synthesizing a musical sound waveform by calculating in real time the amplitudes of sequential sample points of the waveform, converting the waveform amplitude into a musical tone while performing the calculations, individually calculating the composition Fourier components of the musical sound waveform, and A generator is provided which calculates the amplitude of each waveform by summing these Fourier components, and sets the relative amplitude of each Fourier component, that is, the timbre of the generated musical tone, by a set of harmonic coefficients, as described above. A plurality of sets of harmonic coefficients are provided, which may be additionally inserted in the electronic musical instrument adapted to be utilized by said generator by selection via a stop tab switch. a storage device for storing a set of harmonic coefficients supplied to the storage device; a switch for connecting the generating device to utilize harmonic coefficients from the storage device in place of the previously supplied set; an electronic musical instrument comprising an external sound insertion device for inputting a set of harmonic coefficients to obtain a special musical timbre;

[Brief explanation of drawings]

FIGS. 1 to 1F are graphs showing examples of elapsed waveform shapes derived by the circuits shown in FIGS. 2 and 3, and FIG. 2 shows an example of the present invention. 3 is a Budanck diagram of an electronic musical instrument in which a flute-like transition sound is obtained by changing the harmonic amplitude coefficient for one predetermined Fourier component, and FIG. 3 shows an example of a modification of FIG. 2. FIG. 4 is a partial block diagram of an electronic musical instrument in which the elapsed sound starts abruptly from the maximum amplitude. FIG. 4 shows another embodiment of the present invention. FIG. 5 is a partial block diagram of an electronic musical instrument in which wave amplitude coefficients are changed to obtain complex tonal transition sounds. FIG. 5 shows the attack/decay control logic circuit and scale actor memory of the electronic musical instrument shown in FIG. 2. A block diagram illustrating an example, FIG. 6, shows a further embodiment of the present invention, in which a predetermined Fourier component is removed from the calculation of the sequential sample point amplitude value of the musical sound waveform during attack and decay. FIG. 7 is a partial block diagram of an electronic musical instrument in which the harmonics contained in a sound are changed over time. In each of the above embodiments, an auxiliary memory for harmonic amplitude coefficients is provided to enable selection of various tones. It is a block diagram showing an example of a device. 10... Electronic musical instrument, 14... Additional circuit, 85... Transition sound insertion circuit, 49...
Attack decay control logic circuit, 154...additional circuit.

Claims (1)

[Claims] 1. The amplitude value of a sequential sample point of a musical sound waveform is calculated by individually calculating each Fourier component passing through the sample point and summing these components, and the relative amplitude of each Fourier component is a first device that outputs harmonic coefficient change information corresponding to a predetermined order of the Fourier component during a certain period when a musical tone is generated, in an electronic musical instrument that is set by a harmonic coefficient corresponding to each harmonic coefficient; and a second device that changes a harmonic coefficient of a corresponding order based on harmonic coefficient change information output from the first device, the electronic musical instrument changing the timbre of a generated musical tone. 2. In the electronic musical instrument according to claim 1, the value of the harmonic component change information outputted from the first device in accordance with one or more specific orders is temporally changed. An electronic musical instrument that dynamically changes the relative amplitude of harmonic components of the relevant order to obtain a passing sound. 3. The amplitude value of the sequential sample points of the musical waveform is calculated by individually calculating each Fourier component at the sample point and summing these components, and the relative amplitude of each Fourier component is calculated by the corresponding harmonic coefficient. In the electronic musical instrument of the set method, a first device outputs harmonic designation information corresponding to a predetermined order of the Foury component during a certain period when a musical tone is generated; and a second device for removing harmonic components of a corresponding order from calculation of each sample point amplitude value based on harmonic designation information, the electronic musical instrument is configured to change the timbre of a generated musical tone.