JP3223683B2 - Musical sound signal synthesizer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tone waveform signal synthesizer for oscillating and synthesizing tone waveform signals by using a closed waveguide network, and more particularly to simplification and synthesis of control for tone waveform signal oscillation. The present invention relates to a method that can achieve both the enrichment of harmonic components in a given sound.

[0002]

2. Description of the Related Art Japanese Patent Laying-Open No. 63-40199 discloses a basic configuration of a technique for oscillating and synthesizing a musical tone waveform signal using a closed digital waveguide network. This is a signal circulation path (closed digital waveguide network) that circulates digital waveform signals by connecting delay circuits and filters in a closed loop.
And a digital excitation signal is introduced into the circulation path to circulate the circulation path, thereby oscillating a waveform signal and extracting an output tone waveform signal from an appropriate point in the loop. The basic concept of this technique is to model the physical characteristics of a desired natural musical instrument such as a wind instrument or a stringed instrument using a closed digital waveguide network and thereby simulate the sound of the natural musical instrument. That is, the signal circulation path (closed digital waveguide network) models the physical propagation of a vibration signal traveling or reflecting in a medium such as a pipe or a string.

In the case of simulating a wind instrument, the signal circulation path corresponds to the wind section of the wind instrument, and the resonance characteristics of the wind section are set by the signal delay time there. Also, when simulating a stringed instrument,
The signal circulation path corresponds to a string portion. The signal delay time in the signal circulation path is variable according to the parameter, whereby the resonance characteristics in the signal circulation path can be controlled. The pitch of the synthesized tone waveform signal is set or controlled mainly in accordance with the signal delay time, but can be set or controlled by other factors (for example, parameters in a nonlinear circuit for exciting the signal). It is.

[0004] The signal circulation path models the propagation and reflection of a signal at a physical boundary portion (for example, a vibration excitation section such as a lead, an opening of a tube, or an end of a string) in a propagation path of a vibration signal. And a signal junction section for performing the operation. For example, a signal junction unit for modeling a vibration excitation unit includes a non-linear conversion unit. The signal junction for modeling the other physical boundary includes an arithmetic circuit that sorts and combines the traveling wave signal and the reflected wave signal.

A non-linear table is used in the non-linear converter for exciting a vibration signal. By introducing an appropriate electric pressure signal corresponding to a breath pressure or a bowed string operation into the loop of the signal circulation path, a signal obtained by nonlinearly converting the pressure signal by the nonlinear table circulates in the signal circulation path, and vibrates. The waveform signal is excited. JP-A-63-4
In Japanese Patent Publication No. 0199, two types of nonlinear tables are shown. One is a non-linear table having a non-change area and a change area, and in the change area, shows a conversion output characteristic that monotonically changes only in one direction with respect to a change in input. It is suitable for simulating a vibration waveform signal of a wind instrument having The other is a non-linear table showing a conversion output characteristic having a chevron shape (having a bidirectional slope) with respect to a change in input in a change region, and simulates a vibration waveform signal of a bowed instrument such as a violin. Suitable for running.

[0006]

It is known that, in the case of using the former non-linear table showing the unidirectional monotone change conversion characteristic, it is possible to oscillate a vibration waveform signal composed of odd harmonics with even harmonics omitted. I have. Also,
It has been confirmed that the oscillation region for the control parameter is wide, stable oscillation is possible, and that the signal delay time in the loop and the oscillation pitch correspond well, and that the control is relatively easy. On the other hand, even harmonics are omitted as described above, so that only sounds close to a specific instrument such as a clarinet or a flute can be synthesized, and there is a problem that the width of sound creation is narrow.

On the other hand, it is known that the use of the latter non-linear table showing the chevron-shaped conversion characteristic can oscillate a vibration waveform signal having a very complicated overtone structure. In this way, it is possible to oscillate a sound with harmonic characteristics in which all the integer harmonics are arranged neatly, and in particular, if you give a parameter that always oscillates using the conversion characteristic near the top of the mountain, A signal close to white noise can be oscillated. As described above, since a complex overtone structure can be obtained, there is a possibility that a sound having various characteristics can be oscillated. However, as a practical problem, a parameter region for causing the oscillation is narrow, and the adjustment thereof is very difficult. There is a problem that. Furthermore, the signal delay time in the loop and the oscillation pitch do not correspond very well,
There was a problem in control or handling such as easy tuning.

SUMMARY OF THE INVENTION The present invention has been made in view of the above points, and is directed to a musical sound waveform signal synthesizing apparatus which oscillates and synthesizes a musical sound waveform signal by modeling a physical resonance characteristic of a sound source like a closed waveguide network. It is an object of the present invention to provide a musical tone waveform signal synthesizing apparatus capable of synthesizing a musical tone waveform signal including a complicated harmonic with easy control.

[0009]

Means for Solving the Problems A musical tone signal synthesizing apparatus according to the present invention forms a closed loop for circulating a signal and includes delay means for delaying a signal in the loop.
A signal circulating unit whose resonance characteristics are controlled by varying the delay time, an exciting unit for exciting a waveform signal in a loop of the signal circulating unit, and a waveform signal extracted from the loop of the signal circulating unit. Waveform deforming means for deforming the waveform and adding harmonic components, and the waveform deforming means
The extracted waveform signal from the loop of the signal circulating means.
Calculating means for calculating with another waveform signal, an output waveform signal of the calculating means based on a waveform signal extracted from a loop of the signal circulating means and a waveform signal deformed by the waveform deforming means.
Is characterized in that comprises an output means for outputting the issue.

[0010]

The signal circulating means and the excitation means constitute an electronic closed-loop sound source similar or similar to the physical model type sound source using the closed-type waveguide network described above, and the physical resonance of the desired physical sound source is achieved. Properties are modeled. Therefore, in the electronic closed-loop sound source section composed of the signal circulating means and the excitation means, a tone waveform signal simulating the modeled physical sound source is synthesized. The musical tone waveform signal synthesized by the loop of the signal circulating means is taken out of the loop and input to the waveform deforming means, and the waveform of the waveform signal is deformed so that harmonic components are added. With this deformation, it is possible to enrich the overtone structure contained in the waveform signal. Further, the waveform signal deformed by the waveform deforming means is
Signal from another loop of the signal circulation means.
Waveform signal obtained as a result of the calculation
Can be changed in a complicated manner. By outputting the tone waveform signal extracted from the loop of the signal circulating means and the output waveform signal of the arithmetic means based on the waveform signal deformed by the waveform deforming means, and appropriately combining the two waveform signals electrically or spatially. A musical tone obtained by synthesizing both waveform signals is obtained.

The closed loop sound source portion composed of the signal circulating means and the excitation means makes it possible to create a sound peculiar to the above-described physical model type sound source circuit, while using the waveform deforming means .
Overtone components that do not occur in the closed loop sound source
By performing such a waveform deformation, it is possible to abundantly increase the harmonic composition of the waveform signal synthesized in the closed loop sound source portion. Therefore, in the sound generation in the closed loop sound source part, even if a method in which stable oscillation is possible and the control is relatively easy is adopted, a waveform deformation for creating a harmonic component is performed in the waveform deformation part. By combining the two ,
A variety of harmonics can be controlled in a musical tone signal, thereby expanding the range of sound creation. Therefore,
Although the control in the closed loop sound source portion is relatively easy, a tone waveform signal including complicated overtones can be synthesized in the finally obtained tone.

For example, in the closed-loop sound source portion, if the vibration signal is excited by using the above-described non-linear table showing the one-way monotonic change conversion characteristic, stable oscillation is possible, and the control can be compared. It is easy. Also,
If the waveform deforming means includes a differentiating means for differentiating the waveform signal extracted from the loop of the signal circulating means, for example, a square wave input waveform signal composed of odd overtones with even-numbered overtones may be sawtooth-shaped. And an even harmonic component can be added, and the configuration is extremely simple. Also,
In the waveform deforming means, an appropriate partial waveform processing technique such as inversion or cutting of the polarity of a partial phase section of the waveform may be used.

As another example of the waveform deforming means,
It may include a non-linear conversion table, and perform waveform deformation by inputting the waveform signal extracted from the loop of the signal circulating means to the non-linear conversion table and performing non-linear conversion. The shape of the waveform can be greatly deformed by the non-linear conversion, and it is possible to add a harmonic component that was not present in the input waveform. In this case, the amplitude level of the waveform signal passing through the loop of the signal circulating means may be detected, and the amplitude of the waveform signal input to the non-linear conversion table may be controlled according to the detected amplitude level. When the level of the waveform signal taken out of the loop of the signal circulating means fluctuates in response to the fluctuation of the volume envelope level such as attack, decay, etc., the fluctuation of the input waveform level changes the form of the waveform deformation in the non-linear conversion table. Therefore, in order to prevent this, for example, it is preferable to control so that the smaller the detected amplitude level is, the larger the amplitude of the waveform signal input to the nonlinear conversion table is.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the accompanying drawings. In the embodiment of FIG. 1, a physical model sound source of a wind instrument such as a clarinet is constructed according to the same principle as that of a closed waveguide network. In this embodiment, the delay feedback loop unit 10 corresponding to the closed waveguide network is used.
Has a simplified configuration.

In FIG. 1, a delay feedback loop section 10 is formed by forming a closed loop for delaying and circulating a digital waveform signal, and includes a delay circuit 11 for delaying the waveform signal in the loop. By varying the time with the control parameter DC, the resonance characteristics of the loop can be controlled, and the pitch of the synthesized musical tone can be set. The delay feedback loop unit 10 includes a low-pass filter 12 and an all-pass filter 13 in the loop, and further includes an excitation unit 14 having a non-linear characteristic inserted in the loop to excite vibration. In the loop, an adder 15 and an inversion circuit 16 for inverting the polarity are provided on the input side of the excitation unit 14.

The adder 1 provided on the input side of the excitation section 14
Numeral 5 is for introducing a pressure signal P, which is one of the performance parameters, into the loop. Digital data corresponding to the pressure signal P is input to the adder 15 via the adder 17, and a signal obtained by inverting the polarity of the signal circulating in the delay feedback loop unit 10 by the inverting circuit 16 is added thereto. Are input to the non-linear conversion tables 21 and 22. That is, the output signal of the excitation unit 14 is processed by the low-pass filter 12 and the all-pass filter 13,
The signal is delayed by the delay circuit 11 and is negatively fed back to the input side via the inverting circuit 16 and the adder 15. This inverting circuit 16
Acts to excite vibration in the loop section 10 in cooperation with the non-linear conversion tables 21 and 22 as described later, and is regarded as a part of the non-linear conversion element inserted in the loop. Is also good. When viewed from the side of the physical model, the simulation simulates reflection (phase inversion) at the opening end of the clarinet.

The pressure signal P is composed of DC (not a vibration waveform) digital data for exciting a vibration signal, and is continuously output from the performance parameter generation section 20 in accordance with an arbitrary sounding duration. Generated. The performance parameter generating section 20 generates various performance parameter data for controlling various characteristics of a musical tone to be synthesized in real time, and generates each performance parameter data based on a player's intention and other factors. . For example, it may include various performance operators such as selection switches for pitch, timbre, etc., a breath controller, and various data generating means such as an envelope generator. For example, when a player blows a breath controller, a pressure signal P is generated corresponding to the breath pressure. The noise generator 18, filter 19 and adder 17 are for modulating the pressure signal P as desired. By adding the filtered noise signal to the pressure signal P, a fluctuation is given to the pressure signal P to make the tone synthesis form more complicated.

The excitation unit 14 has two non-linear conversion tables 21 and 22 in parallel, and further outputs variable coefficients C 1 and C 2 to the outputs of the non-linear conversion tables 21 and 22.
It has multipliers 23 and 24 for multiplying by two, respectively, and an adder 25 for adding the outputs of the multipliers 23 and 24. Two nonlinear conversion tables 21 and 22 (table #
Examples of the non-linear conversion characteristics in (1, # 2) are as shown in (a) and (b) of FIG. As shown in FIG. 2A, the non-linear conversion characteristic of one non-linear conversion table 21 (table # 1) has a change region and a non-change region, and in the change region, a conversion curve showing a monotonous change in one direction. Consists of The other non-linear conversion table 22 (table #
As shown in FIG. 2B, the nonlinear conversion characteristic of 2)
A certain range in which the input value is small is an invariable region, and the change region is composed of a conversion curve that mostly shows an increasing change but partially shows a decreasing change.

The non-linear conversion characteristics as viewed from the entire excitation unit 14 are characteristics obtained by adding or subtracting the conversion characteristics of the two non-linear conversion tables 21 and 22, and examples thereof are shown in FIGS. It looks like). (C) is (a)
(B) shows an example in which two conversion characteristics are added and synthesized, and (d) shows an example in which subtraction synthesis is performed. (C) or (d)
The curve of the nonlinear conversion characteristic to be synthesized as described above can be appropriately variably controlled by the magnitude and sign of the values of the coefficients C1 and C2 given to the multipliers 23 and 24. For example, when the sign of the coefficient C1 is given as positive and the coefficient C2 is given as negative, a subtraction / synthesis characteristic as shown in FIG. In the combined non-linear conversion characteristic, a small “bump” or “dent” appears partially as shown in the figure. It is advantageous that a nonlinear conversion similar to a conventionally known mountain-shaped nonlinear conversion characteristic is performed in the "kump" or "dent" portion, thereby complicating the generation of harmonic components. In addition, there is an advantage that a relatively stable oscillation can be ensured and the control is relatively easy since an extreme reversal of the inclination does not occur in the "lumps" and "dents". In addition, there is an advantage that the combined nonlinear conversion characteristic can be varied as needed by varying the coefficients C1 and C2.

The coefficients C1 and C2 are, for example, manually from the performance parameter generating section 20 in response to the operation of an appropriate operation element, or automatically in conjunction with a selected timbre, a selected pitch or a selected effect. It can be generated and varied. Of course, in practicing the present invention, the excitation unit 1
4 is not limited to the type that combines two nonlinear conversion characteristics as shown in the figure, and may be configured with only one nonlinear conversion table. Alternatively, three or more different nonlinear conversion characteristics may be combined.

The low-pass filter 12 provided in the closed loop simulates a high-range loss at a reflection end of a wind instrument (for example, an opening end of a clarinet) and a high-range loss at the entire tube. Accordingly, the filter characteristics such as the cutoff frequency are variably controlled. Thereby, the tone color of the musical tone to be synthesized can be adjusted. The filter coefficient group FC1 is also generated, for example, manually from the performance parameter generation unit 20 in response to an operation of an appropriate operation element, or automatically in conjunction with a selected timbre, a selected pitch or a selection effect. And it may be made to be variable.

The all-pass filter 13 provided in the closed loop does not control the amplitude of the passing signal, but varies the amount of phase delay corresponding to the frequency. The all-pass filter 13 physically simulates that the traveling speed of a sound wave propagating in the wind instrument wind varies depending on the pitch frequency. The phase delay versus frequency characteristic of the all-pass filter 13 is controlled by the filter coefficients FC2 and / or FC3.
The filter coefficient FC2 is generated from the performance parameter generating section 20 in accordance with an appropriate factor such as a selected timbre, thereby variably controlling the phase delay versus frequency characteristic in the all-pass filter 13 according to the selected timbre and the like.
It is possible to finely adjust the tone of the synthesized musical tone. The filter coefficient FC3 converts the waveform signal passing through the loop into an amplifier 26.
, And a signal whose level is appropriately adjusted. That is, the filter coefficient FC3 substantially corresponds to the amplitude level of the oscillation waveform signal circulating in the loop. The amplitude level of the oscillation waveform signal fluctuates in accordance with the level of the breath pressure, and the characteristics of the all-pass filter 13 are variably controlled in accordance with the corresponding filter coefficient FC3. Can be adjusted.

The oscillating action of the delay feedback loop unit 10 having the above configuration is generally as follows. In response to the rise of the pressure signal P, the excitation unit 14 outputs a signal obtained by nonlinearly converting the pressure signal P. This signal is supplied to a low-pass filter 12, an all-pass filter 13, and a delay circuit 11.
After the total delay time of the loop has passed, the signal is inverted and fed back to the adder 15 via the inverting circuit 16. Therefore, the adder 15 subtracts the absolute value of the feedback signal from the pressure signal P. Then, the output of the adder 15 is output to the non-linear conversion tables 21 and
2, the signal is nonlinearly converted, inverted through a loop after a predetermined delay time, and fed back to the adder 15. This repetition in the loop will excite vibration. Therefore, in the case of the sound source model shown in FIG. 1, one cycle of the pitch of the oscillated waveform signal is twice the total delay time of the loop.

The control parameter DC for variably setting the delay time of the delay circuit 11 is generated from the performance parameter generator 20 in accordance with the desired pitch. As described above, the pitch of the oscillated waveform signal is determined not only by the delay time of the delay circuit 11 but by the delay time of the entire loop, so that the delay time set by the control parameter DC is , The delay time in the filters 12 and 13 must be subtracted.

The delay feedback loop unit 10 shown in FIG. 1 is a simple clarinet model, and the timbre characteristic of the waveform signal synthesized there is clarinet-like, and lacks even-order harmonics. Accordingly, the nonlinear conversion characteristics in the excitation section 14 are relatively simple, the oscillation is stable, and the control thereof is relatively easy. FIG. 3A shows an example of a waveform signal synthesized by the delay feedback loop unit 10. It can be understood that the shape is close to the shape of the rectangular wave, and the even harmonics are insufficient.

The oscillating waveform signal is extracted from an arbitrary position (the output of the adder 25 in the figure) in the delay feedback loop section 10 and input to the waveform deforming section 27. The waveform deforming unit 27 performs a process of deforming the waveform of the waveform signal extracted from the loop and adding a harmonic component. In FIG. 1, the waveform deforming unit 27 includes a differentiating circuit 28, an absolute value circuit 29, and a comparator 30. Differentiating circuit 28
Is for differentiating the digital waveform signal (for example, a in FIG. 3) synthesized by the delay feedback loop unit 10. Since the digital waveform signal is a digital process, the difference operation is actually performed. An example of the output of the differentiating circuit 28 is as shown in FIG. The absolute value circuit 29 converts the amplitude value of the output of the differentiating circuit 28 into an absolute value. That is, a negative value is converted into a positive value. An example of the output of the absolute value circuit 29 is as shown in FIG. The comparator 30 compares the output of the differentiating circuit 28 with 0, and outputs a value of 0 or more, that is, a positive value. An example of the output of the comparator 30 is as shown in FIG.

The outputs of the absolute value circuit 29 and the output of the comparator 30 are respectively converted into individual coefficients C3 and C3 by multipliers 31 and 32, respectively.
The level is controlled by multiplying by four, and these are added and synthesized by the adder 33. Both the coefficients C3 and C4 can be changed to either positive or negative values, or at least the coefficient C4 has a variable negative value. That is, as is apparent from FIG. 3C, the output of the absolute value circuit 29 has a frequency (one octave higher) twice the original pitch of the tone waveform signal synthesized by the delay feedback loop unit 10. Therefore, not only can the mode of the waveform deformation in the waveform deforming section 27 be controlled by the variable control of the coefficients C3 and C4, but also the frequency can be set to 1
You can change it up an octave.

For example, the coefficient C4 of 0 is output to the output of the comparator 30.
Is added to the output of the absolute value circuit 29, the output of the absolute value circuit 29 is directly output from the adder 33, and the frequency of the waveform signal whose waveform has been deformed by the waveform deforming section 27 is twice the original tone pitch. Becomes Also, if the output of the comparator 30 is multiplied by a coefficient C4 of -1 and added to the output of the absolute value circuit 29, the sawtooth waveform in FIG. 3C is removed every other wave. The output waveform of 33, that is, the frequency of the deformed waveform signal is the same as the original tone pitch. When the coefficient C3 or C4 is appropriately changed, the output of the absolute value circuit 29 and the output of the comparator 30 are added / combined (or subtracted / combined) at a ratio corresponding to the ratio, and the form of the waveform deformation in the waveform deforming unit 27 is controlled. it can. FIG. 3 (c)
Or, as is apparent from (d), it can be understood that the waveform deformed by the differential operation is close to a sawtooth wave and includes even-order harmonics.

The output of the waveform deforming section 27 is input to a high-pass filter (HPF) 34. This high pass filter 3
Numeral 4 is for removing the DC component of the waveform signal output from the waveform deforming section 27. That is, since the output of the waveform deforming unit 27 is offset to one polarity,
Here, the DC component is removed. High pass filter 34
The low-pass filter (LPF) 35 to which the output of
This is for extracting a low-band component from the output signal of the high-pass filter 34. The level of the low-band component signal extracted by the low-pass filter 35 is appropriately adjusted by an amplifier (multiplier) 36, and is added to the output signal of the high-pass filter 34 by an adder 37. The addition of the low-band component by the low-pass filter 35 and the adder 37 is for enhancing the level of the low-band component with respect to the waveform signal subjected to the waveform modification processing by the waveform modification unit 27. The reason is that the low-frequency component originally included in the loop output waveform is attenuated by the differentiation in the differentiating circuit 28, so that the low-frequency component is augmented.
Further, it is conceivable that the low-band component is somewhat attenuated by the DC component removal processing by the high-pass filter 34,
There is also a meaning to reinforce this.

The output of the adder 37 is appropriately level-controlled by a multiplier 38 and input to an adder 40. On the other hand, the waveform signal extracted from the delay feedback loop unit 10 is
After being input to the high-pass filter 41 to remove the DC component, the level is appropriately controlled by the multiplier 39 and input to the adder 40. These multipliers 38 and 39 and adder 40
Is for adding and combining the waveform signal extracted from the delay feedback loop unit 10 and the waveform signal deformed by the waveform deforming unit 27 at a variable ratio according to the multiplication coefficients C5 and C6. As a result, the waveform deforming unit 2 applies the tone waveform signal synthesized by the delay feedback loop
In step 7, the waveform signal subjected to the waveform deformation for adding the overtone component is synthesized at an appropriate ratio, and as a result of the synthesis, a tone waveform signal rich in the overtone component is synthesized. The output of the adder 40 is appropriately tone-adjusted by the filter 42, then converted into an analog signal by a digital / analog conversion circuit (not shown), and acoustically generated. The filter 42 is merely provided for performing further tone color adjustment, and can of course be omitted.

As described above, the delay feedback loop section 10 of FIG. 1 is a clarinet model, and the tone waveform signal generated therefrom has a strong characteristic of a clarinet tone color composed of odd harmonics. On the contrary,
As described above, the waveform deforming unit 27 obtains a waveform signal to which an even-order harmonic is added. If the coefficients C3 and C4 are set so that the output of the absolute value circuit 29 is output as it is, it is possible to have a frequency one octave higher (second harmonic). Therefore, the waveform signal deformed by the waveform deforming unit 27 is added to the waveform signal synthesized by the delay feedback loop unit 10, so that the lacking even-order harmonics of the waveform signal synthesized by the delay feedback loop unit 10 are added. A component can be added and its tone can be changed dramatically. For example, the output waveform of the waveform deforming unit 27 to which the even-order harmonic component is added has a characteristic of a trumpet-type timbre. The addition of harmonics one octave higher can imitate organ-based tone synthesis. In addition, the setting of the coefficients C5 and C6 for setting the synthesis ratio and the setting of other various parameters enable synthesis of an arbitrarily wide range of timbres.

As described above, the coefficients C3, C4, C5, and C6 are obtained from the performance parameter generator 20, for example, manually or in response to the operation of an appropriate operation element, or in the selected tone color, selected pitch, or selected tone. Automatically generated in conjunction with effects, etc.,
And it may be made to be variable. The mixing ratio of the waveform signal deformed by the waveform deforming section 27 is set by the coefficient C
3 and C4, the multiplier 38 and the corresponding coefficient C5 may be omitted. Further, in the circuit in which the coefficients are not shown in the figure, all the circuits which can be controlled by the coefficients (for example, the amplifiers 26 and 36, the HPF, the LPF, the filter 42, etc.) are appropriately calculated by the operation parameter generation unit 20. , And its characteristics can be variably controlled in real time during performance.

The configuration of the delay feedback loop unit 10 itself is not limited to the configuration shown in the figure, but may be a known digital wave loop disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 63-40199 or other patent application publications subsequent thereto. A guide network type sound source may be used as appropriate.

FIG. 4 shows an embodiment in which the configuration of the excitation section 14 in FIG. 1 is changed. Other configurations are almost the same as those shown in FIG. 1. For the circuit elements having the same reference numerals as those shown in FIG. 1, the description in FIG. The excitation unit 43 in FIG. 4 does not have a non-linear conversion table, and is provided with an excitation waveform generation unit 44 instead. The excitation waveform generator 44 generates an appropriate initial waveform in accordance with the pitch data KC, the sound generation instruction data KON, the touch control data TD, etc., which are given from the performance parameter generator 20 in response to the performance operation. When the generated initial waveform signal is input to the adder 45, it is introduced into the loop of the delay feedback loop unit 10. This initial waveform signal is delayed in the loop via the low-pass filter 12, the all-pass filter 13, and the delay circuit 11, is inverted in polarity by the multiplier 46, is level-controlled, and is fed back to the adder 45.
The coefficient “−G” of the multiplier 46 is a negative value, and is variably generated from the performance parameter generation unit 20. With this configuration, as described above, the delay feedback loop unit 1
Oscillation having a period twice as long as the total delay time of 0 is oscillated. In this case, relatively stable oscillation is possible, but the waveform signal synthesized by the loop section 10 has a characteristic of lacking even-order harmonics, as described above. This shortage can be compensated for by the waveform deforming section 27 as described above. Note that the excitation waveform generator 44 may generate an appropriate noise signal in addition to the initial waveform.

FIG. 5 shows another embodiment of the waveform deforming unit 47 different from that of FIG. 1, and the delay feedback loop unit 48 and the excitation unit 49 are also shown in FIGS.
The embodiment shown is different from that shown in FIG. The delay feedback loop section 48 and the excitation section 49 in FIG.
Is a physical model of a saxophone, which is a more complicated model than those shown in FIGS. However, since the configuration is similar to a known configuration as disclosed in, for example, Japanese Patent Application Laid-Open No. 2-294692, the description will be simplified. In addition,
Various performance parameters such as the pressure signal PRES are generated by an appropriate performance parameter generator as described above, but are not shown.

The delay feedback loop unit 48 simulates the operation of the saxophone tube portion, and includes a delay circuit 51L having a relatively long delay time and a delay circuit 51S having a relatively short delay time. Delay circuit 5
1L , low-pass filter 52, multiplier 53, adder 5
4, delay circuit 51S, low-pass filter 55, multiplier 5
6 constitute one loop. Then, the output signal of the excitation section 49 given via the multiplier 57 is input to the adders 50 and 54 and introduced into the loop, and the amplitude is controlled by the coefficients in the multipliers 53 and 56 in the loop, respectively. The signals are added by the adder 58 and the output is fed back to the excitation unit 49, and the excitation unit 49 is incorporated in a loop. The output of the musical tone waveform signal synthesized by the delay feedback loop unit 48 is extracted from the multipliers 53 and 56, respectively, and the amplitude is controlled by the multipliers 59 and 60 by the coefficients m1 and m2, respectively.
It is obtained by adding one. Performance parameters for variably setting the delay time, the multiplication coefficient, and the filter coefficient to variably set the pitch and timbre of the synthesized musical tone are input to the delay circuit, multiplier, and filter in the delay feedback loop unit 48, respectively. , Are not shown.

The excitation section 49 simulates the operation of a saxophone lead, and has two non-linear conversion tables 62 and 63. One non-linear conversion table 62 is a table of a slit function, which simulates the physical characteristics of the displacement of the lead with respect to the pressure. The other non-linear conversion table 63 is a Graham function table, which simulates the characteristic that the differential pressure at the lead gives to the flow velocity, and is used to correct the non-linear conversion by the slit function. Things.

In the adder 64, the pressure signal PRES given as the performance parameter corresponding to the breath pressure is subtracted from the signal fed back via the adder 58 of the delay feedback loop unit 48. This adder 64
Output corresponds to the differential pressure for displacing the lead. The signal corresponding to the differential pressure output from the adder 64 is input to the low-pass filter 66 via the adder 65 and also to the nonlinear conversion table 63.

The low-pass filter 66 has a dynamic characteristic in which the lead does not respond to the high-frequency component, and is used for simulating the dynamic characteristic. Remove components. The output of the low-pass filter 66 is multiplied by a slit gain parameter SG by a multiplier 67, and an embouchure parameter EMB is added by an adder 68. Slit gain parameter SG
Controls the slope of the slit function in the nonlinear conversion table 62 (actually controls the slope of the function by controlling the gain of the input signal), and is greater than 0, that is, a positive coefficient. Embouchure parameter EM
B simulates subtle biting of the lead by making corrections in consideration of the fact that the amount of displacement of the lead is affected by the attitude of the lips, the degree of tightening, and the like. Peak parameter Q and cutoff frequency parameter f for controlling frequency characteristics of low-pass filter 66
c, the slit gain parameter SG, and the embouchure parameter EMB are provided from a performance parameter generator (not shown), as described above, and can be changed in real time during performance.

The output of the adder 68 is input to the non-linear conversion table 62 as a pressure signal that has been subjected to various necessary calculations (differential pressure calculation, high frequency cut, gain adjustment, embouchure correction, etc.). An example of the slit function in the non-linear conversion table 62 is as shown in FIG. 6A, which shows a monotonous change characteristic, and is easy to control. The output of the non-linear conversion table 62 is a multiplier 69
And the output of the non-linear conversion table 63 is multiplied. The non-linear conversion table 63 supplies a multiplier coefficient to the multiplier 69, and an example of the Graham function there is as shown in FIG. 6B. The input signal is a differential pressure signal output from the adder 64, and the output is a flow velocity coefficient. This simulates that even if the differential pressure increases, the flow velocity saturates in a narrow pipeline and the differential pressure is not proportional to the flow velocity. Function. After all, the multiplier 69 multiplies the signal representing the lead displacement (air passage area) corresponding to the pressure given from the nonlinear conversion table 62 by the flow velocity coefficient given from the nonlinear conversion table 63. The output of the multiplier 69 is multiplied by a predetermined coefficient by the multiplier 57,
It is provided to the delay feedback loop unit 48. In addition,
The output of the nonlinear conversion table 62 is input to the multiplier 70, multiplied by an appropriate coefficient, fed back to the adder 65, and added to the differential pressure signal. This is for correcting the differential pressure signal according to the lead displacement amount.

Referring to FIG. 5, the waveform transformation section 47 includes a nonlinear conversion table 71. Adder 68 of excitation section 49
Is output and input to the nonlinear conversion table 71. The non-linear conversion table 71 is composed of a complicated slit function as shown in FIG. The input signal is complicatedly deformed according to a function including several peaks (having both positive and negative slopes). The output of the non-linear conversion table 71 is input to the multiplier 72, and is multiplied by the flow velocity coefficient output from the non-linear conversion table 63. The output of the multiplier 72 is multiplied by a predetermined coefficient by a multiplier 73, further multiplied by a variable coefficient m3 for setting a mixing ratio by a multiplier 74, and input to an adder 75. The adder 75 adds the output of the adder 61, that is, the synthesized tone output signal of the delay feedback loop unit 48. Therefore, the mixing ratio of the synthesized tone output of the delay feedback loop unit 48 and the output of the waveform deforming unit 47 in the adder 75 is determined by the coefficients m1,
It is appropriately controlled by changing m2 and m3.

In the embodiment shown in FIG. 5, the characteristics of the nonlinear conversion table 62 of the excitation section 49 show a relatively monotonous change characteristic as shown in FIG. By controlling the performance signal such as the pressure signal PRES according to the breath pressure and the embouchure parameter EMB, it is possible to easily simulate, for example, a fine pitch change in an attack portion, and synthesize a good-quality saxophone sound. Can be. In addition to this, it is possible to realize more diverse / diversified timbre synthesis by adding the waveform deforming unit 47 without being limited to the saxophone sound. That is, the nonlinear conversion table 7 of the waveform deforming unit 47
Since 1 is out of the loop of the delay feedback loop unit 48, any function (a function that changes in any complicated manner) can be used as a function for the non-linear conversion. Waveform deformation can be performed, and overtone components can be added. Therefore, the tone color of the tone waveform signal output from the adder 75 can be largely changed by variably controlling the coefficients m1, m2, and m3 for setting the mixing ratio.

By the way, in the wind instrument model as shown in FIG. 5, even if the timbre is suddenly changed by the addition of the waveform deforming section 47, the impression that the user is breathing inevitably remains. This is because in this type of physical model sound source, it is difficult for oscillation to grow unless pressure signal energy is gradually applied. That is, in the physical model sound source composed of the delay feedback loop unit 48 and the excitation unit 49, the amplitude envelope of the synthesized tone signal shows a relatively slow rising like a wind instrument.
In the case where the waveform is deformed, the waveform is not sufficiently deformed because the input level of the non-linear conversion table 71 is small, and the waveform deforming effect of the waveform deforming section 47 is weakened at the rising portion, so that the sound played with breathing rises. Impression remains. The improvement measures are shown in FIG.

FIG. 7 shows a modification of the waveform deforming section 47 in FIG. 5. The delay feedback loop section 48 and the excitation section 49 (not shown) have the same configuration as that shown in FIG. Good. 7, the envelope detection circuit 76 includes a delay feedback loop unit 48.
The tone signal TS extracted from an arbitrary portion (arbitrary algorithm portion) of the loop in the excitation section 49 is input, and its amplitude envelope is detected. The level data of the detected amplitude envelope is input to the level conversion table 77. Level conversion table 77
Is a conversion characteristic that outputs data that changes substantially in inverse proportion to the input level (or at least a conversion characteristic that outputs the largest output value when the input level is within a predetermined low level range). FIG. 8A or FIG. 8B shows an example. In other words, most of the conversion characteristics are such that the smaller the level of the amplitude envelope of the tone signal TS is, the larger the value of the output is, but the more the conversion characteristics fluctuate as shown in FIG. 8B. You may.

The data output from the level conversion table 77 is output to the multiplier 78, and the adder 6 of the excitation unit 49
8 is multiplied by the waveform signal taken out from FIG. The output of the multiplier 78 is input to the non-linear conversion table 71.
With this configuration, when the amplitude envelope of the tone signal TS generated in the loops of the delay feedback loop unit 48 and the excitation unit 49 is small, a coefficient having a relatively large value is given from the level conversion table 77 to the multiplier 78. , The level of the waveform signal input to the non-linear conversion table 71 is relatively increased. Therefore, the input signal of the nonlinear conversion table 71 greatly fluctuates, and the waveform can be largely deformed by making full use of the nonlinear conversion characteristics. As a result, even when the tone starts,
Waveform deformation by the waveform deforming section 47 can be effectively performed, and a wind instrument-like rising timbre characteristic that breathes in can be canceled, and the timbre can be changed easily from the time of rising.

In the above description, the level conversion table 77
, The nonlinear conversion table 7 of the waveform deforming section 47
The overtone component of the waveform signal deformed by the waveform deforming unit 47 is controlled by controlling the signal level on the input side of the input unit 1. However, the present invention is not limited to this, and the amplitude may be controlled. That is, the output level of the waveform transformation unit 47 may be controlled by the output of the level conversion table 77. Alternatively, by controlling the amplitude of the tone signal extracted from the delay feedback loop unit 48 based on the output of the level conversion table 77 in accordance with the type of the target tone to be synthesized, the dullness of the rising edge may be eliminated. Good.

By the way, in order to control the number of harmonic components added by the waveform transformation section 47, the nonlinear conversion table 71
May be arbitrarily controlled so as to change and control which region of the nonlinear conversion curve is used. For this purpose, the waveform deforming section 47 in FIG. 5 may be modified as shown in FIG. Although the delay feedback loop section 48 and the excitation section 49 are not shown in FIG. 9, the configuration may be the same as that shown in FIG. In FIG. 9, a multiplier 79 for input gain control is provided on a path for inputting the waveform signal extracted from the adder 68 of the excitation unit 49 to the non-linear conversion table 71, and a gain control coefficient CF of the multiplier 79 is provided.
Is variably set in accordance with various factors such as operation of an operation member (not shown) and a selected tone type. The amplitude level of the waveform signal extracted from the adder 68 of the excitation unit 49 is variably controlled according to the coefficient CF, and the nonlinear conversion table 71
The amplitude level of the waveform signal input to is controlled. As a result, the input area of the nonlinear conversion curve used in the nonlinear conversion table 71 changes, the conversion characteristics of the waveform shape are variably controlled, and the added harmonic components (number and amount of harmonics) can be arbitrarily controlled. become.

FIG. 10 shows the waveform deforming section 47 shown in FIG.
An example in which the same change as in FIG. That is, a multiplier 79 is inserted between the multiplier 78 and the nonlinear conversion table 71.

FIG. 11 shows the waveform deforming section 4 shown in FIG.
7 shows a modification example, in which a multiplier 80 is provided on the output side of the non-linear conversion table 71. The output of the envelope detection circuit 76 is input to the level conversion table 81, and a level conversion output is generated with appropriate conversion characteristics that are substantially directly proportional to the input level. The output of the level conversion table 81 is input as a multiplication coefficient of the multiplier 80, and controls the amplitude level of the transformed waveform signal. This is the level conversion table 7
7 and the multiplier 78, control is performed such that the smaller the amplitude level of the tone signal TS generated in the loop is, the higher the input level of the nonlinear conversion table 71 is. Since the output signal level has increased, this level is to be reduced in accordance with the amplitude level of the original tone signal TS. Instead of providing the level conversion table 81, the output of the envelope detection circuit 76 may be directly or appropriately bit-shifted to lower bits and input as the multiplication coefficient of the multiplier 80. Thus, at the rise of the musical tone, the waveform deformation (addition of harmonics) by the waveform deforming section 47 is sufficiently performed, but the amplitude level can be made gentle (apply an impression of breath pressure). A similar multiplier 80 and level conversion table 81 may be added to the embodiment of FIG.
Alternatively, the multiplier 79 may be omitted in FIG.

FIG. 12 shows a modification of FIG. 7, in which a plurality of different level conversion tables 77a, 77b,.
77n, the output of the envelope detection circuit 76 is input to each of the tables 77a, 77b,... 77n, and the respective outputs are given to the multipliers 82a, 82b,. Then, this is added and synthesized by the adder 83 and input to the multiplier 78.
The multiplication coefficients of the multipliers 82a, 82b,... 82n are variably set in accordance with various factors such as operation of a not-shown operator and a selected timbre. Thereby, the input gain to the nonlinear conversion table 71 can be controlled with free characteristics corresponding to the amplitude envelope level of the tone signal TS synthesized in the loop, and the mode of the harmonic overtone control can be diversified. Can be. A multiplier 79 is provided to further control the input gain to the non-linear conversion table 71 by the coefficient CF. The multiplier 79 may be omitted.

FIG. 13 shows a further modification of the waveform deforming section 47, in which a function operation circuit (slit function operation section 86) is used instead of the nonlinear conversion table 71. Delay feedback loop unit 4 not shown
8 and the excitation unit 49 may have the same configuration as that shown in FIG. In the example of FIG. 13, a signal is extracted from the low-pass filter 66 of the
Is input to a high-pass filter 84 provided on the input side of the. The high-pass filter 84 is for cutting a direct current component. The output of the high-pass filter 84 is input to a slit function calculator 86 via a multiplier 85 for gain control.

The slit function calculation section 86 calculates a predetermined slit function, and receives an input x (where-).
1 <x <1) is multiplied by で in a multiplier 87, and 積 is added to the product in an adder 88, and the sum a = (x + 1)
/ 2 is converted to a logarithm -loga having a base of 2 by a logarithmic unit 89, and the logarithmic value is multiplied by a parameter n by a multiplier 90.
An exponential function with the product b = −nloga as an exponent −b and a base of 2 is obtained by an exponent 91, and the output thereof is added by an adder 92 to −1.
/ 2 is added to obtain the operation result y. After all,
The calculation result y is a value obtained by subtracting (1/2) from a value obtained by raising {(x + 1) / 2} to the nth power. Here, when the parameter n is 1, a monotonous change of y = x / 2 is shown, but by performing variable control with n <1 or n> 1, a non-linear function curve is shown. Therefore, there is an advantage that the nonlinear conversion characteristic can be variably controlled by the variable control of the parameter n.

The output of the high-pass filter 84 is input to a low-pass filter 93 for detecting an envelope, and an output corresponding to the envelope level is obtained. Low-pass filter 9
3 is input to the level conversion table 77 via the multiplier 94, and is output to the level conversion table 77.
Is input to the multiplier 85 via the adder 95. These envelope detection low-pass filters 93
The level conversion table 77 has the same function as the envelope detection circuit 76 and the level conversion table 77 shown in FIG. Although not shown, the multiplier 94
, An appropriate coefficient parameter is input to the adder 95, and an appropriate addition parameter is input to the adder 95.
An appropriate filter coefficient parameter is input to 3, and by varying these parameters, the mode of harmonic overtone control according to the tone signal envelope level can be freely controlled. Also, the slit function calculation unit 86
Is input to the multiplier 96, and the output of the low-pass filter 93 for envelope detection is multiplied. The multiplier 96 is provided for the same purpose as the multiplier 80 in FIG.

In the embodiments shown in FIGS. 5 to 12, the signal input to the waveform deforming section 47 is
8, the signal is not limited to this, and may be a signal extracted from any part of the delay feedback loop unit 48 and the excitation unit 49. Therefore, the envelope detection circuit 76
May be input to the waveform deforming unit 47 as a waveform signal to be deformed. Further, the extracted tone signal TS or the like may be configured to appropriately cut a DC component by high-pass filter processing as shown in FIG. Also, in FIG. 13, the signal input to the waveform transformation unit 47 is not limited to the output of the low-high-pass filter 66 of the excitation unit 49, and may be a signal extracted from any part of the delay feedback loop unit 48 and the excitation unit 49.

In the sound source circuit comprising the delay feedback section 10 and the excitation sections 14 and 43 as shown in FIGS. 1 to 4, a waveform deforming section 47 as shown in FIGS. You may. Conversely, in the sound source circuit including the delay feedback section 48 and the excitation section 47 as shown in FIG. 5, the waveform deforming section 27 as shown in FIG.
May be used. Further, the configuration of the waveform deforming section 47 as shown in FIGS.
The present invention is not limited to a physical-model-type sound source such as a delayed feedback-type sound source or a closed-waveguide-type sound source as shown in the figure. Also, FIGS. 7, 10, and 1
The detection technique of the amplitude envelope level of the musical tone signal TS as shown in FIG. 1 and the control technique of the waveform deformation control mode in the waveform deformation section 47 based on the detection technique is used when the waveform deformation section 27 as shown in FIG. Can also be applied.

FIGS. 1 and 5 show waveform deforming means.
Not limited to those shown in the waveform deforming portions 27 and 47 of FIG. For example, a circuit including a modulation unit that performs frequency modulation (FM) or amplitude modulation (AM) operation in an audio frequency band as a modulation wave or a modulated wave using a waveform signal to be deformed extracted from a loop of the delay feedback units 10 and 48. The waveform deforming means may be constituted by the above. FIG. 14 shows an example in which a waveform signal to be deformed extracted from the loop of the delay feedback units 10 and 48 is appropriately variably controlled by a coefficient k1 by a multiplier 97, and the carrier wave data is transmitted to an arithmetic circuit 98 for FM or AM. Input to input CW. Arithmetic circuit 98 performs an arithmetic operation of frequency-modulating or amplitude-modulating the signal applied to carrier wave data input CW with the signal applied to modulated wave data input MW. The modulation output is output from a multiplier 99 by a modulation index coefficient k2.
, And is fed back to the modulated wave data input MW. Further, the modulation output of the arithmetic circuit 98 is variably controlled by the multiplier 100 by the coefficient k3 as appropriate, and is output for addition and synthesis to the tone signal synthesized by the loop of the delay feedback units 10 and 48. FIG.
The modulation operation is not limited to the self-feedback type modulation operation described above, but may be another type of modulation operation.

The configuration shown in each of the above-described embodiments is not limited to a hard wired circuit, but may be a program processing using a general-purpose processor such as a microprocessor or a computer. Means for generating various performance parameters in real time include the breath controller and keyboard described above.
Not limited to switches, any other devices such as a pedal operator, a joystick, a mouse, and a touch sensor may be used. Further, it is needless to say that appropriate performance parameters and control signals may be received via the MIDI wiring, and thereby the tone synthesis and control operation according to the present invention may be controlled.

Some of the inventions and embodiments extracted from the above embodiments are summarized and listed as follows. 1. A signal circulating means for forming a closed loop for circulating a signal and delaying the signal in the loop, wherein a resonance characteristic is controlled by varying the delay time; and a loop in the signal circulating means. Exciting means for exciting a waveform signal, extracting a waveform signal from a loop of the signal circulating means, deforming the waveform to add a harmonic component, and a waveform signal extracted from the loop of the signal circulating means. A musical sound waveform signal synthesizing device comprising: a synthesizing means for outputting a musical sound waveform signal by synthesizing the waveform signal deformed by the waveform deforming means. 2. The first aspect, wherein the waveform deforming means includes a differentiating means for differentiating a waveform signal taken out of a loop of the signal circulating means.
A musical tone signal synthesizer. 3. 3. The tone waveform signal synthesizing apparatus according to claim 2, wherein said waveform deforming means includes means for converting an amplitude value of the differentiated waveform signal into an absolute value. 4. A comparing means for extracting only a waveform portion having one of a positive or negative amplitude value from the differentiated waveform signal, and comparing the output waveform of the comparing means and the waveform subjected to the absolute value conversion with each other. 4. The tone waveform signal synthesizing apparatus according to claim 3, further comprising means for variably controlling in both positive and negative directions to perform addition and synthesis. 5. 5. The tone waveform according to claim 3 or 4, wherein said waveform deforming means includes means for extracting a low-band component from the differentiated waveform signal, and means for adding the extracted low-band component to an output waveform signal of the waveform deforming means. Signal synthesizer. 6. 2. The musical tone waveform signal synthesizing apparatus according to claim 1, wherein said waveform deforming means includes means for cutting out a part of the waveform. 7. The waveform deforming means includes a non-linear conversion table,
2. The tone waveform signal synthesizing apparatus according to claim 1, wherein a waveform signal extracted from a loop of said signal circulating means is input to said non-linear conversion table and subjected to non-linear conversion to perform waveform deformation. 8. The excitation means includes amplitude control means for variably controlling the amplitude of a waveform signal to be introduced into the signal circulation means, and the waveform deformation means converts the amplitude of the deformed waveform signal output from the non-linear conversion table to the amplitude. 8. The tone waveform signal synthesizing apparatus according to claim 7, further comprising means for performing variable control in conjunction with control by the control means. 9. Means for detecting the amplitude level of the waveform signal passing through the loop of the signal circulating means, and means for controlling the amplitude of the waveform signal input to the non-linear conversion table according to the detected amplitude level. Said 7 further including
Item 8 or the tone wave signal synthesizing device according to item 8. 10. The means for controlling the amplitude of the waveform signal input to the non-linear conversion table according to the detected amplitude level controls the amplitude of the waveform signal to increase as the detected amplitude level decreases. 9. Music signal synthesizer. 11. The means for controlling the amplitude of the waveform signal input to the non-linear conversion table according to the detected amplitude level includes a table for reading an amplitude control coefficient according to the detected amplitude level, and an amplitude control coefficient read from the table. Means for controlling the amplitude of said waveform signal input to said nonlinear conversion table in accordance with said means. 12. There are multiple tables, one of which is selected
12. The tone waveform signal synthesizing apparatus according to the above 11, wherein the amplitude control coefficients read from the two tables are used. 13. 13. The tone waveform signal synthesizing apparatus according to any one of the items 7 to 12, wherein the non-linear conversion table includes an operation unit that executes a predetermined non-linear conversion operation. 14． 14. The method according to any one of claims 7 to 13, wherein the waveform deforming means further includes means for controlling an amplitude of the waveform signal input to the non-linear conversion table in accordance with an operation of a manipulator.
Music signal synthesizers. 15. 2. The musical sound waveform signal synthesizing apparatus according to claim 1, wherein said waveform deforming means includes a modulating means for performing a frequency modulation operation using a waveform signal taken out of a loop of said signal circulating means as a modulated wave or a modulated wave. 16. 2. The musical tone signal synthesizing apparatus according to claim 1, wherein said waveform deforming means includes a modulating means for performing an amplitude modulation operation using a waveform signal taken out of a loop of said signal circulating means as a modulation wave or a modulated wave. 17． 17. The tone waveform signal synthesizing apparatus according to claim 15 or 16, wherein an output signal of the modulation means is used as a modulation wave to modulate a waveform signal taken out of a loop of the signal circulation means. 18. The excitation means is inserted in a loop of the signal circulation means, and has a nonlinear conversion means for nonlinearly converting an input signal, and a means for introducing a signal for exciting oscillation into the loop. Item 18. A tone waveform signal synthesizing apparatus according to any one of Items 1 to 17. 19. 18. The tone waveform signal synthesizing apparatus according to any one of claims 1 to 17, wherein the excitation means includes means for introducing a predetermined initial waveform signal into a loop of the signal circulation means. 20. 20. The tone waveform according to any one of claims 1 to 19, wherein the mixing means includes means for variably controlling a mixing ratio of a waveform signal extracted from a loop of the signal circulation means and a waveform signal deformed by the waveform deformation means. Signal synthesizer. 21. (1) The apparatus according to (1), further comprising means for inputting performance parameters, wherein the input parameters are provided to at least one of the signal circulating means, the exciting means, the waveform deforming means, and the synthesizing means to control the synthesis of the tone waveform signal. 21. A musical sound wave signal synthesizing apparatus according to any one of claims 20 to 20. 22. A waveform generating means for generating a waveform signal, a non-linear conversion means for inputting the waveform signal generated by the waveform generating means and non-linearly converting the input waveform signal, and detecting an amplitude level of the waveform signal generated by the waveform generating means Means for controlling the amplitude of the waveform signal input to the non-linear conversion means in accordance with the magnitude of the detected amplitude level, and the waveform signal generated by the waveform generation means and the non-linear conversion performed by the non-linear conversion means. And a mixing means for outputting a tone waveform signal by mixing the waveform signal.

[0059]

As described above, according to the present invention, the closed loop sound source portion including the signal circulating means and the excitation means enables the creation of a specific sound of a physical model type or similar, while the waveform deforming means. By performing the waveform deformation such that a harmonic component that does not occur in the closed-loop sound source portion is generated, the harmonic composition of the waveform signal synthesized by the closed-loop sound source portion can be enriched.
Therefore, the sound generation in the closed-loop sound source portion can be performed by performing the waveform deformation for generating the harmonic component in the portion of the waveform deforming means, even if a stable oscillation is possible and a relatively easy control method is adopted. The range of sound creation can be expanded based on a rich harmonic composition, and further transformed by waveform transformation means
From the loop of the signal circulation means
The wave obtained as a result of calculation by calculating with the waveform signal of
Runode can be further changed complex shape signal, it is possible to synthesize a high-quality easy <br/> sound waveform signal including a complex overtones yet relatively easy control, good that It has the effect.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

2A and 2B are graphs showing examples of characteristics of a nonlinear conversion table of an excitation unit in FIG. 1; FIGS. 2A and 2B show examples of characteristics of two different nonlinear conversion tables provided in parallel;
(C) and (d) are diagrams respectively showing characteristic examples obtained by combining them at an appropriate ratio.

FIG. 3 is a waveform chart showing an example of a waveform deformation operation in a waveform deformation unit in FIG. 1;

FIG. 4 is a block diagram showing another embodiment of the present invention.

FIG. 5 is a block diagram showing still another embodiment of the present invention.

FIG. 6 is a graph illustrating conversion characteristics in each nonlinear conversion table in FIG. 5;

FIG. 7 is a block diagram showing a modified example of the waveform deforming unit shown in FIG. 5;

8 is a graph showing an example of conversion characteristics of the level conversion table in FIG. 7;

FIG. 9 is a block diagram showing another modified example of the waveform deforming unit shown in FIG. 5;

FIG. 10 is a block diagram showing a modification of the waveform deforming unit shown in FIG. 7;

FIG. 11 is a block diagram showing another modified example of the waveform deforming unit shown in FIG. 7;

FIG. 12 is a block diagram showing still another modified example of the waveform deforming unit shown in FIG. 7;

FIG. 13 is a block diagram showing still another modification of the waveform deforming unit shown in FIG. 5;

FIG. 14 is a block diagram showing another embodiment of the waveform deforming unit shown in FIG. 1 or FIG.

[Explanation of symbols]

 10, 48 delay feedback loop section 11, 51L, 51S delay circuit 14, 49 excitation section 20 operation parameter generation section 27, 47 waveform transformation section 21, 22, 62, 63, 71 nonlinear conversion table 28 differentiation circuit 29 absolute value circuit 44 Excitation waveform generator 76 Envelope detection circuit 77, 77a, 77b, 77n Level conversion table

──────────────────────────────────────────────────続 き Continued on front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G10H 7/08 G10H 7/00 513

Claims (1)

(57) [Claims] 1. A signal circulating means for forming a closed loop for circulating a signal and including a delay means for delaying a signal in the loop, wherein a resonance characteristic is controlled by varying the delay time; Exciting means for exciting a waveform signal in the loop of the circulating means, extracting a waveform signal from the loop of the signal circulating means, deforming the waveform to add a harmonic component, and deforming the waveform by the waveform deforming means . The waveform signal
Calculation using another waveform signal extracted from the loop of the stage
Means, and the computing means based on the waveform signal extracted from the loop of the signal circulating means and the waveform signal deformed by the waveform deforming means.
An output means for outputting an output waveform signal of a stage .