JP3617330B2 - Musical sound synthesizer, musical sound synthesis method and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a musical sound synthesizer, a musical sound synthesis method, and a recording medium suitable for use in application programs in electronic musical instruments, amusement devices (karaoke, game devices, etc.), personal computers, and the like, and in particular, stringed musical instruments such as violins and violas. The present invention relates to a musical tone synthesizer suitable for use in simulation, a musical tone synthesis method, and a recording medium.
[0002]
[Prior art]
In order to automatically synthesize bowed instrument sounds such as violins and violas, physical model sound sources that simulate the behavior of strings and bows are known (Japanese Patent Publication No. 2719655, Japanese Patent Publication No. 7-21715, etc.). In these physical model sound sources, a closed loop circuit that simulates the behavior of a string and an excitation unit that supplies an excitation signal to the closed loop circuit to simulate the operation of rubbing the string with a bow are provided. It has been.
[0003]
By the way, when actually playing a violin, viola, etc., it is common to apply pine resin powder to the bow so that the frictional relationship between the bow and the string is uniform and stable. Here, when heat is generated by the stringing operation, the pine resin is melted and partially liquefied or nearly liquefied, or liquefaction and solidification are repeated. This behavior of rosin produces a subtle change in the frictional relationship between the bow and string.
[0004]
[Problems to be solved by the invention]
However, the conventional physical model sound source cannot simulate the behavior of the above-mentioned pine resin, and it is difficult to perform precise modeling of the bowed instrument.
The present invention has been made in view of the above-described circumstances, and it is a first object of the present invention to provide a musical tone synthesizer, a musical tone synthesis method, and a recording medium that can reproduce subtle nuances by precise modeling of a bowed instrument. It is said. Another object of the present invention is to obtain a new tone color that cannot be obtained with an actual bowed instrument or a conventional physical model sound source.
[0005]
[Means for Solving the Problems]
In order to solve the above-mentioned problems, the present invention includes a delay means, and a closed loop circuit that circulates a waveform signal through the delay means, a waveform signal is extracted from the closed loop circuit, and the extracted waveform signal and performance operation signal are and coupling the signal generating means for generating a combined signal based on the excitation means for applying to said closed loop circuit and converts the combined signal to the excitation signal, obtained on the basis of the waveform signal and the excitation signal taken out from the closed loop circuit Characteristic control signal generating means for integrating the obtained values to generate a characteristic control signal, and conversion characteristic setting means for setting a conversion characteristic between the coupling signal and the excitation signal in the excitation means based on the characteristic control signal; A musical sound synthesizer is provided.
The present invention also provides a delay process for delaying the waveform signal, a combined signal generating process for generating a combined signal based on the performance operation signal and the waveform signal delayed in the delay process, and the combined signal as an excitation signal. A synthesis process for converting and synthesizing the waveform signal to be subjected to the delay process using the excitation signal , and integrating the waveform signal delayed in the delay process and the value obtained based on the excitation signal A characteristic control signal generating process for generating a control signal; and a conversion characteristic setting process for setting a conversion characteristic between the combined signal and the excitation signal in the synthesis process based on the characteristic control signal. A musical sound synthesis method is provided .
In favorable preferable embodiment, the transformation characteristic setting process is the process of determining the amplitude of said excitation signal based on said characteristic control signal.
In another preferred aspect, the characteristic control signal generating step includes a characteristic control signal that reduces the amplitude of the excitation signal as the result of integrating the values obtained based on the waveform signal and the excitation signal increases. comprising generating amplitude control over enough or al.
The present invention also provides a computer-readable recording medium recording a program that causes a computer to execute any of the above-described methods.
[0006]
DETAILED DESCRIPTION OF THE INVENTION
1. Configuration of Embodiment 1.1. Performance controller 1
Next, the configuration of an embodiment of the present invention will be described with reference to FIG.
In the figure, reference numeral 1 denotes a performance operator, which has a form with a bowed musical instrument, and outputs performance information in response to a user operation. Such performance information includes bow pressure Pb, bow speed Vb, tone color TC, and pitch PITCH. Such a performance operator is disclosed in, for example, Japanese Patent Laid-Open No. 5-006169 or Japanese Patent Laid-Open No. 10-078778.
[0007]
1.2. Linear part 3
Next, reference numeral 3 denotes a linear portion, and a closed loop circuit for simulating the behavior of the string of a bowed instrument is formed by the delay circuits 12, 13, 15, 16, the end filters 11, 14, and the adders 17, 19. . A velocity signal indicating the string velocity is propagated in the closed loop circuit. In an actual bowed instrument, when vibration occurs at the contact point between the bow and the string, a traveling wave is generated toward the support points (two points) of the string. Each of the delay circuits 13 and 16 has a delay time corresponding to the delay time until this traveling wave reaches both support points.
[0008]
Further, in the bowed instrument, the traveling wave that has reached the two support points of the string is reflected while the polarity is reversed, and propagates toward the contact point between the bow and the string. Each of the delay circuits 12 and 15 has a delay time corresponding to the delay time until these reflected waves reach the contact point between the bow and the string. Further, the traveling wave and the reflected wave in the bowed instrument are attenuated in the process of propagating in the string and being reflected at the support point. The end filters 11 and 14 invert the polarity of the traveling wave as described above, and have pass characteristics that simulate attenuation in the strings and at the support points and their frequency characteristics.
[0009]
The adder 18 adds the velocity signals output from the delay circuits 12 and 15 and outputs the vibration velocity of the string obtained by synthesizing the reflected waves from both ends of the string. An excitation signal DRIVE (details will be described later) output from the excitation unit 4 is applied to the adders 17 and 19 via the adder 20. This simulates the speed applied from the bow to the string. Here, the adders 20 and 21 simulate the mutual interference between the excitation signal DRIVE and the output signal (string vibration speed) of the adder 18.
[0010]
1.3. Excitation unit 4 and subtractor 7
Next, 7 is a subtracter, which subtracts the bow speed Vb from the speed signal RETURN output from the adder 21, and outputs a virtual relative speed RV between the bow and the string. The details of the excitation unit 4 are shown in FIG. In the figure, 41 is a dynamic friction table, which stores the force applied to the string by the bow when the bow and string are in a sliding state (dynamic friction state) in correspondence with the virtual relative speed RV. Reference numeral 42 denotes a static friction table, which stores the force applied to the string by the bow in correspondence with the virtual relative speed RV when the bow and the string are in close contact with each other (static friction state).
[0011]
However, since the force applied to the string eventually appears as a change in the string speed, the dimensions in the tables 41 and 42 are “speed”. A coefficient generator 43 outputs a coefficient A corresponding to dynamic friction and a coefficient B corresponding to static friction in correspondence with the virtual relative speed RV. The coefficient A is a coefficient that fluctuates in the range of “0” to “1”, and is “0” when the virtual relative speed RV is low, “1” when the virtual relative speed RV is high, and “1” near the crossover point XAB. It changes linearly from “0” to “1”. The coefficient B is a value obtained by subtracting the coefficient A from “1”. Here, the crossover point XAB is set so as to increase as the bow pressure Pb increases.
[0012]
Reference numeral 40 denotes a non-linear transition switching unit, which is a multiplier 44 that multiplies the output signal of the dynamic friction table 41 and the coefficient A, a multiplier 45 that multiplies the output signal of the static friction table 42 and the coefficient B, multipliers 44, And an adder 46 for adding 45 results. Since the sum of the coefficients A and B is always “1”, the non-linear transition switching unit 40 switches the force (speed) applied to the string in the dynamic friction state and the static friction state while cross-fading. Next, a multiplier 47 multiplies the output signal of the nonlinear transition switching unit 40 and the control signal R_cont supplied from the excitation control unit 5, and outputs the multiplication result as the excitation signal DRIVE.
[0013]
1.4. Control unit 2
When the performance information is supplied from the performance operator 1, the control unit 2 sets various parameters in the linear unit 3 accordingly. That is, the control unit 2 sets the delay times DL1, DL2, DR1, DR2 in the delay circuits 12, 13, 15, 16 based on the pitch PITCH, and the filter characteristics TFL in the termination filters 11, 14 based on the tone color TC. Alternatively, the distribution ratio of the delay times DL1, DL2, DR1, DR2 is set. Further, the control unit 2 sets the contents of the tables 41 and 42 of the excitation unit 4 according to the tone color TC.
[0014]
1.5. Excitation control unit 5
Next, the details of the excitation control unit 5 will be described with reference to FIG. In the figure, 61 and 62 are multipliers, and signals LOOP_SIG1 and LOOP_SIG2 are supplied to one end of each. Here, as shown in FIG. 1, the speed signal RETURN is used as the signal LOOP_SIG1, and the excitation signal DRIVE is used as the signal LOOP_SIG2. Multipliers 61 and 62 multiply the signals LOOP_SIG1 and LOOP_SIG2 by level adjustment constants S1 and S2, respectively. Next, the output signals of the multipliers 61 and 62 are multiplied by the multiplier 63, and the absolute value conversion circuit 64 calculates the absolute value of the multiplication result.
[0015]
Here, since the excitation signal DRIVE corresponds to the velocity applied to the string, it correlates with the magnitude of the friction between the bow and the string. Therefore, it is considered that a value corresponding to the heat energy per unit time contributing to heat generation can be obtained by multiplying the excitation signal DRIVE by the speed signal RETURN and obtaining the absolute value thereof.
[0016]
An integration processing unit 65 integrates the output signal of the absolute value conversion circuit 64 to generate a signal corresponding to the temperature rise in the bow. Next, reference numeral 66 denotes a coefficient conversion unit, which is provided with a plurality of (n types) different tables that define input / output characteristics, and performs conversion processing on the output signal of the integration processing unit 65 based on these tables. .
[0017]
As shown in FIG. 4, the tables in the coefficient conversion unit 66 have characteristics that the output signal level decreases as the input signal level increases. The output signal of the coefficient conversion unit 66 is a signal corresponding to the friction coefficient between the bow and the string, and thereby the behavior in which the friction coefficient between the bow and the string decreases as the temperature rises is simulated.
[0018]
By the way, it is considered that the friction coefficient between a bow and a string in an actual stringed instrument does not simply change according to a temperature rise, but includes a noise component according to coating unevenness or roughness of pine resin. Therefore, a fluctuation signal generating unit 69 and a fluctuation signal processing unit 70 are provided in order to give fluctuation based on such noise components. The fluctuation signal generation unit 69 generates a fluctuation signal according to parameters such as a waveform and noise characteristics. The algorithm for obtaining the contents of these parameters and the fluctuation signal may be freely set according to the contents of the tone color settings and the contents of the performance operation.
[0019]
Next, the fluctuation signal processing unit 70 performs a filtering process on the output signal of the fluctuation signal generation unit 69 and outputs the result. The contents of this filtering process are determined based on preset parameters such as filter characteristics. Next, reference numeral 67 denotes a multiplier array, which is composed of n multipliers, and multiplies the output signals of the coefficient conversion unit 66 by the output signals of the fluctuation signal processing unit 70, whereby the friction coefficient fluctuations. Is simulated.
[0020]
Reference numeral 68 denotes a multiplier array, which multiplies the output signals of the multiplier array 67 by level adjustment constants S31 to S3n. The multiplication result is output as n types of control signals R_cont (1) to (n). As described above, the multiplier 47 in the excitation unit 4 multiplies the output signal of the nonlinear transition switching unit 40 by the control signal R_cont. The control signal R_cont is controlled by the control signals R_cont (1) to (n). ) Is selected.
[0021]
1.6. Sound generator 6
In a bowed instrument, the vibration of the string is propagated to the front plate through the piece, and further propagated to the back plate through the internal soul column, causing the entire body to resonate. Reference numeral 6 denotes a sound generation unit, which performs a filtering process on the output signal of the delay circuit 15 and amplifies and sounds the result. By this filtering process, the behavior of the resonance part and the like in the bowed instrument is simulated.
[0022]
2. Operation of Embodiment Next, the operation of this embodiment will be described.
When the user operates the performance operator 1, the tone color TC and the pitch PITCH are supplied to the control unit 2, the bow pressure Pb is supplied to the excitation unit 4, and the bow speed Vb is supplied to the subtractor 7. Based on the tone color TC and the pitch PITCH, the delay times DL1, DL2, DR1, DR2 and the filter characteristic TFL in the linear portion 3 are determined.
[0023]
When the bow speed Vb is supplied to the subtractor 7, the result of subtracting the bow speed Vb from the speed signal RETURN is supplied to the excitation unit 4 as a virtual relative speed RV. Here, since the velocity signal RETURN is “0” in the initial state, the inverted signal of the bow velocity Vb is supplied to the excitation unit 4 as it is. A speed signal corresponding to the force applied to the string is output via the tables 41 and 42 and the non-linear transition switching unit 40.
[0024]
Next, the multiplier 47 multiplies the output speed signal by the control signal R_cont. Since the integration result of the integration processing unit 65 in the excitation control unit 5 is “0” in the initial state, the control signal R_cont is at a relatively high level (see FIG. 4). Accordingly, a relatively high level excitation signal DRIVE is supplied to the closed loop circuit of the linear section 3 via the adders 17 and 19.
[0025]
In the initial state, the speed signal output from the delay circuits 12 and 15 in the closed loop circuit is “0”. For this reason, the excitation signal DRIVE is supplied to the delay circuits 13 and 16 as a velocity signal that forms a traveling wave as it is. These speed signals are delayed through the delay circuits 13 and 16, subjected to filtering processing through the termination filters 11 and 14, and then added through the delay circuits 12 and 15 as speed signals forming reflected waves. 17 and 19.
[0026]
These speed signals are combined in the adder 18 and then fed back to the subtractor 7 as a speed signal RETURN. Accordingly, the bow speed Vb at the time of feedback from the speed signal RETURN is subtracted, and this result is supplied to the excitation unit 4 as a new virtual relative speed RV, and the excitation signal DRIVE based on this virtual relative speed RV is added to the adder 20. , 17 and 19 are supplied to the closed loop circuit again.
[0027]
When the above processing is repeated with the parameters such as the bow speed Vb being constant, each speed signal in the closed loop circuit, the speed signal RETURN, and the excitation signal DRIVE eventually become standing waves having a constant and stable pitch. . When the speed signal in the closed loop circuit is supplied to the sound generator 6, a tone having this pitch is generated.
[0028]
Incidentally, the speed signal RETURN and the excitation signal DRIVE are supplied to the excitation control unit 5 as signals LOOP_SIG1 and LOOP_SIG2, respectively. As a result, a value equal to or greater than “0” is continuously output from the absolute value conversion circuit 64 via the multipliers 61, 62, 63 and the absolute value conversion circuit 64, and the integration result in the integration processing unit 65 is obtained. Gradually rises.
[0029]
When the integration result eventually exceeds the liquefaction start point MP shown in FIG. 4 , the level of the excitation signal DRIVE gradually decreases. Therefore, in the generated musical sound, the influence of the bow pressure Pb and the bow speed Vb gradually decreases, and the vibration inherent to the string simulated by the closed loop circuit is changed to a emphasized sound.
[0030]
3. As described above, according to the present embodiment, the excitation signal DRIVE can be gradually reduced based on the integration result of the signals LOOP_SIG1 and LOOP_SIG2. It is possible to faithfully simulate the behavior of a bowed instrument in which the coefficient of friction gradually decreases, and to reproduce subtle nuances in the tone of the bowed instrument. Furthermore, by changing various parameter settings, it is possible to obtain interesting new timbres that could not be obtained with an actual bowed instrument or a conventional physical model sound source.
[0031]
4). Modifications The present invention is not limited to the above embodiment, and various modifications are possible as follows, for example.
(1) In the above embodiment, the frequency characteristics of the strings are simulated by the termination filters 11 and 14, but the delay circuits 12, 13, 15, and 16 are configured by distributed constant circuits, and the delay and frequency characteristics of the strings are obtained. A more accurate simulation may be performed by simulating.
[0032]
(2) In the above embodiment, the elements such as the delay circuit and the arithmetic circuit are configured by hardware. However, some or all of these elements may be realized by software. Of course, if all the configurations shown in the hardware block diagram are realized by a program, the present invention can be realized by a general-purpose personal computer or the like. In such a case, the present invention can also be realized as a recording medium storing a program describing the operation of each element.
[0033]
(3) In the above embodiment, the output results of the dynamic friction table 41 and the static friction table 42 are cross-faded in the non-linear transition switching unit 40. However, the non-linear transition switching unit 40 is configured by a simple switch, and the output results of both tables are simply used. The configuration may be simplified by switching between.
[0034]
(4) Further, in the non-linear transition switching unit 40, further fine adjustment can be performed by adjusting the output signals of the tables 41 and 42 independently. An example is shown in FIG. In the figure, a multiplier 53 multiplies the control signal R_cont and a level adjustment constant S3. A multiplier 52 is inserted between the adder 46 and the multiplier 44, and a multiplication result of the output signal of the multiplier 44 and the output signal of the multiplier 53 is supplied to the adder 46.
[0035]
Similarly, the multiplier 55 multiplies the control signal R_cont and the level adjustment constant S4. A multiplier 54 is inserted between the adder 46 and the multiplier 45, and a multiplication result of the output signal of the multiplier 45 and the output signal of the multiplier 55 is supplied to the adder 46. In such a configuration, the magnitude of the influence of the control signal R_cont can be set independently for both static friction and dynamic friction by setting the constants S3 and S4 independently.
[0036]
Further, the control signals R_cont supplied to the multipliers 53 and 55 are not limited to the same one, and any one of the control signals R_cont (1) to (n) (see FIG. 3) is arbitrarily and independently selected. Needless to say, the multipliers 53 and 55 may be supplied.
[0037]
(5) In the actual behavior of the bow and string, the point at which the transition from the static friction state to the dynamic friction state does not necessarily coincide with the point at which the transition from the dynamic friction state to the static friction state, and hysteresis as shown in FIG. Has characteristics. In order to simulate such behavior, positive feedback may be applied to the input terminals of the tables 41 and 42 and the coefficient generator 43 from the output signal of the nonlinear transition switching unit 40.
[0038]
FIG. 6 shows an example in which such positive feedback is performed. In the figure, 56 is a multiplier, which multiplies the output signal of the nonlinear transition switching unit 40 by a predetermined feedback constant PFB. An adder 51 is inserted between the subtractor 7, the tables 41 and 42, and the coefficient generator 43, and adds the output signal of the multiplier 56 to the virtual relative speed RV, and the result is corrected virtual relative speed. Output as RV '. As a result, the hysteresis characteristic described above can be applied to the excitation signal DRIVE.
[0039]
(6) In FIG. 6, the positions of the multiplier 56 and the adder 51 may be changed. For example, as shown by the broken line in the figure, the multiplier 56 multiplies the coefficient A output of the coefficient generator 43 and the feedback constant PFB, and the multiplication result is positively fed back to the input of the coefficient generator 43 by the adder 51. Good. Further, in order to change the feedback constant PFB, the feedback constant PFB may be controlled using some control signal R_cont. For example, since it is known that the hysteresis increases as the temperature rises between the actual bow and string, it responds to changes in the contact surface temperature between the bow and string due to the movement and frictional relationship of the bow and string. It is preferable to generate a control signal to control the feedback constant PFB.
[0040]
(7) In the above embodiment, the integration processing unit 65 is used to generate a signal corresponding to the temperature rise in the bow, but a low-pass filter or the like may be used instead of the integration processing unit 65, and the low-pass filter and the integration circuit And may be used in combination.
[0041]
(8) In the above embodiment, fluctuations are given by multiplying the output signals of the fluctuation signal processing unit 70 by a plurality of output signals of the coefficient conversion unit 66, but the calculation for giving fluctuations is multiplication. It goes without saying that it is not limited. For example, it goes without saying that the multiplier array 67 may be constituted by an arithmetic unit array that performs addition, subtraction, multiplication, or a more complicated calculation.
[0042]
(9) In the above embodiment, the speed signal RETURN is used as the signal LOOP_SIG1 supplied to the excitation control unit 5, but a virtual relative speed RV or the like may be used instead.
[0043]
【The invention's effect】
As described above, according to the present invention, the characteristic control signal is generated based on the waveform signal and the performance operation signal, and the conversion characteristic between the coupling signal and the excitation signal is set based on the characteristic control signal. In addition, it is possible to obtain a new tone that cannot be obtained with an actual bowed instrument or a conventional physical model sound source.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an overall configuration of an embodiment.
FIG. 2 is a detailed block diagram of an excitation unit 4;
FIG. 3 is a detailed block diagram of an excitation control unit 5;
4 is an input / output characteristic diagram of a coefficient converter 66. FIG.
FIG. 5 is a friction characteristic diagram in a modified example of the excitation unit 4;
FIG. 6 is a block diagram of a modification of the excitation unit 4;
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Performance operator, 2 ... Control part, 3 ... Linear part, 4 ... Excitation part, 5 ... Excitation control part, 6 ... Sound generation part, 7 ... Subtractor, 11, 14 ... Termination Filter, 12, 13, 15, 16 ... delay circuit, 17-21 ... adder, 40 ... nonlinear transition switching unit, 41 ... dynamic friction table, 42 ... static friction table, 43 ... coefficient generation unit, 44, 45 ... multiplier, 46 ... adder, 47 ... multiplier, 51 ... adder, 52-56 ... multiplier, 61-63 ... multiplier, 64 ... absolute value conversion circuit, 65... Integration processing unit, 66... Coefficient conversion unit, 67 and 68... Multiplier array, 69 .. fluctuation signal generating unit, and 70.

Claims (5)

A closed loop circuit including delay means and circulating the waveform signal through the delay means;
Combined signal generation means for extracting a waveform signal from the closed loop circuit and generating a combined signal based on the extracted waveform signal and the performance operation signal;
Excitation means for converting the combined signal into an excitation signal and providing it to the closed loop circuit;
A characteristic control signal generating means for generating a characteristic control signal by integrating a value obtained based on the waveform signal extracted from the closed loop circuit and the excitation signal ;
A musical tone synthesizer comprising: conversion characteristic setting means for setting a conversion characteristic between the combined signal and the excitation signal in the excitation means based on the characteristic control signal. A delay process for delaying the waveform signal;
A combined signal generating process for generating a combined signal based on the performance operation signal and the waveform signal delayed in the delay process;
Converting the combined signal into an excitation signal and using the excitation signal to synthesize a waveform signal to be subjected to the delay process;
A characteristic control signal generation process for generating a characteristic control signal by integrating the waveform signal delayed in the delay process and a value obtained based on the excitation signal ;
A musical tone synthesis method comprising a conversion characteristic setting step of setting a conversion characteristic between the combined signal and the excitation signal in the synthesis step based on the characteristic control signal. Before Symbol conversion characteristic setting process, the tone synthesis method according to claim 2, wherein the based on the characteristic control signal is the process of determining the amplitude of the excitation signal. Before SL characteristic control signal generation process,
In that it consists of the waveform signal and the excitation signal and amplitude control over degree <br/> result obtained by integrating the value obtained for generating a characteristic control signal and a small amplitude of Daito indeed the excitation signal on the basis of 3. A musical sound synthesis method according to claim 2, wherein 請 Motomeko 2-4 either method a computer-readable recording medium recording a program for causing a computer to execute the described.

Images