JPH0467100A - Musical sound synthesizer - Google Patents

Abstract

PURPOSE:To complicate and fine timbre adjustments without increasing the scale of the above device by leading signals which differ in delay phases out of plural nodes of a closed loop means, adding their multiplication results, and outputting the result as a musical sound signal. CONSTITUTION:A waveform generating circuit 4 generates an initial waveform including many frequency components such as an impulse waveform and this initial waveform is sent from an adder 3 to a delay feedback type musical sound synthesis part to circulate in the loop, thereby generating an attenuated sound. The output signal of the adder and the output signals of delay elements D1 - Dn are supplied to multipliers M0 - Mn are multiplied by coefficients a0 - an corresponding to desired filter arithmetic. Their multiplication results are added by an adder 5, whose addition result is outputted as the musical sound signal. Consequently, the musical sound synthesizer which has a complicate and fine timbre adjusting function is obtained without being increased in device scale.

Description

[Detailed description of the invention]

"Industrial Application Field" The present invention relates to a musical tone synthesis device suitable for use in synthesizing musical tones of natural musical instruments. "Prior Art" A musical tone synthesis device is known that synthesizes natural instrument sounds by operating a model simulating the sound production mechanism of a natural instrument. A musical tone synthesis device for the attenuated sound of a percussed string instrument or a plucked string instrument uses a loop circuit that includes a delay circuit that simulates the propagation delay of vibrations in the strings and a filter that simulates the acoustic loss in the strings, and a loop circuit that includes a filter that simulates the acoustic loss in the strings. Alternatively, a so-called delayed feedback type musical tone synthesizer is known, which includes an excitation circuit that inputs an excitation signal corresponding to excitation vibrations when a string is struck. Note that this type of musical tone synthesis device is disclosed in, for example, Japanese Patent Publication No. 58-58679. In addition, as a musical tone synthesis device for sustained sounds such as wind instruments, for example, a waveguide (bidirectional delay transmission path) that simulates the propagation of air pressure waves in the resonant pipe of a wind instrument,
There is a known configuration in which the resonator is combined with a nonlinear amplification element that simulates the operation of a reed that excites air pressure waves inside the resonant tube. According to this configuration, the output signal of the nonlinear amplification element is transmitted toward the end of the waveguide, and the signal reflected at the end is returned to the nonlinear amplification element through the waveguide in the opposite direction. The action is repeated. That is, the generation and reciprocating propagation of air pressure waves inside an actual wind instrument are faithfully simulated. Then, a target musical tone signal is extracted from an arbitrary node in the waveguide. Note that this type of technology is not disclosed in Japanese Patent Application Laid-Open No. 63-40199. "Problems to be Solved by the Invention" Incidentally, there may arise a demand for applying complex timbre changes to the musical tone signals synthesized as described above. In this case, the above requirements can be met by connecting a filter having a transfer function that allows the desired timbre change to be connected to the output section of a delayed feedback musical tone synthesizer or a musical tone synthesizer using a waveguide. . However, if the desired change in timbre is complex and subtle, it is necessary to provide a high-order filter, which poses a problem in that the entire device becomes a complete mess. It is an object of the present invention to provide a musical tone synthesis device having a complex and delicate tone adjustment function without increasing the scale of the device. ``Means for Solving the Problems J'' The present invention comprises: a closed loop means formed by connecting at least a plurality of delay elements in a closed loop; means for inputting an excitation signal to a predetermined node in the closed loop means; Extracting signals with different delay phases from a plurality of nodes, multiplying each signal by a multiplication coefficient corresponding to the desired filter operation,
It is characterized by comprising timbre adjustment means for adding the multiplication results and outputting the result as a musical tone signal. "Operation" According to the above configuration, the input excitation signal is repeatedly circulated in the closed loop means. Then, f signals with different delay phases are extracted from a plurality of nodes in the closed loop means, each signal is multiplied by a coefficient for filter operation, and the multiplication results are added together and output as a musical tone signal. be done. "Embodiments" Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[Principle of the present invention]

FIG. 1 is a block diagram illustrating the invention in detail. 1 is a delayed feedback type musical tone forming section, each of which has 1
It consists of n delay elements D+-Dn connected in cascade and having a delay time equal to the sampling period, a single filter for adding acoustic loss, a coefficient multiplier for multiplying by an attenuation coefficient, a delay circuit for pitch adjustment, etc. a signal processing circuit 2,
The adder 3 is connected in a closed loop. Then, an initial waveform containing many frequency components, such as an impulse waveform, is generated by the waveform generation circuit 4, and this initial waveform is introduced from the adder 3 into the delayed feedback musical tone synthesis [1] and circulates in the loop. This results in the formation of a damped sound. Output signal of adder 3 and delay element D1
~Dn output signals are respectively given to multipliers M0~Mn to calculate coefficients a0~an corresponding to the desired filter operation.
is multiplied. These multiplication results are then added by an adder 5, and the addition result is output as a musical tone signal. Here, if the transfer function of the signal processing circuit 2 is F (z), the transfer function between the waveform supplied from the waveform generator 4 and the musical tone signal waveform Y obtained from the adder 5 is expressed by the following formula (
+). -zF As shown in the above equation (1), the transfer function of the configuration shown in FIG.
/(1-z F) and the transfer function of an n-th order FIR (finite impulse response) filter, that is, delayed feedback type musical tone synthesis i1,
A transfer function completely equivalent to that obtained when a separate FIR filter is connected is realized. In this way, the means for performing delay processing necessary for filter calculations and the means for performing delay processing in the closed-loop means are shared, making it possible to perform complex and subtle timbre adjustments without increasing the scale of the device. I can do it.

FIRST EMBODIMENT FIG. 2 shows an example of a structure in which the present invention is applied to a musical tone synthesis device for attenuated sounds of pianos, guitars, etc. In this figure, the same reference numerals are given to the parts corresponding to those in FIG. This musical tone synthesis device is supplied with a key-on signal KON generated by operating a performance operator (for example, a keyboard, etc.) not shown, and pitch information PITCH specifying the pitch of a note to be produced. In FIG. 2, the variable delay circuit 14 includes a selector 6 and a soft register formed by cascading delay elements DD to DDm having a delay time of one sampling period. The output of the delay element Dn is input to the first stage delay element DD in the variable delay circuit 14, and each delay output of the delay elements DD to DDm is input to the selector 6. Then, the pitch information PITC is controlled by the coefficient control wI7.
) (delay specification information corresponding to is generated, and its integer part k
is given to the selector 6 as selection information. As a result, the delay output corresponding to k is selected from among the delay outputs of delay elements DD, -DDm and output. That is, if one sampling period is τ, the delay element D
The output of n is delayed by τ, outputted from selector 6, and sent to variable delay circuit 15. In this example,
Since the delay elements D1 to Dn are inserted as fixed delay circuits, the total delay time of the delay elements D1 to Dn is subtracted from the reciprocal of the musical tone frequency corresponding to the pitch information PITCH to obtain the delay time of 1. Specified by delay specification information. In the variable delay circuit 15 , the output signal from the selector 6 is input to the multiplier 10 , and is also input to the multiplier 9 after being delayed by one sampling period by the delay element 8 . These multipliers 9 and IO are given the fractional parts 1 and 1-b of the delay designation information as multiplication coefficients, respectively. Then, the outputs of multipliers 9 and 10 are added by adder 11 and output. According to this variable delay circuit 15, I
Linear interpolation is performed on two signals input before and after the sampling period, and the signal input to the delay circuit 8 is effectively delayed by bτ and output from the adder 11. Each frequency component of the output signal of the adder 11 is attenuated according to the filter characteristics of the low-pass filter 12. Then, the output signal of the low-pass filter 12 is multiplied by an attenuation coefficient 7 determined according to a desired attenuation envelope by a multiplier 13, and the result of this multiplication is fed back to the adder 3. The waveform generator 4a is realized by, for example, a ROM (read only memory), and stores a sample string of waveform values of an initial waveform. When considering the stability of timbre control and pitch control, it is desirable to change the frequency of the initial waveform depending on the pitch of the musical tone to be generated. Therefore, the entire generated sound range is divided into, for example, octaves, and an initial waveform with an optimum frequency is stored for each division obtained by the division. In such a configuration, when a performance operator is operated and a key-on signal KON and pitch information PITCH are generated, an initial waveform corresponding to the pitch information PITCH is read out from the waveform generating section 4a. Then, this successive waveform is calculated by the adder 3-delay circuit, ~Dn=variable delay circuit I
4-variable delay circuit 15-low-pass filter 12-multiplier 13, and thereafter circulates within the closed loop while being attenuated. Then, the output signal of the adder 3 and the delay elements D1 to Dn
Each delayed output is multiplied by coefficients a0 to an for filter calculation, and the multiplication results are added together and output as a tone change signal.

[Second Embodiment] FIG. 3 shows a configuration example in which the present invention is applied to a musical tone synthesis device for a clarinet. In the figure, an excitation circuit 20 corresponds to a mouthpiece portion of a clarinet, and a resonance circuit 40 corresponds to a resonance tube of a clarinet. Moreover, the junk switch 30 inserted between the excitation circuit 20 and the resonance circuit 40 is
This simulates the scattering of air pressure waves at the connection between the mouthpiece and the resonance tube. At this junction 30, the output signal from the resonant circuit 40 and the output signal from the excitation circuit 10 are added by an adder 31 and input to the resonant circuit 40, and the output signal from the adder 31 and the output signal from the resonant circuit 4
The output signals of 0 are added by an adder 32 and input to the excitation circuit 20. The excitation circuit 20 includes a subtracter 21, filters 22 and 23
, adder 24, ROM 25, multipliers 26, 27 and I
It consists of NV. When a musical tone is generated, a signal P specifying the blowing pressure and a signal E specifying the blowing pressure (the pressure when the mouthpiece is held in the mouth) are given from the coefficient control section 50. The subtracter 21 receives a signal inputted from the resonant circuit 40 via the junction 30, that is, a signal P corresponding to the air pressure of the air pressure wave reflected within the resonance tube or at the end of the resonance tube and arriving at the mouthpiece section.
R and a signal P corresponding to the blowing pressure are input. and,
The signal P is subtracted from the signal PR by the subtractor 21, resulting in a signal PA corresponding to the air pressure applied to the reed of the clarinet.
is obtained. The output signal PA of the subtracter 21 is band-limited by the filter 22. This filter 22 is composed of a first-order low-pass filter, and includes an excitation circuit 20 and a resonant circuit 4.
This is inserted to prevent the amplitude of the signal circulating between 0 and 0 from becoming significantly large at a specific frequency. The output signal P of the filter 22 is output from the filter 237.
At the same time, the signal is inverted by the multiplier INV and input to the multiplier 26. The signal P passes through the filter 23 to remove high frequency components. This results in
The response characteristics of a reed absorbing sudden pressure changes are simulated. Then, the adder 24 adds the signal E corresponding to the engine pressure to the output signal P of the filter 23 to obtain a signal P3 corresponding to the pressure actually applied to the lead. This signal P is then given to the ROM 25 as an address. As a result, a table of nonlinear functions stored in advance in the ROM 15 is referred to, and a signal Y corresponding to the cross-sectional area of the gap between the reed and the mouthpiece portion, that is, the admittance to the airflow is output. Then, the signal Y and the signal -Pl are multiplied by the multiplier 26 to obtain a signal FL corresponding to the flow velocity of air passing through the gap between the reed and the mouthpiece portion. Then, the signal PL is multiplied by the multiplication coefficient G by the multiplier 27. Here, the multiplication coefficient G is a constant determined according to the pipe diameter near the mouthpiece section of the resonance tube J, and corresponds to the difficulty of air flow, that is, the impedance to the air flow. It is something. Therefore, the multiplier 27 provides a signal corresponding to the change in air pressure occurring at the mouthpiece side entrance of the resonance tube. This signal is then input to the resonant circuit 40 via the noise junction 30. In the resonant circuit 40, the delay elements ~Dn
Reeds and the first tone hole in the resonance tube
The delay that occurs when air pressure waves propagate back and forth between the two is simulated. Jiankuntan JUI simulates the scattering of air pressure waves in the vicinity of the first tone hole counted from the lead side, and in the case of this embodiment, the multiplier MJ
, ~MJ adder AJ, and a 4-multiplying grid consisting of AJt. Here, the multiplier MJ,
-MJ, is a multiplication coefficient determined according to the open/closed state of the target tone hole, , , 1- k, , -
k, are each given by the coefficient control section 50. At the subsequent stage of the Junk Shicon JU, delay circuits DL+ to DLff having a delay time corresponding to the length between each tone hole in the resonance tube are cascaded via the Junk Shicon JU having the same configuration as the Junk Shicon JU. Connected (as shown). Then, the output of the final stage delay circuit DL& is band-separated by a low-pass filter 41,
Furthermore, it is multiplied by a negative coefficient δ by the multiplier 42 and fed back to the subtracter 21 of the excitation circuit 20 via the cascade-connected junctions and further via the junction 30. In such a configuration, when an operator (not shown) is operated to issue an instruction to generate a musical tone, signals corresponding to the blowing pressure P and the embouchure E are generated by the coefficient control section 50 and supplied to the excitation circuit 20. Also, corresponding to the open/closed state of each tone hole, 1. +, 1 k+,
kl, ~, kl2. ki! , 1-k(!, - are determined and supplied to the respective corresponding junctions. Then, an excitation signal based on the blowing pressure P and the engine E is output from the excitation circuit 20. This signal is The signal inputted via the resonance circuit 30 is then turned back at various points within the resonance circuit 40, and is then input to the resonance circuit 40 again through the resonance circuit 30.
0 to the excitation circuit 20. Then, in the excitation circuit 20, the above-mentioned signal processing is performed based on the signal returned from the resonance circuit 40, the blowing pressure P and the engine E at the relevant time, and a new excitation signal is input to the resonance circuit 40 via the noyanction 30. be done. Similar operations are performed thereafter, and signals are circulated between the excitation circuit 20 and the resonant circuit 40. Then, multipliers M0 to Mn apply coefficients a for filter calculation to the input signals of the delay elements in the resonant circuit 40 and the delay outputs of the delay elements D1 to Dn. -an is multiplied, and the results of each multiplication are added by an adder 5 and output as a musical tone signal. In the above embodiment, an example was explained in which one FIR filter for tone adjustment was provided in the musical tone synthesis device, but it may be provided in a plurality of locations. Further, in this case, the outputs of the plurality of FIR filters may be mixed and output. Also,
The number of stages of the delay circuit used in common with the closed loop means (delayed feedback musical tone forming section and resonance circuit in the above embodiments) to configure the FIR filter corresponds to the order of the filter operation required and can be arbitrarily determined. . Furthermore, if the frequency of the generated musical tone is high, it may be difficult to reduce the total delay time of the closed loop means accordingly. However, in this case, by lowering the order of the filter operation and reducing the number of stages of the delay circuit used as the FIR filter, it is possible to make the total delay time of the closed loop means a value suitable for the desired pitch. "Effects of the Invention" As explained above, according to the present invention, there is provided a closed loop means formed by connecting at least a plurality of delay elements in a closed loop, a means for inputting an excitation signal to a predetermined node in the closed loop means, extracting signals with different delay phases from a plurality of nodes in the closed loop means;
Since each signal is multiplied by a multiplication coefficient corresponding to the desired filter operation, and the multiplication results are added together and a timbre adjustment means is provided to output the result as a musical tone signal, it is possible to reduce the complexity of the device without increasing the size of the device. In addition, there is an effect that a musical tone synthesis device capable of performing subtle tone color adjustments can be realized.

[Brief explanation of the drawing]

FIG. 1 is a block diagram illustrating the basic configuration of the present invention, FIG. 2 is a block diagram showing the configuration of a musical tone synthesizer according to a first embodiment of the present invention, and FIG. 3 is a block diagram illustrating the configuration of a musical tone synthesis device according to a second embodiment of the present invention FIG. 2 is a block diagram showing the configuration of a synthesis device. l...Delayed feedback type musical tone forming section, D, ~D
n... 1 sampling period delay element, M0 to Mn multiplier, 5... adder.

Claims (1)

[Scope of Claims] Closed loop means comprising at least a plurality of delay elements connected in a closed loop; means for inputting an excitation signal to a predetermined node in the closed loop means; and means for inputting an excitation signal to a predetermined node in the closed loop means; Taking out mutually different signals and multiplying each signal by a multiplication coefficient corresponding to the desired filter operation,
A musical tone synthesis device characterized by comprising a timbre adjustment means for adding the multiplication results and outputting the result as a musical tone signal.