JP3298152B2 - Music signal generator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tone signal generator for generating a tone signal in an electronic musical instrument or the like, and more particularly to a tone signal generator capable of giving a special effect to a tone.

[0002]

2. Description of the Related Art Conventionally, simulating a musical tone of a natural musical instrument has been one important technique in musical tone control of an electronic musical instrument. It is also required to generate a variety of musical tones by generating such special musical tones.

[0003]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a musical tone signal generator capable of generating a special musical tone with a simple configuration in an electronic musical instrument or the like.

[0004]

Tone signal generation device of the present invention when taken in order to solve the above problems, there is provided a means for solving] is a predetermined sampling frequency, the original waveform for outputting original waveform data of the designated tone pitch Data output means, and sampling means for sampling the original waveform data at a sampling frequency lower than the sampling frequency of the original waveform data output means, wherein the sampling means can generate an alias component for the original waveform data. And outputting a tone signal including the alias component.

[0005]

In the tone signal generating apparatus according to the present invention, the original waveform data output means has a predetermined sampling frequency .
And outputs the original waveform data of the designated tone pitch, the original waveform data is output is sampled by the sampling means as a musical tone signal.

Since the sampling frequency of the sampling means is set lower than the sampling frequency of the original waveform data output means, the sampling means is converted into a musical sound waveform composed of the original waveform data of the original waveform data output means. If a frequency component equal to or more than 1/2 of the sampling frequency is included, an alias is generated as a result of the sampling by the sampling means, and a special effect is added to the tone signal by the alias component.

The sampling frequency of the sampling means may be selected from a plurality of frequencies, such as 1/2 and 1/4 of the sampling frequency of the original waveform data output means.

[0008]

FIG. 1 is a block diagram of a tone signal generating apparatus according to an embodiment of the present invention. This apparatus constitutes, for example, a sound source incorporated in an electronic musical instrument or a sound source connected to another device. To generate a tone signal.

An FM output waveform is generated at a high-speed sampling frequency of 200 kHz and resampled at a medium-speed sampling frequency of 100 kHz and a low-speed sampling frequency of 50 kHz, thereby adding an alias component to a musical sound signal and thereby generating a special signal. A musical tone can be generated. Note that the apparatus of this embodiment is
kHz low-speed sampling clock φ _S , 100 kHz
Operates based on a medium-speed sampling clock φ _M , a high-speed sampling clock φ _{H of} 200 kHz and an internal basic clock φ _HH of 800 kHz.

In FIG. 1, reference numeral 1 denotes a control unit. Performance information such as pitch, timbre, key press information, etc. is obtained from a performance operation unit such as a keyboard or a circuit (not shown) such as an automatic performance device or a performance computer. And outputs various signals for controlling each circuit according to the performance information.

Specifically, an instruction signal KON instructing the modulated wave generator 2 and the modulated wave generator 3 to generate a waveform.
M, KONC, pitch signal PITCH specifying pitch
M, PITCHC, tone color signals TCM, T for specifying tone colors
CC is output. These tone signals TCM, TCC
Is used, for example, to determine the waveform shape of a modulated wave or a modulated wave. Other circuits include a sampling clock selection signal SCKSEL for selecting a sampling clock for resampling, a tone amplitude control parameter EGPAR for designating an envelope of a tone signal, a modulation index parameter IGPAR for designating an envelope of a modulation signal, and frequency modulation. The offset data OFSET giving the reference point to be multiplied is output.

The modulation wave generator 2 converts a modulation signal for frequency-modulating a carrier signal into a low-speed sampling clock φ.
It is generated in synchronization with _S (50 kHz) and is output to the interpolation filter 4 as an input modulated wave MW. Although the modulation signal may be generated by the modulation wave generator 2 itself,
For example, a modulated signal may be generated and supplied by feedback processing such as feedback FM.

On the other hand, the modulated wave generator 3 generates a carrier signal for FM modulation in synchronization with the low-speed sampling clock φ _S, and uses this as an input carrier CW for the interpolation filter 5.
Output to

The interpolation filters 4 and 5 receive the input 50k
Hz modulated wave MW, carrier C at low sampling frequency
W is the high-speed sampling clock φ _H (200 kHz)
In synchronism with the above, oversampling is performed four times at a high-speed sampling frequency of 200 kHz, and converted into a high-speed sampling modulation wave MWH and a high-speed sampling carrier wave CWH. Since the oversampling technique itself is a well-known technique, a detailed description thereof will be omitted, but a filter having a well-known configuration such as FIR can be applied as a filter in this technique.

The high-speed sampling modulation wave MWH output from the interpolation filter 4 is supplied to a multiplier 6. Further, the modulation index generator 7 in accordance with the modulation index parameter IGPAR and instruction signal KONM from the control unit 1 in synchronism with the fast sampling clock phi _H, and outputs the envelope value ENV for controlling the amplitude of the modulated wave to the multiplier 6. The numbers on the signal lines in the figure represent the number of bits for digital processing.

The multiplier 6 adds the envelope value EN from the modulation index generator 7 to the high-speed sampling modulated wave MWH.
V, and outputs the result INT of the integer part of the high-speed sampling modulated wave MEGWH to the FM processing unit 8. Note that the multiplication of the envelope value ENV gives
The amplitude of the high-speed sampling modulation wave MWH changes to change the depth of frequency modulation with respect to the carrier, and a complex spectrum change, that is, a timbre change is given to an FM output waveform described later.

Meanwhile, FM processor 8 is configured by the read control unit 11 that operates at a high speed sampling clock phi operating at _H (n + 1) stages of the shift register 9 ₀ to 9 _n, the selector 10 and reference clock phi _HH cage, successively takes in the high-speed sampling carrier CWH is the output of the interpolation filter 5 to the shift register 9 ₀ to 9 _n.

The read controller 11 controls the value INT (MEG) of the integer part of the high-speed sampling modulated wave MEGWH after the amplitude control.
Upper 11 bits of WH) and offset data OFS
Depending on the T, of the shift register 9 ₀ to 9 _n [INT
-2] -OFSET, [INT-1] -OFSET
Level, [INT] -OFSET Level, [INT + 1]-
In order to select the output of each of the four successive shift registers of the OFSET stage, the selector 10 is instructed to indicate the respective shift registers by a basic clock φ _HH (800 kHz).
Output in synchronization with z).

Thus, the FM processing unit 8 outputs, for each piece of waveform data of the high-speed sampling carrier CWH input at 200 kHz from the interpolation filter 5, the output of the corresponding shift register selected by the selector 10. The two modulated data SW (waveform data read from the shift register) are sequentially output to the output interpolation calculation unit 12 at a timing of 800 kHz.

On the other hand, the value FRAC (MEG) of the fractional part of the high-speed sampling modulated wave MEGWH output from the multiplier 6
The lower 12 bits of WH are input to the coefficient generator 13. The coefficient generating unit 13 is configured by a memory such as a ROM, for example, and outputs the interpolation coefficient COEF to the output interpolation calculating unit 12 according to the value FRAC of the decimal part. Specifically, the high-speed sampling modulation wave MEGWH
Four interpolation coefficients COEF are sequentially supplied to the output interpolation calculation unit 12 at the timing of 800 kHz for the decimal part value FRAC given every (200 kHz).

The output interpolation calculator 12 includes four modulated data SW output from the FM processor 8 and a coefficient generator 13
Are interpolated by convolution of the corresponding coefficients output from the
H (interpolated FM output) is obtained and output to the sample circuit 15. It should be noted that a specific technique and configuration for performing the same type of musical tone waveform convolution operation is described in, for example, Japanese Patent Application Laid-Open No. 61-905
As disclosed in Japanese Patent Application Publication No. 14 (1999), four waveform data are time-division-multiplied for each coefficient, and the accumulated result is
What is necessary is just to output.

[0022] That is, the output interpolation operation unit 12, for example is constituted by a multiplier 21 and an accumulator 22 as shown in FIG. 6, in synchronism with the basic clock phi _HH, 4
One of the multiplication results by accumulating in the accumulator 22 with the waveform data and coefficients sequentially multiplied by the multiplier 21, in synchronism with the fast sampling clock phi _H outputs the accumulated value from the accumulator 22.

On the other hand, in FIG. 1, a sampling clock selector 14 selects a sampling clock designated by a sampling clock selection signal SCKSEL among three types of a high-speed sampling clock φ _H , a medium-speed sampling clock φ _M and a low-speed sampling clock φ _S. Is output to the sample circuit 15.

The sampling circuit 15 is constituted by a latch and resamples the FM output waveform WH output at a high sampling frequency (200 kHz) in synchronization with the sampling clock input from the sampling clock selector 14 and sequentially samples the waveform. Decimation filter 16 using data as FM output waveform WS
Output to

The decimation filter 16 uses the frequency set by the cutoff coefficient DCMCUTOFF as a cutoff frequency in the FM output waveform WS, for example, to cut frequency components above the audible range, and converts the frequency-cut waveform data to low-speed data. Sampling clock φ _S (50
kHz), and is output to the multiplier 17 as an FM output waveform WOUT. This decimation filter 1
The cutoff frequency in 6 may be set according to the tone color or the like, or may be arbitrarily set.

On the other hand, the envelope generator 18
The envelope values ENG in accordance with the musical tone amplitude control parameter EGPAR at low speed sampling clock phi _S multiplier 1
7 is output. Then, the multiplier 17 outputs the FM output waveform WOUT from the decimation filter 16 and the envelope value ENG from the envelope generator (EG) 18.
And outputs the result of the multiplication as a tone signal W to a sound system (not shown). As a result, the tone signal with the envelope is output to the sound system. It should be noted that such an envelope may be provided before the sample circuit 15.

Next, the operation will be described with reference to the timing chart of FIG. First, when performance information is input to the control unit 1, the control unit 1 transmits the instruction signals KONM, KONC,
Pitch signal PITCHM, PITCC, tone signal TC
M, TCC, sampling clock selection signal SCKSE
L, a tone amplitude control parameter EGPAR, a modulation index parameter IGPAR, and offset data OFSET.

As a result, the input modulated wave MW and the input carrier CW corresponding to the pitch and timbre of the performance information are generated by the modulated wave generator 2.
And the modulated wave generator 3 output a low-speed sampling clock φ _S of 50 kHz. The input modulated wave MW and the input carrier CW of the low sampling frequency are
The signals are converted into a 200 kHz high-speed sampling modulated wave MWH and a high-speed sampling carrier wave CWH by quadruple oversampling of the interpolation filters 4 and 5, respectively. The high-speed sampling modulated wave MWH is multiplied by an envelope value ENV from the modulation index generator 7 to give an envelope.

Next, the integer part of the envelope-attached high-speed sampling modulated wave MEGWH and the high-speed sampling carrier CWH are input to the FM processing unit 8 and subjected to FM modulation processing, and the high-speed sampling carrier CW of 200 kHz is performed.
Four modulated data SW are 8 for each waveform data of H
It is sequentially output at the timing of 00 kHz. Further, a high-speed sampling modulated wave MEGW with an envelope added thereto.
The interpolation coefficient COEF is 800 kHz based on the decimal part of H
Are sequentially output from the coefficient generation unit 13 at the timing of. Interpolation is performed by the output interpolation calculation unit 12 based on the four modulated data SW and the interpolation coefficient COEF, and a 200 kHz FM output waveform WH is output.

As described above, the high-speed sampling carrier wave CWH is FM-modulated by the high-speed sampling modulation wave MEGWH by the FM processing unit 8 and the output interpolation calculation unit 12, and
The FM output waveform WH of kHz is supplied to the sample circuit 15. Then, the sampling circuit 1 responds to the sampling clock selected by the sampling clock selector 14.
5 samples the FM output waveform WH, and outputs the sampling clock φ _H ,
phi _M, FM output waveform corresponding to _{φ S WS (φ H),} WS
(Φ _M ) and WS (φ _S ) are obtained. The FM output waveform WS is output from the decimation filter 16 in synchronization with the low-speed sampling clock φ _S of 50 kHz.
OUT is output.

FIGS. 2 to 4 are diagrams showing an example of the frequency characteristic of each waveform signal in the embodiment. Now, the output interpolation calculation unit 12 obtains the FM output waveform WH as shown in FIG.
The cutoff frequency of the decimation filter 16 is 25
It is assumed that the frequency is set to kHz.

Here, the sampling clock of the sampling circuit 15 is a high-speed sampling clock φ of 200 kHz.
_If it is set to _H , the FM output waveform WS of FIG. 4A is output as it is from the sample circuit 15, but the decimation filter 16 cuts off 25 kHz or more, and the low-speed sampling clock in the decimation filter 16 With the sampling output of φ _S (50 kHz), the FM output waveform W as shown in FIG.
OUT is obtained. Then, what is actually reproduced in analog is a component in the audible range a of 25 kHz or less, and a normal tone is generated. When the cutoff frequency of the decimation filter 16 is set to 12.5 kHz, an FM output waveform WOUT as shown in FIG. 4C is obtained.

Next, assuming that the sampling clock of the sampling circuit 15 is set to the low-speed sampling clock φ _S of 50 kHz, the sampling circuit 15
Since the FM output waveform WH of 50 Hz is resampled at 50 kHz, the frequency component of 25 kHz or more in FIG. 4A becomes aliasing noise, and as shown in FIG.
FM output waveform W including alias in frequency domain below z
S is output from the sample circuit 15.

The decimation filter 16 calculates 2
The FM output waveform WOUT cut off at 5 kHz or more is as shown in FIG. 2 (B), and a special tone including an alias component is generated in the audible range a. When the cutoff frequency of the decimation filter 16 is set to 12.5 kHz, an FM output waveform WOUT as shown in FIG. 2C is obtained, but also in this case, an alias component occurs in the audible range a.

Further, the sampling clock of the sample circuit 15 is a medium-speed sampling clock φ of 100 kHz.
_When set to _M , the low-speed sampling clock φ _S
, The FM output from the sample circuit 15
The output waveform WS is as shown in FIG. 3A, and the FM output waveform WOUT output from the decimation filter 16 is as shown in FIG. When the cutoff frequency of the decimation filter 16 is set to 12.5 kHz, an FM output waveform WOUT as shown in FIG. 3C is obtained, and in any case, a special tone including an alias component is included in the audible range a. Generated.

In the above embodiment, the sampling clock of the sampling circuit 15 is the sampling clock selection signal S
Although the selection is made according to the tone by CKSEL, the selection may be made arbitrarily as desired regardless of the tone.

In the above embodiment, the original waveform data is obtained by the FM modulation. However, the present invention is not limited to this, and the original waveform data of the high-speed sampling clock is resampled by the low-speed sampling clock. I just need.

[0038]

As described above, according to the tone signal generating apparatus of the present invention, the preset sampling frequency
Output the original waveform data of the designated pitch , and sample the original waveform data at a sampling frequency lower than the sampling frequency of the waveform data so that an alias component can be generated. Since a signal can be output, a special musical tone can be generated with a simple configuration in an electronic musical instrument or the like, thereby enriching a variety of musical tones.

[Brief description of the drawings]

FIG. 1 is a block diagram of a musical sound signal generator according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating an example of a waveform signal according to the embodiment.

FIG. 3 is a diagram illustrating another example of the waveform signal in the embodiment.

FIG. 4 is a diagram showing still another example of the waveform signal in the embodiment.

FIG. 5 is a timing chart in the embodiment.

FIG. 6 is a block diagram illustrating an example of an output interpolation calculation unit according to the embodiment.

[Explanation of symbols]

8 FM processing unit, 12 output interpolation calculation unit, 14 sampling clock selector, 15 sample circuit, 16
Decimation filter.

Claims (1)

(57) [Claims] At a sampling frequency set in advance ,
Provided the original waveform data output means for outputting the original waveform data of the designated tone pitch, and a sampling means for sampling said original waveform data at a sampling frequency lower frequency than the sampling frequency of the original waveform data output means, the original A tone signal generating apparatus, wherein an alias component is generated for the waveform data by the sampling means, and a tone signal containing the alias component is output.