JPH04190300A - Musical tone synthesizer - Google Patents

Abstract

PURPOSE:To synthesize a musical tone different in timbre corresponding to the actual performance of a musical instrument by multiplying the loss state of vibration, varying according to a performance method, as the variation of the multiplier of multiplying means. CONSTITUTION:A musical tone synthesize is composed by adding an adding means 201 consisting of an adder 11 to add driving wave form data issued from a memory 21 to wave form data, and multiplying means 202 consisting of a multiplier memory 12 to send out multiplier data P previously stored corresponding to pitch assisting data Ns and a multiplying device 13 to multiply the data P sent out from the memory 12 by using the wave data sent out from an adder 28 as a multiplicand. Since musical tone data obtained by adding transformed wave form data and the driving wave form data vary, musical tone different in timbre can be obtained corresponding to a musical performance.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a musical tone synthesis device using a digital electronic circuit such as an electronic piano or a synthesizer.

2. Description of the Related Art In recent years, with the advancement of digital technology, many tone synthesis devices have been developed that utilize digital electronic circuits, such as electronic pianos and non-synthesizers.

Among these, a musical tone synthesis device has been proposed in which musical tones are synthesized by an arithmetic circuit using a delay means (for example, Japanese Patent Application Laid-open No. 187398/1988 entitled "Wavetable changing device and method for generating musical tones").

Hereinafter, a conventional tone synthesis device will be described without reference to the drawings.

FIG. 2 is a block diagram of a conventional musical tone synthesis device. The musical tone synthesis apparatus shown in FIG. 2 is capable of faithfully synthesizing the sound of a stringed instrument such as the guitar shown in FIG. 3.

The conventional musical tone synthesis device has a memory 21 that stores drive waveform data in advance and a memory 21 that stores drive waveform data in advance.
The drive waveform data stored in 1 is shifted to the shift clock SCK.
a drive waveform generating means 101 consisting of a drive waveform generating means 101 which performs no reading at the generation timing of , and a selector 23 which selects between the drive waveform data read from the memory 21 and the waveform data and outputs the selection result as musical tone data. 8 selection means +02, a delay means 103 consisting of a shift register 24 which softens the data input from the selector 23 at each generation timing of the shift clock SCK and delays it for a certain period of time, and data and values output from the shift register 24. A multiplier 25 that multiplies the data by 0.5, a delay device 26 that delays the data output from the shift register 24 by one sampling clock (corresponding to a shift clock), and the data output from the delay device 26 and the value 05. a multiplier 27 that performs multiplication of
It is composed of a calculation means 104 consisting of. The calculation means 104 constitutes a first-order low-pass filter, and the selection means 102. Delay means 103. The data circulating through the loop-shaped circuit made up of the calculation means 104 is collectively referred to as waveform data.

FIG. 3 is a schematic diagram showing the sound generation mechanism of a stringed musical instrument (for example, a guitar).

In FIG. 3, the string 3 stretched by the strings 1 and 2 on the body 5 is supported by the flare shaft 4 and the strings 2 on the body 5 by pressing with fingers. When string 3 is played by finger B, it vibrates and generates a musical tone. To explain in detail, finger B
By plucking the string, an acceleration force is input to the point where the string is plucked. After that, the acceleration propagates through the string 3
The call Brinnon 2 is reflected and returns to the place where the finger B was stretched. By repeating this action, the three strings vibrate and produce musical sounds. The musical tone synthesizer shown in FIG. 2 approximately realizes this operation using a digital electronic circuit.

FIG. 4 is an example of the envelope of musical tones of the type of instrument (for example, a guitar) that plays strings shown in FIG. 3, and (A) is at time t.

By playing the string, the vibration of string 3 is generated, and the time 1. Envelope of musical sound when the vibration of string 3 is suddenly attenuated, (B)
is the envelope of musical sound data when string 3 is vibrated by playing string 3 at time tI, (C) is the envelope of the composite value of musical sounds of (A) and (B), and (D) is time t. When the string is played, the vibration of string 3 is generated.F) is the envelope of the composite value of musical tones (D) and (E).

(C) is a type of musical instrument (for example, a guitar) in which string 3 shown in Fig. It represents the envelope.The plucking operation by finger B corresponds to time t., tl, and just before the plucking operation at tl, finger A is pressed again, and string 3
This is the case when the vibration of is stopped.

On the other hand, (F) represents the envelope of musical tones in a type of instrument (such as a guitar) in which string 3 shown in Fig. 3 is plucked, for example, when finger B is not pressed again after each plucking operation. There is. Is it a fingertip operation or time t0.
Corresponds to tl. Since the finger is not pressed again at time t1, the time is t. Since the vibration of string 3 due to the plucking operation remains (time t of (D) (envelope thereafter), time 1.
When you play (envelope of (E)), it becomes the musical tone that is a combination of (D) and (E), that is, the envelope of (F).

Regarding the conventional musical tone synthesizer configured as described above,
The operation will be explained below.

First, when the sound generation start information kon reaches the value l, the counter 2
The counter 22 sets the count value (address of the memory 21) to the value O, and then increments the count value at the timing of the shift clock SCK, and reads the memory 21 each time. After re-addressing the entire area of memory 21 from
It is assumed that the count value is halted when the final address of 1 is reached. The memory 21 stores data corresponding to the acceleration given to the string 3 by plucking the string 3 as described above, and is inputted to the selector 23 by the reading operation of the counter 22. Note that the value 0 is stored at the final address of the memory 21, and therefore, after the count value of the counter 22 reaches the final address of the memory 2I, the drive waveform data continues to output the value 0. It is assumed that the selector 23 selects the 8 inputs until all of the drive waveform data stored in the memory 21 is read, and thereafter selects the B input. The drive waveform data is input to the shift register 24 via the selector 23, and is output as desired musical tone data.

In the shift register 241, the noft clock SCK
(corresponding to a sampling clock), the input data is shifted and transferred to the shift register 24.
output after being shifted by the word length. Therefore, if the sampling period is Ts and the word length of the shift register 24 is M, the delay time Td of musical tone data is Td=TsxM[
SEC] -1---(1). In the schematic diagram of FIG. 4, the operation of this shift register 24 corresponds to a propagation operation in which the acceleration input from finger B to the string 3 propagates through the string 3. This corresponds to the delay time Td of equation 1. The waveform data output from the shift register 24 is subjected to high-frequency cutoff in the calculation means 104. The calculation means 104 is a generally known primary FI.
R filter operation, and the filter transfer function H
(z) is given by the following two equations.

H(z)=(1-z-'),'2-(2!
The waveform data output from the calculation means is sent to B of the selector 23.
entered into the input. When the word length of the two-piece memory 21 is N, there is a relationship between the word length M of the shift register 24 and the following equation 3.

N5M ------ (3) Therefore, the selection means 23. Delay means 103.
After being processed by the circuit constituted by the arithmetic means 104, when the waveform data is again input to the B input of the selector 23 as waveform data, the selector 23 has selected one of the eight inputs.

The delay time Tf due to the transfer function of the second equation is 0.5Ts [
SEC], so the period T of the musical tone actually output is given by the following equation 4.

T = Td + Tf [SEC] --- (4) Pitch f [Hz] of musical tone to be output Period T [5EC
If M obtained from J, the first equation, and the fourth equation is not an integer, that is, if M is a decimal, an all-pass filter is added to the delay means 103 and a delay equivalent to a decimal is realized by phase shifting operation. This is a known technique.

Further, the selection means 102 is constituted by an adder that adds the waveform data and the drive waveform data, and the drive waveform generation means 10
The drive waveform data output from 1 is not output once, but is output repeatedly, and the delay means +03 is reset in response to the sound generation start information Kon before synthesis is started. It is a well-known technology to synthesize and output not only sounds but also stationary sounds such as those of a violin or clarinet. In such cases, Section (3)
If the word length of the memory 21 is N, then there is no need to establish a magnitude relationship with the word length M of the shift register 24.

The drive waveform data output from the drive waveform generation means 101 as described above is sent to the selection means 102. delay means 103;
The loop-shaped circuit constituted by the arithmetic means 104 is circulated. Through this cyclic operation, the high frequency components of the waveform data are gradually cut off by the calculation means 104. This corresponds to the operation in which high-frequency components of acceleration propagating through the string are absorbed into the body 5 at the nut 1 and the bridge 2 in the schematic diagram of FIG.

However, in the conventional configuration as described above, the musical tones that are sounded always have the same tone, and there may be differences in tone due to interference between the previously output musical tone and the newly generated musical tone, or The problem is that it is only possible to subtly change the timbre of the musical tone being produced in accordance with the performance.

SUMMARY OF THE INVENTION The present invention is intended to solve the above-mentioned problems, and it is an object of the present invention to provide a musical tone synthesis device that can use different timbres of musical tones depending on how the musical instrument is played, like an actual musical instrument.

Means for Solving the Problems In order to solve the above problems, the musical tone synthesis device of the present invention has the following features:
a drive waveform generation means for generating and outputting drive waveform data in response to a sound generation start instruction; an addition means for receiving the drive waveform data as one input and the waveform data as the other input; and a delay means for time-delaying the waveform data. , comprising an arithmetic means for processing waveform data, and a multiplication means for taking the waveform data as a multiplicand and multiplying it by a multiplier that changes in response to performance input, the addition means, the delay means, the arithmetic means, and the multiplication means. The waveform data input and waveform data output corresponding to each of the above are connected to form a loop, and the waveform data is extracted as musical tone data.

Operation With the above configuration, a musical tone is generated by adding the waveform data circulating in the loop-shaped circuit composed of the adding means, the delay means, the calculation means and the multiplication means, and the drive waveform data sent from the drive waveform generation means. Data is synthesized. In addition, by changing the multiplier of the multiplier in accordance with the performance, the loop gain of the loop-shaped circuit changes, and the waveform data in the transformed loop and the drive waveform data are combined with an adder to obtain musical sound data. changes, so musical tones with different tones can be obtained depending on the performance.

EXAMPLE Hereinafter, an example of the present invention will be described with reference to the drawings. Components that are the same as those shown in FIG.

FIG. 1 is a diagram of the musical tone synthesis device of the present invention.

The musical tone synthesis device of the present invention has a selector 2 in the conventional example shown in FIG.
3, an adding means 201 consisting of an adder 11 that adds the drive waveform data and waveform data output from the memory 21, and outputs multiplier data P stored in advance corresponding to the pitch auxiliary data Ns. A multiplier 1 that multiplies multiplier data P output from the multiplier memory 12 using the waveform data output from the multiplier memory 12 and the adder 28 as a multiplicand.
3, a multiplication means 202 consisting of 3 is added.

The operation of the musical tone synthesis apparatus constructed as described above will be described below with reference to FIGS. 1 and 3 to 6. The basic operation of the musical tone synthesizer shown in Fig. 1 is shown in Fig. 2.
This is similar to the conventional musical tone synthesis device shown in the figure.

First, when the pitch auxiliary data Ns is input, the multiplier memory 1
2 outputs the multiplier data P stored at the corresponding address using the pitch auxiliary data Ns as an input address. Here, the pitch auxiliary data Ns refers to data that detects the performance operation required to determine the pitch. In electronic musical instruments such as guitars or violins, the left hand is usually used to determine the pitch. This data corresponds to the pressing strength of the fingerboard input used by pressing, and is different from so-called touch data that indicates the strength of pronunciation.

For example, in the case of a guitar type instrument, such data can be easily obtained by disposing a known touch sensor on the fingerboard 6 shown in FIG.

When it is assumed that the detected value of pitch auxiliary data Ns takes a value from 0 to 127, the multiplier memory 12 stores data from address OO[HEX] to address 7F[HE
X) (JJ bottom, 00) (written as ,071 (t17)) I: Stores multiplier data P from 0.0 to 1.0. When the value of the multiplier data P is large, it indicates that the fingerboard 6 is pressed strongly, and when the value of P is small, it indicates that the pressing method is weak and the vibration damping is large.

First, the musical tone shown in Figure 4 (C) can be produced by playing the string-struck type instrument shown in Figure 3, for example, by pressing finger A again every time finger B plucks string 3. The operation when synthesizing musical tones will be explained. In this case, the musical tone data that is being generated is first muted before it is generated. Repressing finger A means
This means that the pitch auxiliary data Ns once decreases to 0 and then increases again, after which string 3 is played by the fingertips and new sound generation start information kon is input. . In other words, the value of pitch auxiliary data Ns is, for example, 127 (7FH) = O(OOH) → 127
(7FH), the multiplier data P output from the multiplier memory 12 becomes 1.0-0.0-1.0, and the multiplier]3 is the multiplier data P output from the multiplier memory 12.
and the waveform data output from the calculation means 104 and output the product. Such a human operation takes about 100" and 11 seconds, and within this time, the addition means 201 and the delay means 103
The waveform data circulating in the loop composed of the calculation means 104 and the multiplication means 202 will rapidly attenuate. Once, the musical tone data output from the time when the new sound generation start information kon reached the value 1 (time 1 in FIG. Since it is obtained by addition with the waveform data that has been reset to the value of
It will look like (C) in the figure.

Next, in order to produce a musical tone as shown in (F) in FIG.
An explanation will be given of the operation when synthesizing musical tones produced by a performance method in which the finger B is continuously played while the finger B is held down at all times. In this case, there is no need to mute the musical tone data before making it sound. Therefore, keeping the finger pressed down means that string 3 will be played by finger B and new sound generation start information kon will be input without changing the pitch auxiliary data Ns. do. That is,
Pitch auxiliary data NsO value strength example, 127 (7F
In the case of H), the multiplier data P remains 1, 0, and the waveform data circulating in the loop composed of the addition means 201, the delay means 103, the calculation means 104, and the multiplication means 202 starts a new sound as it is. The driving waveform data output from the driving waveform generating means 101 corresponding to the information kon is added by the adding means 201, and the point in time when the sound generation start information kon newly reaches the value l (time t1 in FIG. 4) The envelope of musical tone data synthesized from is as shown in FIG. 4 (F).

Next, an explanation will be given of the operation when synthesizing musical tones produced by a playing method in which the way the fingers are pressed while the musical tones are being produced is changed. Assume that you press string 3 with finger B strongly and press it weakly after (time t), press it strongly again after time t2, and then press it weakly after time t. Adjust the pitch auxiliary data Ns to time jo + j2 + t3, 127 (7FH) → 64 (40H) -127 (
7FH), the multiplier data P output from the multiplier memory 12 becomes 1.0 → 0.7 → 1.0. Multiplier 13
multiplies the multiplier data P and the waveform data output from the calculation means +04 and outputs the product. FIG. 6 shows the envelope of musical tone data output by such a synthesis operation. In FIG. 6, the envelope of musical tone data is at time t. After the sound is started at , the value of the multiplier data P decreases at time t2, resulting in a rather rapid attenuation, and at time t, the value of the multiplier data P increases again, so the attenuation changes to normal attenuation again. I will do it.

In this way, according to the present embodiment, when producing a musical tone as shown in FIG. 4C, that is, in an instrument of the type in which string 3 in FIG. The musical sounds produced by playing methods such as pressing the keys again are also
The musical sound shown in (F) in Figure 4 can be produced by playing a musical instrument of the type shown in Figure 3, in which string 3 is plucked, for example, by continuously plucking with finger B while keeping the finger pressed down. It is also possible to synthesize musical tones that are being produced, and it is also possible to change the musical tone data being produced in response to playing methods such as changing the force with which the fingers are pressed.

Also, since the musical tone data is synthesized by adding the waveform data and the drive waveform data, the drive waveform data continues to be generated in response to the sound generation start information kon. It is also possible to synthesize stationary tones.

In this embodiment, the delay means 103 may be configured by the soft recorder 24, or may be configured by using a memory.
It is generally known that 3 can be constructed.

Further, the addition means 201. Delay means 103. Arithmetic means 10
4. The loop-shaped circuit constituted by the multiplication means 202 may change the connection order thereof. Furthermore, the musical tone data may be taken out after the addition means 201, for example after the delay means 103, after the calculation means 104, or after the multiplication means 202.

Effects of the Invention As described above, according to the present invention, the vibration loss state that changes depending on the playing method is applied to the waveform data circulating in the loop circuit composed of the addition means, the delay means, the calculation means, and the multiplication means. By multiplying by changing the multiplier of the multiplication means, it is possible to synthesize musical tones with different tones corresponding to actual musical instrument performances.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a musical tone synthesizer according to an embodiment of the present invention, FIG. 2 is a block diagram of a conventional musical tone synthesizer, and FIG. 3 is a sound generation mechanism of a stringed instrument (for example, a guitar). Schematic diagrams, Figures 4 (A) to (F) and Figure 6 are respectively state diagrams showing an example of the envelope of musical tones of the type of instrument (for example, a guitar) that plays the strings shown in Figure 3, and Figure 5 is a state diagram showing an example of the envelope of musical tones of the type of instrument (for example, a guitar) that plays the strings shown in Figure 3. 1
It is a graph which shows the contents of the multiplier memory of a figure. 101... Drive waveform generation means, 103... Delay means. 104...Arithmetic means, 201...Addition means. 202...Multiplication means.

Claims (1)

[Scope of Claims] 1. Drive waveform generation means for generating and outputting drive waveform data in response to a sound generation start instruction; addition means for receiving the drive waveform data as one input and the waveform data as the other input; and waveform data. a calculation means for processing the waveform data; and a multiplication means for multiplying the waveform data by a multiplier that changes in response to the performance input, the addition means and the delay means A musical tone synthesis device characterized in that the corresponding waveform data input and waveform data output of the calculating means and the multiplication means are connected to form a loop, and the waveform data is extracted as musical tone data.