JP3291965B2 - Music synthesis apparatus and music synthesis method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for synthesizing a musical tone such as an electronic musical instrument, a game device, a personal computer system, an application program thereof, and a musical tone processor. In particular, the present invention simulates the physical properties of a natural musical instrument. The present invention relates to a musical tone synthesizing apparatus and a musical tone synthesizing method suitable for an electronic musical instrument.

[0002]

2. Description of the Related Art Conventionally, there has been known a musical tone synthesizing apparatus which synthesizes musical tones by simulating a sounding mechanism of a natural musical instrument using an electric model. Here, FIG. 11 shows an example of a signal forming circuit in a conventional tone synthesizer. This signal forming circuit is for synthesizing a musical tone of a wind instrument among natural musical instruments. As shown in the figure, a tube simulating section simulating the physical properties of a tube in a wind instrument electrically. 20 and an excitation circuit 10 that generates an excitation signal based on a performance operation performed by the player and supplies the excitation signal to the tube simulation unit 20. Various parameters in the excitation circuit 10 and the tube simulation unit 20 are also controlled based on the performance operation.

First, the subtracter 1 in the excitation circuit 10
A signal corresponding to the pressure p (oral pressure) blown into the mouthpiece is generated and supplied to a subtraction input terminal (-) of 1 in accordance with a sensor (not shown) for detecting a performance operation. On the other hand, to the addition input terminal (+) of the subtractor 11, a signal indicating the pressure q in the mouthpiece, which is a signal simulated from the tube, is supplied as described later. Therefore, the output signal of the subtractor 11 is a signal corresponding to the air pressure difference Δp in the gap between the mouthpiece and the lead. The low-pass filter 12 simulates the movement of a lead, and outputs an input signal after band-limiting the input signal. The reason why the band is limited in this way is to simulate the responsivity of the lead to the pressure change. More specifically, when the pressure of the lead changes, inertia and the like act on the lead itself, causing a delay in the displacement of the lead.Furthermore, when the frequency of the pressure change increases, the lead gradually stops responding. This is to simulate. The low-pass filter 12 has a parameter F for controlling its characteristics according to a performance operation or the like.
c and Q are supplied, and the cut-off frequency and the Q value are respectively set by these parameters.

A signal E, which is embouchure data indicating the pressure applied to the lead, is added to the output of the low-pass filter 12 by an adder 13 to obtain a signal corresponding to the pressure actually applied to the lead. Then, this signal is converted by the nonlinear table 14 into a signal corresponding to the cross-sectional area S of the gap between the mouthpiece and the lead, and supplied to one input terminal of the multiplier 15.

On the other hand, a signal indicating the air pressure difference Δp in the gap between the mouthpiece and the lead is transmitted to the nonlinear table 16.
Is also supplied. The non-linear table 16 simulates that the flow velocity saturates in a narrow pipe line and the differential pressure is not proportional to the flow velocity even when the pressure difference of the air becomes large. A signal corresponding to the air pressure corrected in consideration of the influence on the flow velocity can be obtained.

The output signal of the non-linear table 16 is supplied to the other input terminal of the multiplier 15 and is multiplied by a signal indicating a gap (a cross-sectional area of a portion) S between the leads. A signal f corresponding to the volume flow velocity in the gap between the lead and the lead is supplied to the multiplier 17. This signal f is multiplied by a signal indicating the characteristic impedance z (corresponding to resistance) of the mouthpiece by the multiplier 17 to become a sound pressure signal fz corresponding to the air sent from the mouthpiece to the tube, and as an excitation signal It is supplied to the tube simulation unit 20. In this way, in the excitation circuit 10, the excitation section including the lead of the wind instrument is electrically simulated.

Next, the tube simulating section 20 feeds back the sound pressure signal fz as a signal q, and a low-pass filter and a delay circuit are inserted in the feedback path (not shown). . The low-pass filter simulates the shape of the tube, especially the resonance tube, and the delay circuit adjusts the length of the resonance tube and the length of the tone hole from the mouthpiece. This simulates the state in which the wave returns to the mouthpiece as a reflected wave. In this case, the delay time of the delay circuit is controlled by the pitch of the generated musical tone. Strictly speaking, a delay occurs in the low-pass filter in the feedback path, so that the pitch of the generated musical tone is controlled by the tube simulation unit 2.
This is performed by controlling the delay time of the delay circuit while taking into account the delay time of the low-pass filter so that the delay time per cycle corresponds to the pitch.

[0008] Since the flow of air generated in the wind instrument is simulated by the electric model by the signal forming circuit configured as described above, it is possible to synthesize a tone signal approximate to the tone of the actual wind instrument. Has become. The output signal at this time is supplied to the excitation circuit 10 or the tube simulating unit 2 in the signal forming circuit.
It can be taken from any point at 0.

[0009]

By the way, it has been pointed out that the tone of ff (fortessimo) of the synthesized sound by such a configuration is weak even if the signal corresponding to the intraoral pressure p is increased. Therefore, the present inventor considered that there was a problem in modeling the nonlinear characteristic in the excitation circuit 10 while searching for the cause, and decided to examine the characteristic again.

First, the conventional excitation circuit 10 can be roughly classified into a portion that approximates the dynamic characteristics of a lead (a low-pass filter 12), a portion that approximates its non-linear characteristics (a non-linear table 14), and the air flow in the lead gap. The flow can be divided into parts that approximate the flow (nonlinear table 16). Here, how the synthesis is performed will be considered by fixing the gap between the leads.

That is, since the reed moves slowly as compared with the air flow and pressure phenomena, when the air flow and pressure phenomena are considered instantaneously, the reed is fixed and does not move. It can be treated as if it were. Therefore, the lead gap (output of the nonlinear table 14) S in the excitation circuit 10 is set to “1.0”, “0.5”, “0.2”.
5 "(no unit), and consider the characteristics of the volume flow rate f when the pressure difference Δp changes. FIG. 9A shows the characteristics in this case.

Next, the pressure difference Δp in an actual natural musical instrument
FIG. 4B shows the relationship between the volume flow rate f and the volume flow rate f. Here, comparing the two characteristics of FIGS. 9A and 9B, the conventional excitation circuit 10 well approximates the air flow of the natural musical instrument in consideration of the fact that the saturation functions are very similar. However, the present inventor focused on the following two differences. Although the relationship between the pressure difference Δp and the volume flow rate f originally does not depend on the lead gap S as shown in FIG. In the conventional modeling, as shown in FIG. 2A, the entire function has a vertically compressed shape, and the inclination becomes smaller as the lead gap becomes smaller. The relationship between the pressure difference Δp and the volume flow rate f is, as shown in FIG. 3 (b), the volume flow rate f increases and the saturation does not occur as the absolute value of the pressure difference Δp increases. However, as shown in FIG. 3A, it is completely saturated and becomes a constant value.

The inventor of the present application believes that the tone of fortesimo is weak is mainly based on the above difference, and has made the present invention to solve the problem.
The present invention has been made in view of the above-described problem, and an object of the present invention is to more accurately simulate the flow of air in a natural musical instrument and synthesize a musical tone that does not weaken the tone applied to ff. It is an object of the present invention to provide a possible tone synthesis device and a tone synthesis method.

[0014]

In order to solve the above-mentioned problems, according to the present invention, a loop means having at least a delay means and a low-pass filter is provided.
An excitation signal generation means for generating an excitation signal and supplying the excitation signal to the loop means, wherein the excitation signal generation means comprises a level characteristic of the excitation signal. and against the performance operation amount to increase the volume of the to be synthesized musical tone is changed in accordance with a predetermined function, the predetermined function, with respect to the direction of change to increase the tone of the performance operation amount, the excitation signal Level increases monotonically , and
The change in the excitation signal when the
The characteristic is that the chemical content is constant with respect to the fluctuations of other performance operation amounts.
It is characterized in that. According to a second aspect of the present invention, in the first aspect of the present invention, the excitation signal generating unit is configured to generate the excitation signal in accordance with a difference between a signal from the loop unit and a signal corresponding to a performance operation. It is characterized by generating a signal. In the invention according to claim 3, in the invention according to claim 1 or 2, the predetermined function includes at least a root function, and the excitation signal is transmitted via a conversion table that stores the root function. It is characterized by being generated. In the invention described in claim 4 , in the invention described in claim 1 or 2, the predetermined function has a non-linear characteristic in which an input-output characteristic curve is non-linear. The characteristic curve is controlled so as to move on coordinates in accordance with the performance operation amount. According to the fifth aspect of the present invention, there is provided a synthesizing step of generating an excitation signal, and synthesizing a tone signal by circulating and extracting the excitation signal through a loop unit having at least a delay unit and a low-pass filter. Wherein the generating step changes a level characteristic of the excitation signal in accordance with a predetermined function with respect to a performance operation amount for increasing the volume of a musical sound to be synthesized, and the predetermined function is In response to a change in the direction of increasing the musical tone of the performance operation amount,
The level of the excitation signal monotonically increases ,
Excitation signal performance operation amount in the case of zero to increase the sound
The amount of change in the number is constant with respect to the fluctuation of other performance operation amounts
It is characterized by characteristics.

[0015]

In this type of musical instrument, the characteristics of the loop means and the excitation signal are controlled by simulating physical properties such as the shape of a natural musical instrument. Saturates against the amount of performance operation that increases. According to the first aspect of the present invention, the excitation signal is supplied to the loop means having the delay means and the low-pass filter by the excitation signal generation means, and the characteristic of the excitation operation is to increase the volume of the musical sound to be synthesized. Monotonic increase with quantity. Also,
Even if the playing operation amount is close to zero, the change
If there is, it can secure the change of the excitation signal
You. According to the second aspect of the present invention, the signal (feedback signal) returned from the loop means is reflected in the generation of the excitation signal. According to the third aspect of the present invention, the function that determines the characteristics of the excitation signal has at least a root function. However, when the excitation signal is generated, the function passes through the conversion table. The time required is reduced. According to the fourth aspect of the present invention, the characteristic curve of the predetermined function is controlled so as to move on the coordinates according to the performance operation amount, so that the characteristic curve approximates the nonlinear characteristic of the actual musical instrument. Can be done. Claim 5
According to the invention described in (1), similarly to the invention described in (1), the characteristic of the excitation signal is monotonically increased with respect to the performance operation amount for increasing the volume of the musical sound to be synthesized. Also playing
Even when the manipulated variable is close to zero, the change is somewhat
If so, a change in the excitation signal can be secured.

[0016]

【Example】

1: Embodiment An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a signal forming circuit according to a musical sound synthesizer of the present invention. The signal forming circuit shown in this figure is intended to solve the above-mentioned problem by performing synthesis of the functions by the multipliers 15 and 17 and the nonlinear table 16 in FIG. . Further, this embodiment has a junction portion 30 for simulating the turbulence of the air flow at the joint between the mouthpiece and the tube.

1-1: Junction Section First, the junction section 30 will be described for the sake of convenience.
It is configured as shown in (a). In the configuration shown in the figure, while the feedback signal q _i of the sound pressure signal fz and tube simulating portion 20 of the multiplier 17 are added by the adder 31 is supplied as the signal q _O to tube simulating unit 20 The feedback signal q _i of the tube simulation unit 20 and the output signal q _O of the adder 31 are added by the adder 32 and supplied to the subtractor 11 as a signal q _h . With this configuration, at the junction, scattering of air pressure at the junction with the pipe is simulated. However, in this configuration, since sound pressure signal fz is fed back directly to the exciting circuit 10 as a component of the signal q _h, the oscillation phenomenon is likely to occur. Therefore, FIG.
Is slightly changed so that the junction 3 in FIG.
0. That is, at the junction in FIG. 12A, the addition result of the adder 31 (signal q
_o ) is the sum of the sound pressure signal fz and the signal q _i , so that the adder 32 adds the addition result of the adder 31 and the signal q _i.
Is a multiplier 34 for multiplying the signal q _i by a coefficient “2”, a result of the multiplication and a sound pressure signal fz, as shown in FIG.
Can be equivalently replaced by an adder 33 that adds. In this configuration, the sound pressure signal fz is
1, that is, the configuration in which the adder 33 is omitted is the configuration itself of the junction unit 30 in FIG. 1.

1-2: Nonlinear Part Model Next, the nonlinear part model 200 will be described. The nonlinear part model 200 includes a signal indicating the lead gap S converted by the nonlinear table 14 and an air pressure difference Δp
And the signal indicating the characteristic impedance z of the mouthpiece are appropriately combined with the sound pressure signal fz.

1-2-1: Nonlinear Function Synthesized by Nonlinear Part Model The nonlinear function synthesized by the nonlinear part model 200 will now be described. First, the inventor of the present application decided to return to the original physical mechanism and study the modeling of an actual wind instrument. The flow velocity at the lead portion of the mouthpiece is given by the following equation (1) from Graham's theorem in hydrodynamics.

(Equation 1) In this equation, p is the pressure in the mouth, q is the pressure in the mouthpiece, and ρ is the density of air. When (p−q) is negative in consideration of the air flow direction, the positive characteristic is symmetric with respect to the origin.

Here, the pressure q in the mouthpiece can be expressed by the following equation. q = q _i + q _O Next, the signal q _o in FIG. 1 is a sound pressure signal fz and a signal q _i
Therefore, this equation can be further expressed as the following equation. q = 2q _i + fz _{O In} this equation, assuming that 2q _i is q _h , the pressure p in the mouthpiece is a signal indicating the inflow pressure from the joint to the mouthpiece as shown in the following equation. q _h ,
Conversely, it can be expressed as the sum with the sound pressure signal fz indicating the outflow pressure from the mouthpiece to the joint. q = q _h + fz _O ... Assuming that the sound velocity is c and the cross-sectional area of the mouthpiece immediately below the lead is s ₀ , the characteristic impedance z ₀ at the lead becomes as shown in the following equation. z ₀ = ρc / s ₀ ... Then, when q in the equation is substituted into the equation (1) and f is also solved using the equation, the following equation (2) is obtained.

[0021]

(Equation 2)

Here, in order to compare the equations (1) and (2), the opening of the slit is "0.05 cm" and the width is "1.3".
When the product is set to 6 cm, and the product is used as the gap S between the leads, and the characteristic impedances are further equalized, the characteristics of both are as shown in FIG. As can be seen from this figure, equation (1)
Is moved so as to be in contact with the straight line of f = −1 / z ₀ and to pass through the origin, Expression (2) is obtained.
Therefore, when the equation (2) is to be realized digitally, it is sufficient to provide a four-calculation circuit together with a table for obtaining the characteristics of the equation (1) and a function generator (root function). It should be noted that the characteristics in this figure show the case where (q _h -p) corresponding to Δp is negative, and the characteristics when (q _h -p) is positive, taking the air flow direction into consideration. (Q _h
This is a characteristic obtained by symmetrically moving the characteristic when −p) is negative about the origin. Then, the nonlinear part model 200
Is the sound pressure signal fz which is the product of the flow velocity f and the impedance z _0, and by rearranging the two sides of the equation (2) by the impedance z ₀ , the following equation (3) is obtained. .

[0023]

(Equation 3) In this way, the present inventor has proposed the nonlinear part model 200
The non-linear function to be synthesized in is set in Expression (3) including the root function. Note that this equation (3)
Indicates a sound pressure signal when (q _h -p) is negative.

1-2-2: Configuration of Nonlinear Part Model Next, the configuration of the nonlinear part model 200 for synthesizing the nonlinear function of the equation (3) will be described. FIG. 2 is a block diagram showing a configuration of the nonlinear part model 200. As shown in this figure, the pressure difference Δp is supplied to a positive / negative detector 201 and an absolute value calculation circuit 202. Positive / negative detector 201
Is provided to make the function result symmetric at the origin according to the positive or negative of the pressure difference Δp. Specifically, in the positive / negative detector 201, Δp is positive (including the case where it is zero).
In this case, “+1” is output, and if negative, “−1” is output. In addition, for positive / negative detection,
For example, it is determined by the most significant bit of Δp.

The absolute value calculation circuit 202 calculates the absolute value of the pressure difference Δp including the flow direction by full-wave rectification and calculates the absolute value.
03 to the subtraction input terminal (-). The addition input terminal (+) of the subtractor 203 is supplied with “S ² z ₀ ² / 2ρ” calculated in advance in the preceding stage (not shown), from which the absolute value of the pressure difference Δp is subtracted. Route function table 2
04 is input. The root function table 204 is a table in which, for an input x and an output y, a root function result of y = (2x / ρ) ^1/2 is stored in advance.
Is supplied to one input terminal of the multiplier 205. The other input terminal of the multiplier 205 is supplied with “Sz ₀ ” calculated in advance in a preceding stage (not shown), multiplied by the output y of the root function table 204, and added to the addition input terminal ( +). Subtractor 206
Is supplied with “S ² z ₀ ² / ρ” calculated in advance in the preceding stage (not shown),
The result of multiplication by 5 is subtracted and supplied to the multiplier 207.

The output of the positive / negative detector 201 and the coefficient "-1" are also supplied to the multiplier 207, and the result of the multiplication by the multiplier 205 is determined based on the positive or negative of the pressure difference Δp. It is designed to be symmetrical about the center. That is, this is to expand the calculation result of Expression (3) in consideration of the air flow direction.

The following equation (4) shows the nonlinear function synthesized by the nonlinear part model 200 having such a configuration.

(Equation 4)

In this configuration, other than the pressure difference Δp can be treated as a constant, and the calculation of the route function requiring the longest time for processing is performed by conversion of the route function table 204. Can be performed in real time. Since the relationship between the pressure difference and the flow velocity occurring in an actual natural musical instrument is more accurately simulated, the flow velocity does not saturate even if the pressure difference increases.

Further, the pressure difference Δp is generated based on a signal corresponding to the pressure p detected by a sensor (not shown), and other constants can be arbitrarily set by the user. Even in this case, the relationship between the pressure difference Δp and the sound pressure signal fz is f = −− as shown in FIG.
By moving so as to be in contact with the straight line of 1 / z ₀ and to pass through the origin, the set condition is satisfied.

Also, with respect to equation (4), q _h -p is differentiated as x, for example, and the gradient (change) when x = 0 (ie, when the pressure difference is zero) is changed from the positive side to x = 0. and when brought closer, when determined separately in the case of close from the negative to the x = 0, the variation, since both are the -1 / z _0, becomes constant without depending on the read gap S, Also in this regard, it can be said that the relationship of the flow velocity occurring in the actual natural musical instrument is more accurately simulated.

1-3: Pipe Simulator Next, the pipe simulator 20 will be described. In this embodiment, several algorithms are prepared according to the shape, type, etc. of the tube to be simulated, and any one of them is configured to be selectable. The following describes two representative algorithms.

1-3-1: First Algorithm of Pipe Simulator First, the first algorithm will be described. FIG. 3 is a block diagram showing a configuration of the pipe simulation unit 20 based on the first algorithm. This algorithm simulates a tube shape (a collection of cylinders with different diameters) as shown in FIG. 6, and combines all the tone holes and register tubes of the wind instrument (including the half-open state). And is closest to an acoustic instrument that can realize a tube of an arbitrary shape. In FIG. 3, what is indicated by a symbol SR and a suffix is a shift register, which simulates a propagation delay of an air pressure wave in a pipe. What is indicated by the symbol J and the suffix is a junction portion, which simulates the scattering of the air pressure wave generated at the point where the diameter of the tube changes. The LPF is a low-pass filter that simulates energy loss or the like when an air pressure wave is reflected at the end of the tube.

In the example shown in FIG. 3, the junctions J1, J2, and J5 have only a step without a tone hole (two-port junction). ) To (b). Also, the junction J
Reference numerals 3, J4, and J6 each have a high tone hole (3-port junction), and the configuration is as shown in FIG. Here, the parameters α, β, γ in the first algorithm depend on the diameter φ of the tube and the diameter ト ー ン of the tone hole in FIG. 6, and the number m3, m4, m6 of the shift register SR _{t3 and the} like.
Is determined by the height t (t3, t4, t6) of the tone hole. Further, each delay time of each shift register SR corresponds to the length L1 to L7 of each tube in FIG. That is, control is performed so as to correspond to the pitch of a musical tone to be synthesized. The parameters γ _t1 , γ _t2 , γ _t3 are reflected so that the tone hole in FIG. 6 is negative when the tone hole is open and positive when the tone hole is closed. As described above, each parameter for determining the simulation mode by the tube simulation unit 20 is output by the control processing unit (not shown) based on the shape of the tube to be simulated. In this algorithm, the presence / absence of a tone hole at the junction can be arbitrarily set.

1-3-2: Second Algorithm of Pipe Simulator Next, the second algorithm will be described. FIG. 7 is a block diagram showing the configuration of the pipe simulation unit 20 based on the second algorithm. This algorithm is a tube model without a tone hole, and the length to the end of the tube determines the sound frequency. In addition, such a tube model includes a type in which a tube portion is realized by cascade connection using a two-port WGN (waveguide network), and a case in which a conical portion, which is a part of the tube, is a simple WGN.
This algorithm corresponds to the latter type.

This type is derived from an approximate expression of the input acoustic impedance of a cone. In the algorithm shown in FIG. 7, two cylinders WGN are connected in parallel. Also in this algorithm, various parameters are calculated according to the tube shape. That is, similarly to the first algorithm, these parameters are output from a control processing unit (not shown). The first and second algorithms are described in detail in, for example, Japanese Patent Application Laid-Open No. 5-80761.

1-4: Effects of the Embodiments Conventionally, as a result of simulating that the flow velocity saturates in a narrow pipeline even when the air pressure becomes large and the air pressure is not proportional to the flow velocity, the excitation signal has a certain pressure difference. , The saturation becomes a constant value, and the tone applied to ff could not be successfully synthesized. According to such a configuration, when the pressure difference increases, the slope decreases due to the root function but continues to increase monotonously without saturation, so that an excitation signal corresponding to the pressure difference can be generated.
Thus, the timbre relating to f can be successfully synthesized. Also,
Even if the pressure difference Δp changes near zero, the change in the flow velocity f is constant regardless of the gap S between the leads, so that the change in the sound pressure signal fz as the excitation signal can be secured. . Thereby, when the gap S between the leads is small, it is possible to avoid the problem that the tone change by the tube simulation unit 20 is small.

Although not particularly described in the above embodiment, the lead gap S obtained by the non-linear table 14 is fed back to the adders 13 and 18 via the multipliers FBNL and FB, respectively. It may be configured to provide feedback. This allows the leads to be simulated to have a hysteresis characteristic,
Further, if each of the multiplication coefficients is set to zero, it can be set to one having no characteristic.

Next, a modification of the above-described embodiment will be described with reference to FIG. As shown in this figure, in this modification, the sound pressure signal Fz is multiplied by a coefficient
8 and is added to the pressure difference Δp. With such a configuration, feedback control can be performed in calculating the lead gap S.

Although the above embodiment has been described with reference to the case of hardware configuration, even if such a signal processing procedure is described in a program and executed by a DSP or an MPU, equivalent musical sound synthesis is possible. In addition, a tone synthesis can be performed even in a mixed system of both. That is, in the present application, not only the hardware configuration but also the tone synthesis can be performed by the algorithm indicating the signal processing procedure.

[0040]

According to the present invention described above, the following effects can be obtained. Since the level of the excitation signal monotonically increases with respect to the performance operation amount for increasing the volume of the musical sound to be synthesized, it is possible to prevent the synthesized tone from becoming weak even if the performance operation is performed on ff. I can . Also,
Even if the performance operation amount is close to zero, the change
If there is a degree, it is possible to secure the change of the excitation signal
Therefore, the function of the loop means should be more reliable.
(Claims 1 and 5) . To generate the excitation signal,
Since the signal (feedback signal) returned from the loop means is reflected, it is possible to make use of the features of the natural musical instrument (claim 2). The function that determines the characteristics of the excitation signal has at least a root function. However, the generation of the excitation signal is performed via the conversion table, so that the time required for generating the excitation signal can be shortened. Item 3). The characteristic curve having a predetermined function can be approximated to the nonlinear characteristic of the actual instrument (claim 4).

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a first embodiment according to the present invention.

FIG. 2 is a block diagram illustrating a configuration of a nonlinear unit model in the embodiment.

FIG. 3 is a block diagram showing a configuration of a first algorithm used in the tube unit in the embodiment.

FIG. 4 is a block diagram showing a configuration of junctions J1, J2, J5 in a first algorithm.

FIG. 5 is a block diagram showing a configuration of junctions J3, J4, J6 in a first algorithm.

FIG. 6 is a schematic configuration diagram showing a tube shape simulated by a first algorithm.

FIG. 7 is a block diagram showing a configuration of a second algorithm used in the tube unit in the embodiment.

FIG. 8 is a block diagram showing a configuration of a second embodiment according to the present invention.

FIG. 9A is a diagram showing a relationship between a pressure difference and a flow velocity according to a conventional configuration, and FIG. 9B is a diagram showing the same relationship with an actual natural musical instrument.

FIG. 10 is a diagram showing the relationship between the characteristics according to the example and the characteristics shown in FIG. 9 (b).

FIG. 11 is a block diagram showing a configuration of an excitation circuit in a conventional tone synthesizer.

FIG. 12 is a block diagram illustrating a configuration of a junction unit in the excitation circuit.

[Explanation of symbols]

10 excitation circuit (excitation signal generation means), 20 pipe simulation unit (loop means), 204 root function table (conversion table), fz sound pressure signal (excitation signal)

Claims (5)

(57) [Claims] A loop means having at least a delay means and a low-pass filter; and an excitation signal generation means for generating an excitation signal and supplying the excitation signal to the loop means, and synthesizing a tone signal by extracting the excitation signal. In the musical sound synthesizer, the excitation signal generating means changes a level characteristic of the excitation signal according to a predetermined function with respect to a performance operation amount for increasing a volume of a musical sound to be synthesized, wherein the predetermined function is The level of the excitation signal monotonically increases in response to a change in the direction of increasing the musical tone of the performance operation amount.
The change in the excitation signal in the case of
A tone synthesizer characterized by having characteristics that are constant with respect to fluctuations . 2. The tone synthesis according to claim 1, wherein said excitation signal generating means generates said excitation signal in accordance with a difference between a signal from said loop means and a signal corresponding to a performance operation. apparatus. 3. The musical sound synthesizer according to claim 1, wherein the predetermined function includes at least a root function, and the excitation signal is generated via a conversion table storing the root function. . 4. The predetermined function has a non-linear characteristic in which an input-output characteristic curve is non-linear, and the excitation signal generating means moves the characteristic curve on a coordinate according to the performance operation amount. 3. The musical tone synthesizer according to claim 1, wherein the control is performed as follows. 5. A tone synthesis method comprising: a generation step of generating an excitation signal; and a synthesis step of synthesizing a tone signal by circulating the excitation signal through a loop means having at least a delay means and a low-pass filter. In the generation step, the level characteristic of the excitation signal is changed according to a predetermined function with respect to a performance operation amount for increasing the volume of a musical tone to be synthesized, and the predetermined function is a musical tone of the performance operation amount. The level of the excitation signal monotonically increases in response to a change in the direction in which
The change in the excitation signal in the case of
A tone synthesis method characterized in that it has characteristics that are constant with respect to fluctuations .