JPS60195592A - Functional waveform generator for electronic musical instrument - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] (Technical Field of the Invention) The present invention relates to a function waveform generator for an electronic musical instrument, and in particular to a function waveform generator for an electronic musical instrument, in which a value can be changed from the beginning to the end from an initial value to a target value with a simple configuration using digital technology. The present invention relates to a function waveform generator for an electronic musical instrument, which generates a function waveform that changes in an S-shape, with gradual changes in the middle and more rapid changes in the middle.

(Background of the Invention) In electronic musical instruments, the frequency (pitch) of musical tones is continuously changed in order to obtain voltament, attack pitch, glide effects, and the like. In this case, in conventional electronic musical instruments, the pitch of the musical tone is controlled by the charge/discharge voltage of the capacitor using, for example, an OR time constant circuit, so the change in pitch is initially large, as shown by the solid line in Figure 1. The changes gradually became gradual. For this reason, unlike pitch changes in portamento performances of natural instruments, where the change is gradual at the beginning and end and largest in the middle, as shown by the broken line in Figure 1, the pitch changes suddenly at the beginning of the pitch change. There was an inconvenience in that it gave an audible sense of discomfort.

In addition, focusing on the above points, in analog circuits,
There is a known method (Special Publication No. 57-56079) that changes the pitch of a musical tone exponentially with respect to time, as shown by the solid line in Figure 2, but when the pitch change ends, the change in pitch stops abruptly. It still feels strange. Furthermore, as shown by the dashed line in Figure 2, Chinoku Special Publication No. 5B-20440) has been proposed in which the pitch change speed is changed midway through to approximate the pitch change of a natural musical instrument, as shown by the dashed line in Figure 1. ,
Since it is a polygonal line approximation, there is also the disadvantage that there is a sense of discontinuity.

(Object of the Invention) In view of the above-mentioned conventional problems, the present invention has been made to improve the digital circuit configuration in order to enable a digital electronic musical instrument to achieve natural pitch changes similar to those of the above-mentioned natural instruments. An object of the present invention is to provide a function waveform generation device for an electronic musical instrument that generates an S-shaped function waveform.

(Summary of the Invention) In order to achieve the above object, the present invention sets first and second reference values having a predetermined relationship with an initial value and a target value, and first sets a function waveform value currently being output (hereinafter referred to as A new function waveform value to be output next is generated by adding a change value corresponding to the difference between the current value and the first reference value to the current value until the current value) reaches a predetermined value from the initial value. (hereinafter referred to as the next value) and repeat this process, and when the current value reaches a predetermined value, change the change value corresponding to the difference between the second reference value and the current value to the current value. By adding the next value and repeating this until the current value reaches the target value, the first reference value becomes the asymptotic value of -■ time, and the second reference value becomes +■ time of darkness. An S-shaped function waveform with an asymptotic value is formed. Here, the above-mentioned initial value and target value regulate the area of the function waveform to be formed, and are set in advance or prior to the function waveform forming operation.

(Explanation of the principle of this invention) A detailed explanation of this invention is as follows.

First, take the difference between the current value f(x) and a certain reference value a, find this difference m, and add it to the current value f(x). ) to (6).

In Figure 3(a), the current value is f(x) and the reference value is a
Assuming that the weighting constant is b (however, a number larger than 1), the relational expression between the next value f (x+Δx) and f(x) is as shown in Equation (1).

f (x+Δx)=f (x)+ (f (x)-8)
・(1/b)・ΔX ・・・・・・(1) This formula (1)
Transposing and rearranging, (f (x+ΔX)−f (X))/ΔX= (1/b
)−f (x) −a/b −・= (2>Here, Δ
When X→O, d/dx −f (x)= (1/b)−f (X)−
a, "b... (3) The general solution to this differential equation (3) is as shown in equation (4).

f (x)-C-exp(x/b)+a ・=-(4)
(However, C is an integral constant.) Also, in Figure 0 (b), the current value is ○ (X), the reference value is a', and the weighting constant is b' (however, a number larger than 1). Then, the current value Ω(X> and the next value o(x+ΔX
) is the inverse of the subtraction term between the reference value a and the current value f(x) in equation (1), so Ω(X+ΔX)-G (X)+(a'-Q (X))
・(1/b') ・ΔX・・・・・・(5) If you solve this equation (5) in the same way as equations (1) to (4), you get the equation (
6) is obtained.

f (X)=C'-eXll(-X/b') +a
'・・・・・・ (6) (However, C' is an integral constant.) Therefore, by appropriately connecting equations (4) and (6), we obtain a slippery slope as shown in Figure 4. An S-shaped curve is obtained. In this case, the curve t'(x) and the curve 1aa(x
) is the cleanest and most preferable method.

Hereinafter, the present invention will be explained in detail using the drawings.

(Embodiment 1) FIG. 5 shows a block configuration of a function waveform generator according to an embodiment of the present invention. The function waveform generation diagram in the same figure shows initial value setting circuit 1, target value setting circuit 2.1, curve generation section 3.0
It is composed of a curve generation section 4, a switching control section 5, etc., and the initial value SV generated from the initial value setting circuit 1 is used as an asymptote at -Q, and the target value T output from the mm measurement circuit 2.
A function waveform is output that sequentially changes in an S-shape from the initial value Sv to the target value TV, with gradual changes at the beginning and end and larger changes in the middle, with V as an asymptote at 10. For this purpose, in this function waveform generator, the above initial value Sv is set as the first reference 11TisD, the target value TV is set as the second reference 11TD, and the clock pulse φ
For every cycle of , one curve generating section 3 r l; L l 1
Itl! A part of the exponential function f(X) with ll5D as the asymptotic value in one flash is formed from the initial value SV to the target value TV, and the 0 curve generating section 4 generates an index that asymptotically approaches the second reference value TD from the initial value 1asv. Form a function 0(X). The switching control unit 5 first outputs the numerical value f(x) formed by the 1-curve generation unit 3 as the current value Cv, and then outputs the function 9(x) formed by the 0-curve generation unit 4. Switching control is performed so that

In FIG. 5, an initial value setting circuit 1 and a target value setting circuit 2 generate an initial value Sv and a target value of 117V, respectively.

The 1-curve generation section 3 includes a subtracter 30, a weighting circuit 31, an adder 32, a comparator 33, an initial value correction circuit 34, a selector 35, etc., and generates this function waveform every cycle of the clock pulse φ. The output value of the device, that is, the current f[cV=
f (X) Nico's presence icv and first standard 1i1
Add the change d' which is proportional to the difference from sD and calculate the leap number (×
) to form the next value CVf'. That is, in this 1-curve generating section 3, the subtracter 30 inputs the current value Cv output from the delay circuit 56 to the input terminal, and inputs the initial value Sv output from the initial value setting circuit 1 to the 8-input terminal. is standard 1
It is input as 11sD. Therefore, the subtracter 30 outputs a difference signal Df (
Dr-CV-8D). The weighting of this difference signal Of is multiplied by 1/kf in the circuit 31, and this weighting is carried out from the circuit 31 by the difference signal D "change d of a sufficiently smaller value.
" is output.

The adder 32 receives the current value Cv output from the delay circuit 56.
is input to the 8 input terminal, and the 8 input terminal is weighted (B circuit 3
The change amount (N) output from 1 is input. As a result, the adder 32 changes the current value C
and the change d input to the 8 input terminal, and the addition result is the added value C M (CVf' = CV + d
f) is output to the A input terminal of the t-reflex 15.

The initial II (II correction circuit 34 has means for generating an initial change Δ consisting of a predetermined positive minute value, and addition and subtraction means, and this initial change Δ is outputted from the initial value setting circuit 1. The calculated initial values Sv+Δ and Sv−Δ, which are formed by adding and subtracting from the initial value Sv, are supplied to the B and C input terminals of the selector 35, respectively.

The comparator 33 compares the target value TV outputted from the target value setting circuit 2 and the initial value Sv outputted from the initial value setting port 1. Then, depending on the magnitude relationship between the target value TV and the initial value Sv, if TV is larger than SV, a discrimination signal AGB-'''1'' is generated, and if TV is smaller than SV, a discrimination signal BGA-''1'' is generated. Further, when TV and Sv are equal, a match signal EQ-"1" is generated. Each of these signals AGB1BGA and EQ has its output terminal connected to the input selection terminal SB of the selector 35.
, AND circuit 36.3 connected to SC and SD
7 and 38, respectively.

Further, the output terminal of an inverter 39 is connected to the six input selection terminal SA of the selector 35.
and the other input terminals of the AND circuits 36, 37 and 38 are supplied with a start pulse S P -"1" to be input to the function waveform generator. Further, the D input terminal of the selector 35 is supplied with the initial value Sv output from the initial value setting circuit 1. As a result, in the selector 35, when the start pulse SP is OP, the inverter 39
The output of , that is, the level of the six input selection terminal SA becomes 1'', and the addition 1icvr' of the input terminal is selected and output.On the other hand, when the start pulse SP-``1'' is supplied, the level of the comparator 33 becomes 1''. If the discrimination signal @AGB is 1'', the output of the AND circuit 36, that is, the level of the B input selection terminal SB becomes 1'', and the calculation initial value Sv+Δ of the B input terminal is selected and output, and the discrimination signal of the comparator 33 is If BGA is 1'', the output of the AND circuit 31, that is, the level of the C input selection terminal SC becomes 1'', the calculation initial value SV-Δ of the C input terminal is selected and output, and the match signal EQ of the comparator 33 is output. is “1”
Then, the output of the AND circuit 38, that is, the D input selection terminal S
When the level of D becomes "1", the initial value SV of the D input terminal is selected and output.The output Cvt of this selector 35 is supplied to the B input terminal of the selector 54 of the switching control section 5.

The Q curve generator 4 includes a subtracter 40. The weighting device is equipped with a circuit 41, an adder 42, etc., and the output value of this function waveform generator, that is, the current value CV, is calculated every cycle of the clock pulse φ.
=o Second 1 in (x)! Quasi-1ifiTD and current value C■
The next value Cvg of the function g(X) is formed by adding the change dΩ proportional to the difference between the two. That is, in this Q curve generating section 4, the sieve reducer 40 receives the target value TV outputted from the target value setting port 2 at the input terminal as the reference value TD, and the delay circuit 56 at the 8-point terminal. The current value CV output from is input. Therefore, the subtracter 40 is Kawaishi mc
Difference signal DO (Dlll −T
D-CV). This difference signal 1)G is weighted by 1/kil1 in a circuit 41, and this weighting is given by a change dQ which is sufficiently smaller than the difference signal Dg.
is output.

In the adder 42, the weighting change dQ output from the weighting circuit 41 is input to the A input terminal, and the delay circuit 56 is input to the B input terminal.
The current value CV output from is input. As a result, the adder 42 adds the change amount dQ and the current *CV, and the added value CV(1(CVg-CV+
dil > is output to the t-rector 54 (7) A input terminal.

The switching control section 5 includes a comparator 50.57, a differentiating circuit 51.5
8, equipped with flip-flop circuits (FF) 52, 59, selectors 54, 55, and a delay circuit 56, output CVf from 1 curve generating section 3, output C from 0 curve generating section 4.
Either Va or the target value TV from the target value setting circuit 2 is selected and output as the current value C■ of the open number waveform.

The comparator 50 compares the weighting of the 1-curve generation section 3 with the variation dO output from the circuit 31 (H and the weighting of the 0-curve generation section 4 with the variation dO output from the circuit 41, and calculates dg. df
When the value is larger than that, a discrimination signal AG82m "1" is generated. The differentiating circuit 51 logically differentiates the rising and falling edges of this discrimination signal AGB2 to obtain this signal AGB2.
When changes from 1" to O" or from "lOl" to 1, the change detection pulse C)-IP1=" has a width equivalent to one cycle of the clock pulse φ (hereinafter referred to as one clock time).
This change detection pulse CHP1="1" is generated.
is supplied to the set terminal S of the FF 52. This FF52
The start pulse SP is input to the reset terminal R, and when this start pulse SP-"1" is input, it is reset and the output Q becomes O", and the change detection pulse CHPI-1" is input. Then, it is set and the output Q becomes ``1''. The output Q of this FF52 is
The signal is supplied to the six input selection terminals SA of the selector 54, and is also supplied to the eight input selection terminal SB of the selector 54 via the inverter 53. As a result, this selector 54
is a start pulse SP-“1” to this function waveform generator.
'' is input, first select and output the output CVf from the 1-curve generating section 3, and calculate the change d of the 1-curve generating section 3.
After the magnitude relationship between " and the change dO of the 9-curve generating section 4 is reversed, the output CVg from the 0-curve generating section 4 is selected and output.

Cono selector 54 output cV'(CVf; l+, tc
vc+) is supplied to eight input terminals of the selector 55.

Function waveform output (current value) in this function waveform generator
CV has the second reference value TD, that is, the target value TV, as an asymptotic value, and does not pass through this asymptotic value. Therefore, in the switching control section 5, the comparator 51, the differentiation circuit 58, and the flip-flop circuit (FF) 59 and selector 55 are not particularly essential, but these circuits can be used to change the current value Cv to the target value, such as when the function waveform generator is modified and the reference value TD is set to a value different from the target value TV. value tv
This is provided in case the person may pass through.

The comparator 51 compares the target 11TV and the current value Cv,
When TV is larger than Cv, discrimination signal AG83m"
1”. The differentiating circuit 58 generates this discrimination signal @AG
By logically differentiating the rising and falling edges of B3, this signal AGB3 changes from 1'' to 0'' or from O'' to 1''.
When the change occurs, a second change detection pulse CHP2-"1" of one clock time width is generated. Fl"59 inputs this change detection pulse CHP2 to the set terminal S and inputs the start pulse SP to the reset terminal R. Therefore, when the start pulse SP="1" is generated, the FF59 is reset and the output Q is When the change detection pulse CHP2-"1'" is generated, it is set and the output Q becomes 1". The output Q of the FF 59 is supplied to the six input selection terminal SA of the filler 55, and also to the inverter 60. is supplied to the B input selection terminal SB of the selector 55 through the selector 55.Thereby, when the start pulse SP-"1" is input to this function waveform generator, the selector 55 first selects the output from the selector 54. CV' is selected and output, and after the current value C■ passes the target value TV, the A input terminal is selected and the target value TV supplied from the target value setting circuit 2 is selected and output. The output of this selector 55 CV# (CV'?-4; tTV)G is the output of the delay circuit 5.
6. The delay circuit 56 delays the output Cv'' of the selector 55 by one clock time and outputs it as the next current value OV.

Next, the operation of the grooved numerical waveform generating device will be explained.

Start pulse 5P-11 is applied to this function waveform generation I'll.
When 1II is input, the selector 35 of the 1 curve generation section 3
In this case, the output of the inverter 39, which is the input to the six input selection terminals SA, becomes the "O" level, and the input selection terminals SB, S
Depending on the output of the comparator 33, one of the outputs of the AND circuits 36°3t and 3B, which are the inputs of C and SD, becomes 1.
Furthermore, since [F52 and 59 of the switching control section 5 are reset and the outputs Q are both set to the "0" level, the selectors 54 and 55 are both in the 8-input terminal selection state. Therefore, the one selected by the selector 35 among the dark, numbing initial values SV+Δ, SV-Δ and SV in which the start pulse SP is #IN is supplied to the delay circuit 56 via the selectors 54 and 55.

This calculation initial value is delayed by one clock time in the delay circuit 56 and output as the initial value CVo of the current value CV.

2. f function V When the start pulse SP goes from 1'' to 1101+, 1
In the curve generating section 3, the input of the inverter 39 is 0'', so the output is 1'', and the AND circuit 36.37.3
8, one input is II O11, so the outputs are all 0". As a result, the selector 35 has the 6 input selection terminal SA set to 1°", enters the A input terminal selection state, and starts the following calculation. .

That is, 6 ICV is currently input to the subtracter 304 ctA input terminal, and the reference fiflsD output from the initial value setting circuit 1 is input to the B input terminal. As a result, this subtracter 30 divides these current values Cv and reference values SD.
The circuit 31 outputs a difference signal Df = CV-8D between
is formed and output to the B input terminal of the adder 32. and,
The adder 32 adds the current value Cv input to the input terminal to the change df input at the time of sunset, and outputs an added value Cvf' as the addition result. This additional value CVf
' is selected by the selector 35 as the output CVf, selected by the selector 54 as the output CV', and the selector 55
This next value Cv'' is delayed by one clock time in the delay circuit 56, and is output as the current value Cv in the next period of the pulse φ. Therefore, in the first curve generating section 3, based on the current value Cv and the first reference value SD, CVf' = CV
+ (CV-8D)/kr next time 1icvr' is formed every cycle of the O-switch pulse φ. As a result, this function waveform generator generates a function waveform Cv that changes exponentially from the initial value SV to the target value TV.

3. Even while forming the z function in the motion 1 curve generating section 3, the 0
In the curve generating section 4, a subtracter 40 outputs a difference signal Da-TD-CV between the second reference value T() inputted to the input terminal and the current value Cv manually inputted to the input terminal, and calculates the weight. The circuit 31 converts this difference signal Da into 1/ko (8L,re(
IJ) do (do = rl /kc+). On the other hand, in the switching control section 5, the comparator 50
The change d' generated in the curve generation section 3 is compared with the change dQ generated in the 0 curve generation section 4, and the differentiation circuit 51 monitors the change in the output of the comparator 50.1 curve generation In section 3, the current value Cv is the reference value SD (initial value SV)
Since the differential signal @D of the compressor 30 and the weighted change value df of the circuit 31 change sequentially from [o-1 to a certain value < (the absolute value of df is On the other hand, in the 0 curve generating section 4, the current value Cv gradually approaches the reference 11TD (target value TD), so that the difference of the 'M calculator 40 is equal to the current value Cv.

And weighting is the change value dQ of the circuit 41 from a certain value to "
(The absolute value of do decreases sequentially). Therefore, when the current value cv sequentially changes exponentially as described above, at a certain point the magnitude relationship between the change values df and d(l is reversed, and the magnitude discrimination signal AGR2 of the comparator 50 becomes 1. "0" or "1"
Changes to

Then, the differentiating circuit 52 detects the change and generates a change detection pulse CHPI having a one clock time width.

As a result, FF52 is set and the output Q is “1''
Therefore, the selector 54 switches to the six input selection state.

In this state, in the 0 curve generation section 4, the subtracter 40 and the weighting circuit 41 input the change d(l) to the input end of the adder 42, and the weighting circuit 41 inputs the 8 inputs of the weighting 12. The current value cv is input to the end.Thereby, the adder 42 adds the change dQ and the current value Cv and outputs the added value CVo as the addition result.This added value CVg is inputted by the selector 54. The output value Cv' is selected as the next value cv'', and the selector 55 selects it as the next value cv''.

This next value c v ″ is delayed by one clock time in the delay circuit 56 and output as the current value Cv. Therefore,
The 0 curve generating section 4 calculates CVIJ-CV0 (CV-
The next value CVO of TD)/k(l is formed every cycle of the clock pulse φ.Thereby, from this function waveform generator, starting from the current value C■ at the time of switching, the target value TV is logarithmically A function waveform Cv which is asymptotic is generated.

Note that the 1 curve switches to the 0 curve (change amount 1i11
(When the magnitude relationship between lr and dg is reversed), the added value CVf' (=CV+df) of the adder 32 and the added value CVO (-CV+dQ) of the adder 42 are almost the same value, so when switching, the current value Cv No sudden changes occur. Therefore, the transition from the 1st curve to the 9th curve is performed smoothly. The point represented by point P in FIG. 4 is the turning point from the 1 curve to the 0 curve, and the slopes of both curves at this point are almost equal.

In addition, in FIG. 4, a corresponds to the initial value Sv, and a′
corresponds to the target value TV and X corresponds to time. Furthermore, weighting is done by selecting the values of coefficients 1/kf and 1/In.
The timing of transition from the 1 curve to the 0 curve can be set arbitrarily.

4. Startup In the above, the reason for using the calculation initial Il value SV±Δ, which is formed by adding or subtracting the initial change amount Δ to this initial value Sv, instead of the initial value Sv as the initial value CVo of the current value CV is It is as follows.

That is, if the initial value 111cVo of the current value Cv is equal to the initial value Sv, the current value Cv supplied to the input terminal of the subtractor 30 is the initial value SV, while the The first reference value SD is also equal to the initial value Sv. Therefore, the difference signal [)f of the subtracter 30 is input to the 8 input terminals of the adder 32 (H becomes "0", and the addition of the adder 32 FjIIicVr', !-L
, T current 1i1CV (tl:rt+Chi initial 1iSv)
will be output as is. Therefore, the current value CV output from the delay circuit 56 remains unchanged as cvo =sv, and an exponential function waveform cannot be obtained. Therefore, by outputting the calculation initial value SV±Δ formed by adding and subtracting the initial change amount Δ to the initial value Sv as the initial 1flcVo of the current value CV, the subtracter 30
The first difference (r=@Or ecVo -8D-(SV+
Δ) -8Vk- and the change (H) is set to a non-zero value. In this way, if the next value different from the current value Cv is calculated in the first calculation, the subsequent difference signal Df and change df
is always a non-zero value, and it is possible to generate a current value Cv that sequentially changes according to an exponential function as shown in equation (4).

The above can also be explained from the above equation (4). That is, when determining the integral constant C by setting the sum of the first reference value SD and the initial variation Δ as the initial value f(0) of the function f(x), equation (4) becomes f(x)−Δ・ext)
(x/kf)+5D-(7). Therefore, Δ
If NO, f(x) becomes an exponential function with Δ as a coefficient, while if Δ-〇, f<X>=SV)=constant value, so in order to generate an exponential function, Δ It can be seen that must be a non-zero value.

Note that when the initial value Sv and the target value TV are equal, there is no need to generate a function waveform, but in this case, the selector 35 of the one-curve generating section 3 outputs the initial value Sv.

By the way, in the above description, the reason why the comparator 33 compares the magnitude relationship between the target value TV and the initial value S■ and selects the calculation initial value Sv+Δ or SV−Δ according to this magnitude relationship is based on the current value Cv. Since the direction of change in the current value CV depends only on the sign of the initial change Δ, such as increasing if the initial change Δ is positive and decreasing if it is negative, the direction of change in the current value CV is changed from the initial value SV to the target value. This is to set it in the direction toward the TV.

In addition, in order to adjust the rate of change of the function waveform generated by this function waveform generator, the period of the clock pulse φ,
Either the initial change amount Δ or the weighting coefficients kf and k(+) may be changed.

FIG. 6 shows a modification of the device of FIG. As shown in FIG. 6, the selector 54 selects the change d(l or d(l), and the adder 32 of the one-curve generator 3 combines the selected change df or dg with the current (11cV). If the addition is performed, the adder 42 of the 0 curve generation section 4
(Fig. 5) can be omitted. Furthermore, when the coefficients ``and klJ'' are equal, if the difference signals Df and Dg are switched by the selector 54, the weighting can be further applied to the circuit 3.

One of 41 can be omitted. In this case, the comparison circuit 50 may compare the difference signals Dr and D(+).

Further, the first reference value and the second reference value may be switched.

In addition, in the embodiments shown in FIGS. 5 and 6, the variation df
The selection operation of the selector 54 is switched at the time when and d(l are equal, but instead of this, the selection operation is changed at the time when the current mcv reaches a desired value between the initial value Sv and the target value TV, or when the start pulse SP The selection operation of the selector 54 may be switched depending on the elapse of time after the occurrence of the event.

Also, in the embodiments of FIGS. 5 and 6, ! is input to the 8 input terminals of the subtracter 30! Although the initial value Sv was used as the l quasi-value SD, instead of this, l! As the quasi-value SD, the value “Sv±
.DELTA.", or a value obtained by adding or subtracting the value .DELTA. from the current value Cv may be added to the input terminal of the subtracter 30, and various modifications are possible. In this case, the initial value Sv can be used as is as the initial value CVo of the current value Cv.

Furthermore, in the embodiments shown in FIGS. 5 and 6, the initial value SV
and the target value TV are arbitrarily set and awaited, and in this case, when the initial value Sv is always set to "0", the initial value setting circuit 1 and the subtraction 30 can be omitted. Also, M half value S
D and TD are initial value Sv and target value T, respectively.
Instead of setting the same value as V, an arbitrary value of 1 (however, a value close to the initial value S■ or the target value TV is desirable) can be set.

Furthermore, in the embodiments shown in FIGS. 5 and 6, the output (next value) CV'' of the selector 55 may be extracted as a function waveform.
It may be added to the 8 input terminals of 7.

(Embodiment 2) Figure 7 shows an example of the overall configuration of a single-note electronic musical instrument that uses the numerical waveform generator of the present invention to obtain a voltament effect. In the figure, the keyboard section 61 is equipped with a large number of key switches corresponding to duplicate keys, and the key press detection circuit 62 detects the on or off operation of each key switch of the keyboard section 61, and detects the pressed key. Identification key information (hereinafter referred to as key code KC) and key-on signal KON indicating key depression
Output. Note that when a plurality of keys are pressed simultaneously on the keyboard section 61, the pressed key detection circuit 62 outputs a key code KC and a key-on signal KON for the single pressed key selected by giving priority to treble or adhesive priority. The frequency information memory 63 receives the key code KC output from the key press detection circuit 62 and outputs a frequency information value F corresponding to the pitch of the key indicated by the key code KC. This frequency information number 11F is, for example, "0°0" for high tone C2.
52325 (decimal display)", r for pitch CI
A value proportional to the frequency of a musical tone, such as ``1.674400 (in decimal notation)'', where 1 bit is an integer part and the other 1 bit is
This data is expressed as a 15-bit binary number with 4 bits as a decimal part. In this way, the frequency information memory 6
The frequency information value F corresponding to the pitch of the pressed key outputted from 3 is supplied to a voltament control circuit 64 constructed using the numerical waveform generator of the present invention.

The voltament on/off switch 65 switches the voltament control circuit 64 between normal operation and voltament operation. When this switch 65 is off, the voltament control circuit 64 is in normal operation, and when the switch 65 is turned off, the voltament control circuit 64 is in normal operation.
When turned on, it becomes voltament operation. and,
This voltament IFilI1m circuit 64 supplies the frequency information value F outputted from the frequency information memory 63 as it is to the accumulator 66 as a changed frequency information value F' during normal operation, and on the other hand, outputs the frequency information value F from the frequency information memory 63 during voltament operation. The calculated frequency information number IF is arithmetic processed and converted into a changed frequency information number F' whose value changes sequentially, and is output to the accumulator 66.

The accumulator 66 generates an accumulated value qF obtained by sequentially adding the modified frequency information value F' at the timing of the clock pulse φ0.
' Output (q-1, 2, 3...).

The output of the accumulator 66 generated in this way (
The accumulated value QF') is supplied to a tone waveform generation circuit 61, which includes a waveform memory that stores, for example, sequential sample point amplitude values of a desired tone waveform. The musical tone waveform generating circuit 61 reads out the integer part of the accumulated value QF' from the waveform memory as an address signal, and sequentially outputs musical waveform data MW representing the amplitude value of each sample point of the musical tone waveform. The musical sound waveform data MW is multiplied by the envelope waveform data ENV outputted from the envelope 1 waveform generation port 69 in the multiplier 68 and given an amplitude envelope. ' will be converted to '. this! #Iris waveform signal MW
' is converted into sound by a sound system 92 consisting of a filter, an amplifier, a speaker, etc., and is emitted as a performance sound.

It should be noted that an electronic musical instrument configured to sequentially accumulate frequency information values F corresponding to the pitches of pressed keys and generate musical sound waveforms according to the accumulated output is disclosed in Japanese Patent Application No. 48-4199, for example.
4 (Japanese Unexamined Patent Publication No. 49-130213), detailed explanation of each part will be omitted.

The sounding width detection circuit 93 is an envelope waveform generation circuit 69.
When the generated envelope waveform data ENV is not "0", it outputs a sound generation signal @ENVE of 1". This sound sound raw signal ENVE is supplied to the voltamento lllll11 circuit 64, and in the portamento control circuit 64, this sound sound raw signal ENVE is During the period 'O'', changes in the modified frequency information value F' are prohibited.

FIG. 8 shows a detailed block configuration of the voltament control section 64. Note that parts common to the dark number waveform generation equipment shown in FIG. 5 are represented by the same reference numerals, and corresponding parts are represented by the same reference numerals with an A appended thereto.

In the figure, a differentiating circuit 71 delays a key-on signal KON outputted from a key press detection circuit 62 by one clock time in a delay circuit 72 and further inverts the signal by an inverter 13. An AND circuit 71 calculates the logical product of the key-on signal KON and the key-on signal KON. It is output as a key-on pulse KONP. That is, this differentiation circuit 71 logically differentiates the rising edge of the key-on signal KON output from the key press detection circuit 62 in FIG. When there is a key), a key-on pulse KONP is generated.

This key-on pulse KONP corresponds to the start pulse SP in FIG.

The latch circuit 1A corresponds to the initial value generation circuit 1 shown in FIG.
is supplied to the O-do terminal, at this time the delay circuit 56
The previously pressed key as the current value being output from (
The modified frequency information value F' corresponding to the pitch (hereinafter referred to as the first pressed key) is taken in and latched. Then, this latch circuit 1A holds the latched changed frequency information value F' until the third key press is performed after the current second key press.
is retained and output. The voltament control circuit 64 receives the changed frequency information value F' latched in this latch circuit 1A.
(Hereinafter, this will be referred to as F'old) is the initial value (reference value), and the frequency information value F output from the frequency information memory 63 by pressing the second key is set as the target value, and changes sequentially exponentially. A modified frequency information value F' is formed and output.

The subtracter 30 outputs the number of changed frequency information 11TiF' as the current value output from the delay circuit 56 and the latch circuit 1A.
The difference signal D (D-F
'F'old). The shift Oo path 31A shifts this difference signal by n pits in the direction of the lower pits and outputs a change value d'' which is multiplied by 1/(K=21)L and is sufficiently smaller than this difference signal.

The comparator 33A compares the frequency information value ``which is the target value with the changed frequency information number 11F' which is the current value, and when the target value F is larger than the current value 11F', the determination signal AGB=''
1". This signal AGB is passed through an inverter I5 to an exclusive OR circuit (EXOR) 76 and an adder 7.
It is supplied to the carry input terminal Ci of No. 7. EXOR76
, when the key-on pulse KONP-'"1" is generated, that is, the initial value F'o is set as the current value F'.
Only the next discrimination signal AGB to which +c+ is output is used. Therefore, in this case, the comparator 33A compares the initial value 1'' old with the target value [.

In other words, the output of EXOR76 is the key-on pulse K
Only when ONP=1111+ occurs, AND circuit 7
8 to the input end of the adder 11. This E
The XOR 76 outputs a positive initial change numerator Δ when the signal AGB is 1''. Also, when the signal AGB is 0'', the 1'' signal output from the inverter 75 causes each of the initial changes Δ to be output. The bit is inverted by the EXOR 76, and "1" is input to the carry input terminal Ci of the adder 71. That is, by adding "1J" to the inverted output of Δ in the adder 11, the value -〇 It is possible to form a two's complement number, ie -Δ.In addition, the comparator 33Δ
is the target value F and the current value F' (initial WiF' old)
When they match, a match signal EQ="1°" is generated.

This coincidence signal E Q = “1” is the inverter 79
is input to the AND circuit 78 via
Turn B off. As a result, the target 111 IF and initial MF
'old is equal to 1-, that is, when the C pitch of the first key pressed and the pitch of the 2111'l key are equal, the addition of the initial change Δ is prohibited and the start of the portamento operation is prohibited. Do 1. Therefore, in the above points, comparator 338 corresponds to comparator 33 in FIG. Comparator 33A also has the same function as comparator 51 in FIG. 5, but this point will be described later.

The inverters 75, 79, EXOR 76, AND circuit 78, adder 17, etc. described above correspond to the initial value correction circuit 34, selector 35, etc. in FIG.

The gate 80 is the envelope waveform generating circuit 69 in FIG.
The raw sound output signal ENVE='″1′.

When the generation of musical tones stops, that is, when the sound generation raw signal ENVE is 0, the latitudinal component dM output from the adder 77 is turned off.
is prohibited from being input to the input terminal of adder 320B.

The adder 32 outputs an added value Ff'' which is the sum of the current value F' at the input terminal and the latitudinal value dM at the 8 input terminals, and supplies it to the 8 input terminals of the selector 54.

In the Ω curve generator 4, the subtracter 40 outputs a difference signal DΩ (−F−F′) between the target value F and the current value F′, and the shift circuit 41A converts this difference signal r)g in the direction of the lower pit. The bits are shifted and a change dQ of "24×DΩ" is output. The gate 81 supplies the change amount dlJ to the adder 42 only when the sound generation raw signal ENVE is 1''.The adder 42 adds the change amount d(] and the current value F', and this added value FO
” is supplied to the input terminal of the selector 54.

The comparator 50 receives the change d output from the shift circuit 41A.
When g is larger than the change d output from the shift circuit 31A, a discrimination signal AGB2-"1" is generated. A change detection pulse CHP1 = rr 1 ++ is generated.F
F52 is reset by the key-on pulse KONP-'"1" and set by this change detection pulse C)-IPl-"1". This causes the selector 54 to output the key-on pulse KONP! When “1” occurs, the addition value Ft” input to the 8 input terminal is selected and output, and when the magnitude relationship between the changes dQ and df is reversed, the addition value F (+“ The output F11 of the selector 54 is supplied to the 8 input terminal of the selector 55, and if the selector 55 is in the B input terminal selection state, it is supplied as is to the delay circuit 56, It is later output as the current value F'.

Further, the discrimination signal AGO of the comparator 33A is further supplied to a differentiating circuit 58. This differentiating circuit 58 includes a delay circuit 85
and E
P2-"1" is generated (this change detection pulse CI-I
P2 is input to the set terminal S of the FF59. As a result, when the current value F' sequentially changes and reaches the target 1iiIF, and the value of the discrimination signal AGB of the comparator 33A is inverted, the FF 59 is set by the change detection pulse CHP2, and the "1"1" signal is set by the change detection pulse CHP2. This is supplied to one input terminal of the AND circuit 87.The other input terminal of this AND circuit 87 is supplied with an FF
52 is inputted via an inverter 88, and as a result, the AND circuit 87 selects the six inputs of the selector 55 after the current value F' exceeds the target value F while one curve is being generated. 1 to terminal SA via OR circuit 89.
As a result, the selector 55 selects the input terminal to output the frequency information value F, which is the target value, as it is as the changed frequency information value F'.

The delay n path 56 delays the target value F or the added value F'' (FM or FO') output from the selector 55 by one o clock time and outputs it as the next current value (changed frequency information numerical value) F'.

The electronic musical instrument shown in FIG. 7 enters normal operation by turning off (opening) the voltament on/off switch 65. That is, in FIG. 8, when the voltament on/off switch 65 is turned off, the output signal of the inverter 90 becomes 1.
'', and this signal ``1'' is supplied to the 6 input selection terminal SA of the selector 55 via the OR circuit 89.As a result, the selector 55 selects the input terminal, so that the sound of the pressed key on the keyboard section is Frequency information memory 63 corresponding to high
Frequency information value F supplied from Voltament 1tlJIj unit 64 changes frequency information number 1i11F as it is
', which causes the sound system 92
A musical tone is generated at the same pitch as the pressed key.

2. Voltament movement When performing voltament movement in the electronic musical instrument shown in FIG. 7, the voltament on/off switch 65 is turned on (closed). As a result, the inverter 9 in FIG.
The signal input from 0 to the OR circuit 89 is always 110''.

In this state, when a new key is suppressed on the keyboard section 61, the key-on pulse KONP="1" is output from the differentiator circuit 11 in response to this new key depression (second key depression).
FF52 and 59 are reset and their output Q is
o''. As a result, the selector 54 enters the B input selection state, and as the output signal of the AND circuit 81 becomes 0'', the output signal of the OR circuit 89 also becomes 0'', and the selector 55 also enters the 8 input selection state. becomes. Therefore, the added value Ff'' of the adder 32 of the f-curve generating section 3 is input to the delay circuit 56 via the selectors 54 and 55.

At the same time, the key-on pulse KONP-"1" is supplied to the latch circuit 1A, so that the latch circuit 1A outputs the changed frequency information value F' from the delay circuit 56 at the rising edge of this pulse KONP (this is the same as the previous value). Frequency information value F corresponding to the pitch of the pressed key (first press #I)
) is latched as the initial value F'old. Further, when the key-on pulse KONP-"1" is generated, the AND circuit 78 becomes operable and the initial change value Δ is supplied to the A input terminal of the adder 77. At this time,
The change value (H) supplied to the 8 input terminals of the adder 77 is because the difference signal Df of the subtracter 30 is rOJ (because the current value F' input to the subtracter 30 and the initial value 1i11F'old are equal) , rOJ.As a result, the initial change value Δ is output as is from the adder 77 as the change @df', and is supplied to the B input terminal of the adder 32 via the gate 80.The adder 32 Current value F' and change @cJf'
(=Δ) and the added value Ff"("F'+Δ"
or "F'-Δ") is supplied to the delay circuit 56 via selectors 54 and 55.

Therefore, the current value (changed frequency information numerical value) 1' output from the delay circuit 56 first changes by the initial change value ΔD with the generation of the key-on pulse KONP. In this case, the sign of the initial change value Δ is controlled by the magnitude discrimination signal AGB from the comparator 33A, so that the pitch of the second pressed key is the same as the pitch of the first pressed key.
When the pitch is higher than the pitch of the pressed key (when AGB is ``1''), it is set to positive, and when it is lower than the pitch of the pressed key (when AGB is ``0''), it is set to negative.

In this way, the current value F output from the delay circuit 56
When ' changes by the initial change value Δ, the difference signal D' output from the Me device 30 becomes the value Δ.

This difference signal Of is multiplied by 24 by the shift circuit 31A to become a change value df, which is supplied to the B input terminal of the adder 71. At this time, the key-on pulse KONP is always O'' and the AND circuit 78 is inactive, so the initial change value Δ is not input to the input terminal of the adder 77. From the vessel 77, the change is 1i11d
f is output as is and is passed through the gate 80 to the adder 32.
supplied to As a result, the added value Ff“ of the adder 32 is
changes by the change value df with respect to the current value F'. Along with this, the current value F' output from the delay circuit 56 also changes. Similarly, by repeating this operation according to the clock pulse φ,
The current value, that is, the number of changed frequency information 111F' outputted from the delay circuit 56, first changes exponentially as in the embodiment shown in FIG. 5 described above.

In this case, when the current IF' sequentially changes exponentially as described above, the magnitude relationship between the change values df and dg is reversed at a certain point, as explained in the embodiment of FIG. Then, the magnitude determination signal AGB2 of the comparator 50 becomes 1''.
to l Q II, or from 0" to "1", and the change detection pulse CHP of II I 11 changes from the differentiating circuit 51.
I is output.

As a result, the FF 52 is set, the selector 54 enters the six input selection state, and the addition value F from the 0 curve generation section 4 is set.
Select and output “o”.

The adder 42 of the 0 curve generating section 4 receives the change amount 11do of the shift circuit 41A input through the gate 81 and the current value F.
′ is added, so the added value F(1″ changes sequentially in a logarithmic manner as shown in equation (6) above. As a result, the current output from the delay circuit 56 11
(Change frequency information value) F' is change detection pulse CHP1
- From the time when "1" is generated, the value changes sequentially toward the target value F in a logarithmic manner.

In this way, the modified frequency information value F' output from the delay circuit 56 sequentially changes in an S-shape from the initial value F'01 (l) toward the target value F, and accordingly, the sound system 69 The pitch of the musical tone generated from (FIG. 7) changes in an S-shape from the pitch of the previously pressed first key to the pitch of the second pressed key at the end, resulting in a voltament effect.

In the embodiment of FIG. 8, the coincidence signal EQ of the comparator 33A is input to the AND circuit 18 via the inverter 79 for the following reason. That is, when a certain key is pressed, the modified frequency information value F' output from the delay circuit 56 changes sequentially as described above, and finally becomes the frequency information number (IF) of the key. In this state, if the key is pressed once and immediately pressed again, the changed frequency information value F' output from the delay circuit 56 will always be the frequency information value of the key. Since it is F, there is no need to change the numerical value F'.

However, with this key press again, the key-on pulse KON
P-"1" is generated, which causes FF52, 59
is set and the selectors 54 and 55 enter the B input selection state, so at this time the key-on pulse KO is activated as described above.
Due to the occurrence of NP, when the initial change value Δ is output as a change value (H') from the adder 77, the next simple F n becomes Δ
As a result, the changed frequency information value F' momentarily changes by the initial change value Δ. Therefore, in order to prevent this, the coincidence signal EQ of the comparator 33A is applied to the AND circuit 18 via the inverter 19, and when the changed frequency information value F' is equal to the frequency information value F (as described above, When the key is pressed again), the AND circuit 78 is disabled so that the initial change value Δ is not outputted from the adder 71.

In the embodiment shown in FIG. 8, the raw sound generation signal ENVE for opening and closing the gate 80.81 is converted into envelope waveform data ENVE in the envelope waveform generation circuit 69 (FIG. 7).
This indicates whether or not V is occurring, and the corresponding data EN
Gate 8 only while V is occurring and the musical tone is being sounded.
Open 0.81, add voltament effect, and after releasing the key,
When the sound generation is finished, gates 80 and 81 are closed and adder 3
2. The input data to the 8 input terminals of 42 are both set to fOJ, and the voltament is stopped. As a result, when a new key is pressed, the voltament starts from the pitch at which the sound of the previous key press ended. In addition, this gate 80.81 is
In particular, when the key is pressed and immediately released, the changed frequency information value F'
This is provided in consideration of a case where the sound generation ends before the frequency information value F is reached.

Further, the embodiment shown in FIG. 8 can also be modified in various ways as described for the embodiments shown in FIGS. 5 and 6.

By the way, the voltament tiIa section 64 shown in FIG.
A logic circuit 84 as shown in FIG. 9 is connected between the output terminal Q of the FF 52 and the six input selection terminal SA of the selector 54.
By connecting the 1 curve shown by f(x) in FIG. 4, the 0 curve shown by 9(x), and the S
A desired one can be selected from among the curves.

That is, in FIG. 9, when the Q input terminal is set to 0'', the 6-input selection terminal SA of the selector 54 always becomes 11011, and in this case, the addition 11Ff'' of the 1-curve generating section 3 is always output. Also, the f input terminal is 11 Q II , Q
When the input terminal is set to "1", the six input selection terminal SA of the selector 54 always becomes "P1", and in this case, the added value FO" of the 0 curve generating section 4 is always output. Roughly speaking, when both the f and Q input terminals are set to "1", the output Q of the FF 52 is supplied to the six input selection terminal SA of the selector 54, so that an S-shaped curve is output as described above.

In this case, when one curve is selected, one curve occurs i!
Since the additional value Ff" of ls3 tries to change further even after reaching the target value F, it is necessary to stop the change in the current value F' when the current value F' reaches the target value F. The comparator 33A, the differentiating circuit 58, the FF 59, and the selector 55 also have the function of stopping this operation.In addition, in this case, the switch 65 and 9 are not omitted.
The logical sum of the signals at the input terminals may be input to the OR circuit 89.

(Scope of Application and Modifications of the Present Invention) The function waveform generator according to the present invention is not limited to the above-mentioned voltament, but can also be used for pitch control of other musical tones such as attack pitch or glide. for example,
In the case of the attack pitch system, in the embodiment shown in FIG. 5, the target value TV is set to [1] and the initial value S
For example, set the start pulse S to [0.8J, and then set the start pulse S
The key-on pulse KONP explained in FIG. 8 is used as P, and the current icv is changed in an S-shape from "0.8" to 1'IJ by the generation of the key-on pulse KONP. Then, this current mcv may be multiplied by the frequency information value F outputted from the frequency information memory 63 in FIG. 7 and then supplied to the accumulator 66. In this way, when producing a musical tone corresponding to the pitch of the pressed key, it is possible to obtain an attack pitch effect by continuously changing the pitch of the musical tone from a desired pitch amount lower to the original reference pitch. can. Further, the function waveform generator according to the present invention can also be used to control other musical tone parameters such as timbre, volume, control waveform (envelope waveform), etc. Further, in the function waveform generator according to the present invention, for example, by exchanging the target value and the initial value each time the current value reaches the target value, it is possible to obtain a periodic function waveform that repeatedly changes smoothly.

Further, in the above-described embodiment, an example was shown in which the present invention was applied to a single-tone electronic musical instrument, but this invention employs a known digital time-division processing method to produce a multiple-tone electronic musical instrument that simultaneously produces multiple tones. It can also be applied to musical instruments.

(Effects of the Invention) As described above, according to the present invention, a function waveform in which the value changes sequentially in an S-shape from an initial value to a target value has a gradual change at the beginning and end and a larger change in the middle. This can be generated by a simple configuration using a digital circuit. Also,
By using this function waveform to change the pitch of a musical tone, for example, it is possible to realize natural pitch changes such as voltament, attack pitch, and glide.

[Brief explanation of drawings]

1 and 2 are graphs showing pitch changes in a conventional electronic musical instrument, FIG. 3 is a diagram for explaining the present invention in detail, and FIG. 4 is a graph showing pitch changes generated by the function waveform generator of the present invention. FIG. 5 is a block diagram of a function waveform generator according to an embodiment of the present invention; FIG. 6 is a block diagram showing a modification of the device in FIG. 5; FIG. FIG. 8 is a circuit diagram showing a specific example of the voltament control shown in FIG. 7. FIG. 9 is a circuit diagram showing a specific example of the voltament control shown in FIG. 8. FIG. 7 is a partial circuit diagram showing a modification of the control section. 1... Initial value setting circuit, 2... Target value setting circuit, 3
...1 curve generation part, 4...0 curve generation part, 5...
- Switchable m section, 30, 40...subtractor, 31.41...weighting circuit, 32, 42...adder.

Claims (1)

[Scope of Claims] 1. A function waveform generator that generates an S-shaped function waveform whose value changes gradually from the initial value to the target value at the beginning and end and more rapidly at the middle, Forming a first change value corresponding to the difference between the current value of the waveform and a predetermined first reference value and smaller than the eye difference, and converting this first change value into the above-mentioned current value at a predetermined timing. A first mode in which the next current value is formed and output by adding or subtracting with a second mode for forming a second change value and adding or subtracting the second change value from the current value at a predetermined timing to form and output the next current value; , a function waveform generator for an electronic musical instrument, characterized in that the calculation means is initially controlled to perform calculation in a first mode, and is controlled to switch so that calculation is performed in a second mode midway through. Device. 2. The electronic musical instrument function according to claim 1, wherein the control means compares the first variation value and the second variation value and performs the switching control according to the comparison output. Waveform generator. 3. The control means sets the initial value and the target value as the first and second reference values in the calculation means, respectively, and first outputs the initial value as the current value of the function waveform; Next, the calculation in the first mode is started by calculating the current value and a predetermined initial change value and outputting the calculation result as the next current value. 2. The function waveform generator for electronic musical instruments according to item 2. 4. The control means includes means for setting an arbitrary initial value and a target value, and determines the magnitude relationship between the target value and the initial value and adds or calculates the current value and the initial change value. 4. A function waveform generator for an electronic musical instrument according to claim 3, which instructs one of the subtractions. 5゜The control means includes means into which arbitrary initial values and target values are input, and adds to the initial values a positive or negative initial change value according to the magnitude relationship between these target values and the initial values. to form the first reference value.
Or the function waveform generator for electronic musical instruments according to item 2. 6. The control means sets a frequency information value corresponding to the pitch of the previously pressed key as the initial value, and sets a frequency information value corresponding to the pitch of the newly pressed key as the target value. 6. The function waveform generator for an electronic musical instrument according to claim 4, wherein the pitch of the generated musical tone is controlled by the current value or the next current value.