JP2820205B2 - Music 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 musical tone synthesizing apparatus, and more particularly to a musical tone synthesizing apparatus capable of simulating vibration of a soundboard provided in a piano of a natural musical instrument.

[0002]

2. Description of the Related Art Conventionally, there is a sound source circuit technology for simulating a natural musical instrument such as a piano with a physical model sound source. For example, Japanese Patent Publication No. 58-48109 discloses a tone synthesizer having closed loop means in which delay means are connected in a closed loop and outputting a signal circulating through the closed loop means as a tone signal. This is because a tone waveform is added to an input terminal of a filter having a desired frequency characteristic, and a tone waveform appearing at an output terminal of the filter is delayed by delay means and added again to an input terminal of the filter so that the tone waveform is changed. The closed loop means is circulated, and a musical sound waveform that changes sequentially according to the characteristics of the filter each time it passes through the filter is extracted. According to this,
It is possible to simulate a musical tone similar to a musical tone of a natural musical instrument utilizing vibration of a tube, a string or a membrane.

[0003]

However, in the above-described conventional tone synthesizer, only the vibration of a tube, a string, or a membrane in a natural musical instrument is simulated.
On the other hand, in the piano of a natural musical instrument, there is a phenomenon that vibration of a string is transmitted to a soundboard and the soundboard vibrates, and the vibration of the soundboard is transmitted to the string in reverse. In the above-mentioned prior art, it was not possible to imitate such a phenomenon in a natural musical instrument.

[0004] The present invention relates to a musical sound synthesizer using a physical model sound source that simulates vibration of a tube, a string or a membrane.
By imitating not only the vibration of the tube, string or membrane, but also the vibration of the members accompanying it (for example, the soundboard of a piano)
It is an object of the present invention to faithfully reproduce a phenomenon in a natural musical instrument, thereby generating a high-quality tone signal.

[0005]

To achieve this object, the invention according to claim 1 includes a delay means in a closed loop.
A plurality of closed loop means whose signal delay characteristics of the closed loop correspond to the pitch are independently provided for each pitch , and these closed loops are provided.
A tone synthesizer for generating a tone waveform signal of a designated pitch by means , provided in each of the closed loops;
Input operation information according to the user's operation, and input the operation information
Accordingly, the characteristics of the cyclic signal looping through each closed loop
A plurality of loop characteristic changing means for changing the characteristics;
An excitation signal is input from outside the loop means, and the excitation signal and the
By synthesizing the cyclic signal that has passed through the loop characteristic changing means,
Cyclic signal synthesizing means to be input to the loop characteristic changing means,
Based on signals output from the plurality of closed loop means
Means for obtaining a combined output signal;
Of which, the closed loop means corresponding to the pitch at which the pronunciation was instructed
Means for inputting the excitation signal, and
Means for re-inputting the synthesized output signal to each of the closed loop means.
And a step .

[0006]

When a drive waveform is input to the closed loop means, the waveform circulates through the closed loop means, and a signal is output from an appropriate position in the closed loop means. The signal that loops through the closed loop means simulates the vibration of a tube, string or membrane of a natural musical instrument, the pitch of which is determined by the delay time of the loop of the closed loop means. Furthermore, the tone synthesis of the present invention
The device is assumed to have a plurality of closed loop means
After the tone signals output from the plurality of closed loop means are synthesized, the characteristics of the synthesized signal are changed and re-input to each of the plurality of closed loop means. This is because the musical tone output by the vibration of the tube, string or membrane of a natural musical instrument causes members such as the soundboard to vibrate.
It simulates affecting the vibration of each tube, string or membrane. What kind of characteristic change is performed and re-input to the closed loop means defines, for example, a relationship between a string and a soundboard in a natural musical instrument piano.

[0007]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a block diagram of an electronic piano to which a tone synthesizer according to an embodiment of the present invention is applied.

The electronic piano of this embodiment has an 88-key keyboard (keyboard) 2 and a damper pedal 4 as a damper operator.

The control unit 6 detects depression of each key of the keyboard 4 and operation of the damper pedal 4, and outputs key-on signals KON1 to KON88 and a damper operation amount DMPR for each key according to the detection result. Further, the control unit 6 generates delay times DLY1 to DLY88 and string characteristic filter coefficients STRG1 to STRG88, which are parameters for synthesizing a tone corresponding to the pitch corresponding to each key and a musical tone having the characteristic, and outputs them to the sound source unit 8. . Further, the control unit 6 includes a sound source unit 8
Is output.

Reference numeral 8 denotes a sound source section (musical sound waveform output means).
The sound source unit 8 includes 88 sound sources 8-1 to 8-8 corresponding to each key.
8 18 is an information conversion unit. Information converter 18
Receives the damper operation amount DMPR and the key-on signals KON1 to KON88 from the control unit 6 and applies the damper characteristic filter coefficients DMPF1 to DMPF88, DMPG1 to DMPG88 to the sound sources 8-1 to 8-88 of the sound source unit 8.
And feedback levels FBLVL1 to FBLVL
88 is output.

The sound source 8-n (n = 1 to 88) has a key-on signal KONn, a delay time DLYn, a string characteristic filter coefficient STRGn, and a damper characteristic filter coefficient DMPFn, D
MPGn and feedback level FBLVLn are input, and a tone waveform having a predetermined pitch is output according to these parameters. The output from the sound source 8-n is synthesized by an adder (musical sound waveform synthesizer) 10 and output as a musical sound signal (synthesized output).

On the other hand, the output tone signal (combined output of the adder 10) is input to the multiplier 12 and the soundboard resonance applying section 14. Multiplier 12 multiplies the input signal by overall feedback level FB and outputs the result to adder 16. The soundboard resonance imparting unit 14 is a filter that simulates the characteristics of the soundboard of a piano, which is a natural musical instrument. The soundboard resonance imparting unit 14
The output signal having the resonance effect of the soundboard is added to the adder 16.
Output to. The adder 16 adds these inputs and feeds them back to each of the sound sources 8-1 to 8-88.

Next, referring to FIG. 2, one sound source 8-n
Will be described.

The sound source 8-n (n = 1 to 88) is connected to the adder 2
6, a delay circuit 20, a string characteristic filter 22, and a damper characteristic filter 24 in series with a loop circuit 29. The sound source 8-n outputs an initial drive waveform corresponding to the key-on signal KONn.
Having. The initial drive waveform output from the initial drive waveform generator 28 is supplied to the loop circuit 29 via the adder 26.
Supplied to

The delay circuit 20 of the loop circuit 29 delays the input tone waveform by a delay time DLYn and outputs it. The pitch of the tone signal generated by the tone generator 8-n is determined by the delay time of the entire loop circuit, and is controlled here by the delay time DLYn. When using a high-order FIR or other filter having a delay amount sufficient to obtain a desired pitch, a separate
It is known that it is not necessary to prepare a delay circuit, and that the delay amount can be controlled by appropriately setting the coefficient of the filter.

The string characteristic filter 22 is a filter equivalent to the frequency characteristic of a sounding body to be simulated (in this embodiment, a piano string), and changes and outputs an input musical sound waveform based on the frequency characteristic. The frequency characteristic is set for each sound source according to the string characteristic filter coefficient STRGn.

The damper characteristic filter 24 is a variable filter that approximates the characteristics of the damper, and changes and outputs the input tone waveform based on the damper characteristics. The frequency characteristic of the damper characteristic filter 24 is based on the damper operation amount DMPR
Changes in accordance with the frequency characteristic coefficient DMPFn corresponding to. The gain characteristic of the damper characteristic filter 24 is a gain characteristic coefficient DMPGn corresponding to the damper operation amount DMPR.
It changes according to.

The adder 26 of the loop circuit 29 adds and outputs the initial drive waveform from the initial drive waveform generator 28, the output waveform of the damper characteristic filter 24, and the feedback input from the multiplier 30.

The multiplier 30 receives the feedback of the combined output in FIG. 1 as a feedback input FBIN, and inputs a feedback level FBLV for each sound source.
Ln is multiplied and output.

The output of the adder 26 is the final output as the sound source 8-n.

Next, the damper characteristic filter 24 shown in FIG. 2 will be described with reference to FIG.

The damper characteristic filter 24 used here
Is basically a low-pass filter. The damper characteristic filter 24 includes a subtracter 26, a multiplier 28, an adder 3
0 and a loop circuit in which delay circuits 32 are connected in series. The output of the delay circuit 32 is fed back not only to the subtractor 26 but also to the adder 30. The input waveform INn to the damper characteristic filter 24 is injected into this loop circuit via the subtractor 26 and circulates through this loop circuit.

The multiplier 28 controls the cutoff frequency of the filter 24. The frequency characteristic coefficient DMPFn, which is a multiplier of the multiplier 28, changes according to the operation amount (depressed amount) of the damper pedal, for example, as shown in FIG. In FIG. 7, the horizontal axis indicates the damper position, that is, the state where the damper pedal is not depressed at the damper position “0”, and the state where the damper pedal is fully depressed at the damper position “1”. As the amount of depression of the damper pedal increases, the frequency characteristic coefficient DMPFn, which is a multiplier in the multiplier 28, increases, and the cutoff frequency of the damper characteristic filter 24 increases.

The output of the damper characteristic filter 24 is extracted from the output of the adder 30 via a multiplier 34. The multiplier 34 is for controlling the level of the filter 24. Gain characteristic coefficient DMP which is a multiplier of multiplier 34
Gn changes, for example, as shown in FIG. 8, according to the operation amount (depressed amount) of the damper pedal. As shown in this figure, as the amount of depression of the damper pedal increases,
The gain characteristic coefficient DMPGn, which is a multiplier of the multiplier 34, increases, and the gain of the damper characteristic filter 24 increases.

It should be noted that the damper characteristic filter may be simply composed of only a multiplier. FIG. 4 shows an example in which the damper characteristic filter 24 'is composed of only the multiplier 36 having the gain characteristic coefficient DMPGn as a multiplier. However, in order to obtain the effect of pressing the strings with felt like a damper pedal of an actual piano, the configuration of FIG. 3 is more preferable.

Next, the information conversion unit 18 shown in FIG. 1 will be described with reference to FIG.

The information conversion section 18 is a circuit for calculating a coefficient or the like defining a damper characteristic to be given to the damper characteristic filter 24. The information conversion unit 18 has 88 individual conversion units 40-1 to 40-88 corresponding to each key. Since each conversion unit has the same configuration, the conversion unit 40
-1 will be described.

The conversion section 40-1 includes N or (OR) circuits, a filter coefficient table 44, and a gain coefficient table 4.
6, and a feedback input coefficient table 48.

The OR circuit 42-1 has a key-on signal KON1.
(1 bit) and the first bit of the damper operation amount DMPR (predetermined number of bits) are input, and the result of the OR operation is output. Similarly, the OR circuit 42-n (n = 1 to N) receives the key-on signal KON1 and the n-th bit of the damper operation amount DMPR, and outputs the result of the OR operation. Therefore, the N OR circuits 42-1 to 42-N provide the damper operation amount DMPR when the corresponding key is not turned on.
And outputs all bits “1” when the corresponding key is turned on.

The filter coefficient table 44 is a table storing a list of frequency characteristic coefficients DMPF1 corresponding to the damper operation amount DMPR. The gain coefficient 46
Is a table storing a list of gain characteristic coefficients DMPG1 corresponding to the damper operation amount DMPR.

The filter coefficient table 44 and the gain coefficient table 46 are read in accordance with the outputs of the OR circuits 42-1 to 42-N, and the frequency characteristic coefficient DMPF1 and the gain characteristic coefficient DMPG1 are output. When the corresponding key is turned on, data of all bits “1” is input to the tables 44 and 46. This means that the damper operation amount DMPR is forcibly set to "1" (that is, the maximum depression state).
Is equivalent to

On the other hand, the 1-bit key-on signal KON1 and the damper operation amount DMPR of a predetermined bit are input to the feedback input coefficient table 48. The feedback input coefficient table 48 shows the damper operation amount D for each case when the corresponding key is turned on and when it is not.
It is a table storing a list of feedback levels FBLVL1 corresponding to MPR.

The feedback level FBLV corresponds to the presence / absence of key-on and the damper operation amount DMPR.
L1 is read and output.

It should be noted that the key-on signal KON1 is normally set so that the string that is key-on should not be involved in resonance.
Is "1", the feedback input coefficient table 48 may be set so that the feedback level FBLVL1 is set to "0" regardless of the value of the damper operation amount DMPR.

The feedback input coefficient table 4
8 stores only the information when the key is not turned on,
The table output may be gated by a key-on signal.

FIG. 6 shows a connection example of such a feedback input coefficient table. In this figure, the feedback input coefficient table 50 has a damper operation amount DMP.
R and the corresponding feedback level FB
LVLn is output. This output is input to an AND circuit 54. On the other hand, the key-on signal KONn is inverted by the NOT (NOT) circuit 52, and
To enter. Therefore, when the key-on signal KONn is "0", that is, when the corresponding key is not turned on, the feedback level FBLVLn is output from the AND circuit 54. When the key-on signal KONn is "1", that is, the corresponding key is turned on. When all the bits are “0” as the feedback level FBLVLn
Is output from the AND circuit 54.

The fact that the feedback level FBLVLn is all "0" means that the feedback input FBIN is cut off in the tone generator of FIG. 3, and the tone generator does not participate in resonance.

Next, the operation of the electronic piano will be briefly described. When the keyboard 2 of the electronic piano is pressed, the control unit 6
Detects the pressing, and detects a corresponding key-on signal KONn, a delay time DLYn, and a string characteristic filter coefficient STRGn.
Is output. Further, the control unit 6 detects the operation amount of the damper pedal 4 and outputs the corresponding damper operation amount DMPR and the overall feedback level FB.

Key-on signal KO output from control unit 6
A signal such as Nn is input to the corresponding sound source 8-n. In the sound source 8-n, the key-on signal KONn is input to the initial drive waveform generator 28, whereby the initial drive waveform is supplied to the adder 26 of the loop circuit 29. The initial drive waveform circulates through the loop circuit 29 to generate and output an output waveform. The output of the loop circuit 29, that is, the output of the sound source 8-n is synthesized and output by the adder 10 with the output from another sound source.

The output of the adder 10 is fed through the multiplier 12 and the adder 16, and further through the soundboard resonance imparting section 14 and the adder 16, to the feedback inputs FBIN of all the sound sources 8-1 to 8-88. To enter. Each sound source 8-1 to 8
At −88, the feedback input FBIN is multiplied by the feedback level FBLVLn, and the loop circuit 2
Inject into 9.

As described above, in a musical instrument such as a piano, when a certain key is struck and sounded, the vibration propagates through the piano frame or sound board to excite the other strings as a whole, and a resonance sound is generated. The situation that occurs can be simulated.

On the other hand, the damper operation amount DMPR output from the control unit 6 is input to the information conversion unit 18 together with the key-on signals KON1 to KON88. The information conversion unit 18 reads and outputs a frequency characteristic coefficient DMPFn, a gain characteristic coefficient DMPGn, and a feedback level FBLVLn corresponding to the damper operation amount DMPR. At this time, values corresponding to the maximum value of the damper operation amount DMPR are forcibly output as these parameters to be given to the sound source corresponding to the key that is turned on, and the feedback level FBLVLn for the resonance effect is forcibly output. Outputs "0". This makes it possible to prevent the key-on sound source from generating a resonance sound.

The frequency characteristic coefficient DMPFn, the gain characteristic coefficient DMPGn, and the feedback FBLVLn output from the information conversion unit 18 are input to the corresponding sound source 8-n. The frequency characteristic coefficient DMPFn and the gain characteristic coefficient DMPGn change the characteristic of the damper characteristic filter 29 of the loop circuit 29 of the sound source 8-n. As a result, the effect of the damper is provided, and a musical tone waveform of the resonance component is output. Further, it is possible to output the resonance component by controlling the feedback level of the synthesized output by the feedback level FBLVLn.

As described above, in the above embodiment,
Since the physical model sound source is configured using the loop circuit, it is possible to generate a musical tone that is similar to a musical tone of a natural musical instrument such as a piano. In addition, the combined output waveform obtained by combining the outputs from the respective sound sources is supplied again to the loop circuit of each sound source, which means that the resonance and the reverberation effect are realized by the physical model sound source. Furthermore, a variable filter that approximates the characteristics of the damper is provided in the loop circuit of each sound source, thereby realizing the effect of the damper with a damper pedal or the like.

In the above-described embodiment, a plucking model for inputting an initial waveform is used as an example of a physical model. However, the present invention is not limited to this, and various excitation signals may be used. For example, a model simulating a hammer striking a string may be used instead of the initial waveform.

Although the damping characteristic of the damper is approximated by a variable filter inserted into a string loop, a physical model of the damper may be used.

Although the number of strings is set to 88, it is also possible to use a model in which each has a plurality of strings, or a model in which the number of strings is reduced to less than 88 and the key length is changed.

The delay time DLY and the string characteristic filter coefficient STRG given to the sound source are input from the control unit 6, but may be set to fixed values in advance.

In the conversion section 18 shown in FIG. 5, the feedback input coefficient table 48 is provided with tables corresponding to two cases, ie, a case where the key is turned on and a case where the key is not turned on, as shown in FIG. Key-on signal KON
Gating may be performed by n, and the feedback input coefficient table may have only data when the key is not turned on.

However, it is desirable to variably control the level of the feedback input signal FBIN to the loop circuit of the sound source depending on whether the key is turned on or not, so that the pedal operation amount DMPR is set as in the above embodiment.
It is preferable to refer to the feedback input coefficient table by using a signal obtained by parallelizing the key input signal KONn with the feedback input coefficient table as an address.

Since the treble strings of a real piano do not have a damper mechanism, the sound sources corresponding to the treble strings in that range are fixed in the damper-open state, and the corresponding converters 40-i are omitted. May be.

The output from the loop circuit may be taken from anywhere in the loop circuit. Further, outputs may be taken out from a plurality of different positions of the loop circuit and synthesized.

Without providing a delay circuit as in the above embodiment, for example, only a filter having a delay amount sufficient to obtain a desired pitch may be used in the loop circuit. in this case,
The delay amount may be controlled by appropriately setting the filter coefficient.

[0055]

As described in the foregoing, according to the present invention, the musical tone synthesizing apparatus which outputs a musical tone signal by synthesizing the signal circulating a plurality of closed loop means that a closed loop connecting the delay means, a plurality of closed loop means After the synthesized tone signal is synthesized, the characteristics of the synthesized signal are changed and re-input to each of the closed loop means. The sound board vibrates, and the vibration of the sound board is transmitted to the strings in reverse. Therefore, it is possible to faithfully reproduce the phenomenon of a natural musical instrument, thereby generating a high-quality tone signal.

[Brief description of the drawings]

FIG. 1 is a block diagram of an electronic piano to which a tone synthesizer according to an embodiment of the present invention is applied;

FIG. 2 is a block diagram of a tone generator of the tone synthesizer.

FIG. 3 is a block diagram of a damper characteristic filter of a sound source.

FIG. 4 is a configuration diagram showing another example of a damper characteristic filter.

FIG. 5 is a block diagram of an information conversion unit of the musical sound synthesizer.

FIG. 6 is a block diagram showing a connection example of a feedback input coefficient table.

FIG. 7 is a graph showing a correspondence between a damper position and a frequency characteristic coefficient.

FIG. 8 is a graph showing a correspondence between a damper position and a gain characteristic coefficient.

[Explanation of symbols]

2 ... keyboard, 4 ... damper pedal, 6 ... control section, 8 ... sound source section, 8-1 to 8-88 ... sound source, 10, 16 ... adder,
12 Multiplier, 14 Soundboard resonance imparting section, 18 Information conversion section, 20 Delay circuit, 22 String characteristic filter, 24 Damper characteristic filter, 26 Adder, 28 Initial drive waveform generator, 29 ... Loop circuit, 30 ... Multiplier, 42-1
~ 42-N ... OR circuit, 44 ... Filter coefficient table,
46: gain coefficient table, 48: feedback input coefficient table.

Claims (1)

(57) [Claims] 1. A signal of unrealized closed loop delay means in a closed loop
The closed loop means whose signal delay characteristic corresponds to the pitch
Are provided independently of each other, and each of these is provided by these closed loop means.
In a tone synthesizer for generating a tone waveform signal of a designated pitch, the tone synthesizer is provided in each of the closed loops and responds to a user operation.
Input the operation information, and according to the operation information,
Multiple loops to change the characteristics of the cyclic signal looping means
An excitation signal is input from outside the loop characteristic changing means and the closed loop means, and the excitation
Combining the signal with the cyclic signal that has passed through the loop characteristic changing means
The cyclic signal input to the loop characteristic changing means.
Generating means , based on signals output from the plurality of closed loop means.
Means for obtaining a synthesized output signal; and a pitch indicated to be sounded among the plurality of closed loop means.
Inputting the excitation signal to the closed loop means corresponding to
And outputting the combined output signal to each of the closed loop hands according to operation information.
Means for re-inputting to a step .