JP2730420B2 - 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 synthesizer capable of easily producing sounds.

[0002]

2. Description of the Related Art Conventionally, a sound model of a natural musical instrument has been represented by DS.
2. Description of the Related Art There is known a technique of synthesizing a musical tone by performing simulation using a P (digital signal processor) or the like. As an example, FIG. 7 shows an outline of an algorithm of a physical model sound source used for wind instrument simulation.
This will be described with reference to FIG.

In FIG. 1, reference numeral 101 denotes a non-linear operation unit.
This simulates a non-linear part in a wind instrument, that is, a lead part. A linear operation unit 102 simulates a linear part of a wind instrument, that is, a resonance tube. Reference numeral 107 denotes a resonance system that extracts a pressure wave signal or the like propagating through the non-linear operation unit 101 or the linear operation unit 102, adds a resonance effect to the extracted signal, and outputs the signal. Reference numeral 103 denotes a DAC (digital-to-analog converter) which converts an output signal of the resonance system 107 into an analog signal and outputs the analog signal as a tone signal.

A non-linear operation unit 101 is externally supplied with a blowing pressure signal pre indicating the player's blowing pressure and an embouchure signal emb indicating the player's lip tension. From 102, the pressure reflected wave signal q
_i is supplied. The non-linear operation unit 101 outputs a pressure traveling wave signal q ₀ based on these data and signals.
The output pressure traveling wave signal q ₀ is supplied to the linear operation unit 102, propagates while being reflected and attenuated therein, and is returned to the nonlinear operation unit 101 as a new pressure reflected wave signal q _i .

Next, details of the linear operation unit 102 will be described with reference to FIG. In the figure, reference numerals 113, 114, and 119 denote delay circuits provided to simulate a propagation delay of a pressure wave in a resonance tube of a wind instrument. The delay times of these delay circuits are constants DLEB, DLEL and DL depending on the instrument to be simulated and the pitch.
Given by EA. Reference numerals 115 and 120 denote LPFs (low-pass filters), which simulate the propagation loss of pressure waves in the resonance tube. Also, 110,1
Reference numeral 16 denotes a multiplier, which simulates losses at the front end and the rear end of the resonance tube by multiplying the passing signals by predetermined reflection coefficients REFS and REFL.

A junction 200 simulates a tone hole provided in a resonance tube of a wind instrument by appropriately reflecting and filtering the pressure reflected wave signal q _i and the pressure traveling wave signal q ₀ . The components described above are connected in a loop, and a signal propagating through each component simulates a pressure traveling wave and a pressure reflected wave inside the resonance tube.

Next, the pressure traveling wave signal q ₀ supplied from the non-linear operation section 101 is supplied to multipliers 121 and 118. In the multiplier 118, the supplied pressure traveling wave signal q ₀ is multiplied by the input gain constant NLSO, and the multiplication result is added to the pressure reflected wave signal via the adder 117. Thus, the effect of the blowing pressure on the pressure reflected wave is simulated. Similarly, in the multiplier 122, the pressure traveling wave signal q ₀ is multiplied by the input gain constant NLLO, and the result of the multiplication is added to the pressure traveling wave signal via the adder 112 and given to the pressure traveling wave by the blowing pressure. The effect is simulated.

Next, the pressure traveling wave signal output through the LPF 120 and the multiplier 110 is multiplied by the linear portion output gain NLSI through the multiplier 111 and supplied to the adder 123. Similarly, the pressure reflected wave signal at the previous stage of the adder 117 is output to the linear unit output gain via the multiplier 121.
NLLI is multiplied and supplied to the adder 123. These signals are added in the adder 123, and the addition result is supplied to the nonlinear operation unit 101 as a pressure reflected wave signal q _i .

Thus, the linear operation unit 1 shown in FIG.
02, the delay circuits 113, 114, 119, L
The pressure wave signal is propagated through the loop including the PFs 115 and 120, and the pressure reflected wave signal q _i is formed under the influence of the pressure traveling wave signal q _{0 and} is fed back to the non-linear operation unit 101.

Next, details of the non-linear operation section 101 will be described with reference to FIG. In FIG. 9, reference numeral 140 denotes a subtractor, which subtracts the blowing pressure signal pre from the pressure reflected wave signal q _i and outputs the subtraction result as a pressure difference signal Δq. Pressure difference signal Δ
The q is supplied to the digital controlled filter 142 via the subtractor 141 and the multiplier 147
Is supplied to the Graham function table 148 via the.

Digitally controlled filter 14
2 is a second-order low-pass filter whose cutoff frequency and amplitude rise ratio (Q) are variables lpfr and ql, respectively.
given by r. Also, the embouchure data emb is supplied to the digital controlled filter 142,
Based on this, characteristics such as a cutoff frequency are set. Further, the digitally controlled filter 14
2 is added to the embouchure data emb via the adder 143. As a result, a frequency characteristic is given to the pressure difference signal Δq, and a frequency selection operation according to the state of the lips of the player is simulated.

Next, the output signal of the adder 143 is weighted by being multiplied by a variable SLTGIN via a multiplier 144 and supplied to a slit function table 145. Slit function table 145, based on the pressure difference signal Δq which the frequency property has been granted, and outputs the opening area signal S _L of an aperture area of the player's lips. The aperture area signal S _L is supplied to a multiplier 149, is also multiplied by a feedback coefficient β via a multiplier 146, and is fed back to a subtractor 141. Thus, each component 141 in FIG.
146, the pressure difference signal Δq, the variable lpfr, the variable ql
Based on r and embouchure data emb, it is understood that the opening area signal S _L is obtained.

Next, according to Graham's law, the flow rate (air velocity v) flowing per unit area per unit time
Is represented by the following equation (A1). v = {2 (Δq) / ρ} (A1) where ρ is the air density. Graham function table 1
48 outputs an air velocity signal v indicating the air velocity when the pressure difference signal Δq is supplied based on the relationship of the above equation (A1). In the preceding stage of the Graham function table 148, a multiplier 147 is used to adjust the influence of the Graham function.
Is provided, where the pressure difference signal Δq is multiplied by a predetermined variable (graham gain) GRMGIN.

As described above, when the opening area signal _SL and the air velocity signal v are obtained, the two are multiplied by the multiplier 149, and the result of the multiplication is output as the flow rate signal f. Next, the flow signal f is supplied to the multiplier 153, and is multiplied by a variable z indicating the input impedance of the resonance tube. Then, the result of the multiplication is output as the pressure traveling wave signal q ₀ via the adder 155.

By the way, in an actual brass instrument, a so-called "chatter" is generated by reflection of a pressure wave in a mouthpiece. In Figure 9, which has Reparushingu function table 150 is provided to simulate and wherein a pressure difference signal Δq is supplied to a chatter signal S _RP indicating the chattering noise is output. This chatter signal S _RP is transformed via an HPF (High Pass Filter) 151 and supplied to a multiplier 152. Note that H
The characteristics of the PF 151 are set by supplying the coefficient HPFR.

The chatter signal S _RP output via the HPF 151 is multiplied by a pressure difference signal Δq via a multiplier 152, further weighted via a multiplier 154 by a variable REPGIN, and supplied to an adder 155. . As a result, the pressure traveling wave signal q ₀ is influenced by the chatter sound. Here, the waveform and frequency spectrum of the analog output signal output from the DAC 103 are converted to FFs.
FIG. 10 shows an example measured by a T (fast Fourier transform) analyzer. FIGS. 11 to 13 show waveforms and frequency spectra of the opening area signal S _L , the air velocity signal v, and the chatter signal S _RP .

[0017]

By the way, in order to create a musical tone in the above-mentioned algorithm, it is necessary to change various parameters (variables). However, in some cases, it is difficult to predict which parameter changes the tone, and it is difficult to clarify the correspondence between the tone color and the parameter. In addition, some parameters change the pitch by changing them, or stop the oscillation.
Therefore, considerable skill was required until various parameters were appropriately changed to produce a desired sound.

It is preferable that the electronic musical instrument is provided with various controls so that the player can appropriately change the parameters during the performance. However, the parameters entrusted to the player are limited for the reasons described above, and Was difficult to improve. The present invention has been made in view of the above-described circumstances, and has as its object to provide a musical sound synthesizer capable of easily producing sounds.

[0019]

According to the present invention, there is provided a linear arithmetic means for receiving a traveling wave signal, propagating and delaying the traveling wave signal as appropriate, and outputting the signal. Non- linear operation means for non-linearly converting the output signal of the linear operation means based on the control information and outputting the output signal as the traveling wave signal, and a plurality of signals generated in the linear operation means or the non-linear operation means are extracted and extracted. Of the signals , the linear operation means or the non-linear operation
The first signal extracted from one of the
After the fluctuation, a second signal extracted from the other is added, and a tone modifying means for outputting the addition result as a tone signal is provided.

[0020]

When the traveling wave signal is input, the linear operation means
The traveling wave signal is propagated and delayed as appropriate and output.
On the other hand, the non-linear operation means non-linearly converts the output signal of the linear operation means based on the musical tone control information and outputs it as a new traveling wave signal. The musical sound modification means extracts a plurality of signals generated in the linear operation means or the non-linear operation means,
Among the extracted signals , the linear operation means or the non-linear
The first signal extracted from one of the
After performing the pitch change, the second signal extracted from the other is added, and the addition result is output as a tone signal. Therefore, a tone signal is formed without affecting the traveling wave signal at all.

[0021]

Embodiment A. Configuration of embodiment A-1. The overall configuration of the embodiment below with reference to FIG. 1～6,14～16 described embodiments of the present invention. In these figures, FIGS.
The same reference numerals are given to portions corresponding to the respective portions of No. 9 and the description thereof is omitted.

FIG. 1 is a block diagram of an electronic musical instrument according to one embodiment of the present invention. In the figure, reference numeral 108 denotes a sound-making effect device, which outputs a pressure reflected wave signal q _i , a signal S ₁ , a signal S ₂ , an opening area signal S _L , an air velocity signal v, and a chatter signal S from the non-linear operation unit 101 and the linear operation unit 102. _RP is supplied. Reference numeral 105 denotes an operator, which includes a keyboard and various tone control operators (not shown). Operation information of the operation unit 105 is supplied to the control unit 106 as a MIDI signal. The control unit 106 outputs musical tone control information based on the supplied operation information. The tone control information designates the modulation depth of AM modulation and FM modulation (details will be described later) executed in the sound creation effect device 108, or the width and period of pitch fluctuation.

The sound making effect device 108 synthesizes a tone signal based on each signal supplied from the non-linear operation section 101 and the linear operation section 102 and control information supplied from the control section 106, and resonates the synthesized tone signal. System 107
Is supplied to the DAC 103 via the.

A-2. The structure of the sound making effect device 108
Formed it will be described with reference to FIG. 2 the configuration of the sound creation effecting device 108. In the figure, reference numeral 1 denotes a carrier MIX unit, which outputs a carrier signal based on each of the signals q _i , S ₁ , S ₂ , S _L , v and S _RP . FIG. 3 shows the circuit configuration. In FIG. 3, each of the signals q _i , S ₁ , S ₂ , S _L , v and S _RP is multiplied by predetermined weighting variables MIXC _{1 to} MIXC ₆ via multipliers 11 to 16, respectively. Are output from the adder 17. After an output signal of the adder 17 is subjected to an LCF (low-pass cutoff filter) 18 to remove a DC component, equalization, that is, sound creation is performed sequentially through an HPF 19 and an LPF 20, and is output as a carrier signal. In addition, HPF19 and LPF
The cutoff frequency of 20 is determined by the predetermined variables HPFC and LPFC
Is set by

Next, in FIG. 2, reference numeral 2 denotes a modulator MI.
X section, and each signal q _i , S ₁ , S ₂ , S _L , v and S _RP
And outputs a modulation signal based on. The modulator M
The configuration of the IX unit 2 is the same as that of the carrier MIX unit 1 (FIG. 3) except that the variables MIXC _{1 to} MIXC ₆ , HPFC and
The variable corresponding to the LPFC is, of course, set independently of the carrier MIX unit 1.

A-3. Configuration of the FM modulator 3 Next, carrier signal outputted from the carrier MIX unit 1 is supplied to the FM modulator 3, the modulator MIX 2
Are FM-modulated based on the modulation signal output from the control unit 106 and the control signal output from the control unit 106. Fig. 4 shows the details.
This will be described with reference to FIG. In FIG. 4, reference numeral 31 denotes a delay circuit, which is constituted by storage elements having a plurality of addresses.
The delay circuit 31 latches the carrier signal for each sampling period, reads the carrier signal into the first address, shifts the information stored in each address sequentially to the subsequent address, and discards information overflowing from the last address.

When a read address is supplied from the multiplier 32, the delay circuit 31 outputs data stored at this address. Therefore, if the read address is kept constant, the carrier signal is output after being delayed for a predetermined time. However, if the read address is changed back and forth, it is understood that the FM modulated carrier signal is output.

Here, the modulation signal output from the modulator MIX unit 2 is supplied to a multiplier 36, where the modulation signal is multiplied by a variable M_DEPTH which specifies a modulation depth. Next, the variable C_DELAY is added to this modulated signal via an adder 35,
The result of this addition is supplied to the multiplier 33 via the multiplier 34. The variable C_DELAY is a variable that specifies the delay time of the carrier signal together with the note-on pitch length data NOPL described later. The multiplier 34 multiplies the modulated signal by the note-on pitch length data NOPL.
Here, the note-on pitch length data NOPL is data indicating the pitch of a musical tone at the time of note-on.

Here, the reason why the data NOPL is multiplied by the modulation signal is as follows. That is, when FM modulation is performed on the carrier signal at a predetermined depth, the change width of the read address becomes larger as the pitch becomes larger (as the pitch becomes lower). Therefore, when the pitch is large, it is necessary to use the carrier signal stored for a longer time, so that the read address needs to be large. On the other hand, when the pitch is small, the read address may be small. The variable C_DELAY also specifies the center of the read address, and a larger value is selected as the modulation depth increases.

As described above, the center of the read address, that is, the read address when the modulation factor is "0" is determined based on the variable C_DELAY and the note-on pitch length data NOPL. Next, the modulated signal is
Is multiplied by the variable C_DEPTH. Where the variable C_DEP
TH is a variable for controlling the modulation depth.
The information is supplied from the control unit 106 together with the above note-on pitch length data NOPL based on various operation information in 05.

Next, the modulated signal output from the multiplier 33 is supplied to the multiplier 32 and multiplied by "1/32". In order to improve the accuracy of FM modulation, the modulation signal is multiplied by "32" in advance, and is multiplied by "1/32" immediately before being supplied to the delay circuit 31.

By providing each of the above components,
In the FM modulator 3, the modulated signal and the controller 106
The carrier signal is delayed and FM-modulated based on various control signals supplied from the control unit. For example, the variable M_
If DEPTH is “0”, variable C_DELAY is “1”, and variable C_DEPTH is “1”, the carrier signal is delayed by one cycle. Similarly, when the variable C_DELAY is “0.5”, the carrier signal is delayed by a half cycle, and when the variable C_DELAY is “0.25”, the carrier signal is delayed by a 周期 cycle. Where the variable C_DEPTH is
It is preferable that the setting be made based on the MIDI signal output from the operation element 105.

That is, the MIDI signal can take a value of "00H" to "7FH" in hexadecimal code, which is converted into a value of "0.0" to "1.0" by the control unit 106, and converted. When the result obtained is given as the variable C_DEPTH, the phase delay of the carrier signal can be controlled in real time. Here, assuming that the variable M_DEPTH is “1”, the variable C_DELAY is “1”, and the modulation signal is a sine wave with an amplitude of “1”, the phase delay of the carrier signal in the delay circuit 31 is as shown in FIG.

As described above, the modulation depth can be changed by reducing the variable M_DEPTH.
Is preferably configured to be able to be appropriately set by an operator or the like. In addition, the delay time is given to the carrier signal in the FM modulator 3 so that the FM signal applied here is
A time lag occurs between the modulation and the AM modulation performed in the AM modulator 4 (details will be described later). Therefore, it is understood that the delay time given in the delay circuit 31 has a great influence on the tone color formation of the musical tone.

A-4. Configuration of AM Modulation Unit 4 Next, the carrier signal output from the FM modulation unit 3 is A
The signal is supplied to the M modulator 4 and is further AM-modulated. The details will be described with reference to FIG. In the figure, the carrier signal output from the FM modulator 3 is multiplied by a variable C_P_LV indicating a carrier power modulation level via a multiplier 41. Similarly, modulator MI
The modulated signal output from the X unit 2 is multiplied by a variable C_M_LV indicating a modulator power modulation level via a multiplier 44. Therefore, both variables C_P_LV, C_
By appropriately setting M_LV, it is possible to determine the ratio between the AM modulation carrier signal and the modulation signal.

Next, the modulated signal output from the multiplier 44 is further multiplied by the variable P_DEPTH via the multiplier 45. This variable P_DEPTH is a variable that specifies the depth of AM modulation, and is generated by the control unit 106 based on the MIDI signal output from the operation unit 105. That is,
As described above, the MIDI signal is “00” in hexadecimal code.
H "to" 7FH ".
In step 6, the value is converted to a value of "0.0" to "1.0", and the converted result is supplied as a variable P_DEPTH. Next, the multiplier 42 multiplies the carrier signal output from the multiplier 41 by the modulation signal output from the multiplier 45, and as a result, the carrier signal is AM-modulated by the modulation signal. The output signal of the multiplier 42 is further output through an LCF (low-pass cutoff filter) 43 from which a DC component is removed.

Next, in FIG. 2, the output signal of the AM modulation section 4 is supplied to an addition circuit 5 and is combined with the carrier signal output from the carrier MIX section 1. Then, the combined result is supplied to the input MIX unit 6 and the output MIX unit 8.

A-5. Input MIX unit 6 and output MIX
Configuration of Unit 8 Next, the configurations of the input MIX unit 6 and the output MIX unit 8 will be described with reference to FIG. The output signal of the adding circuit 5 is supplied to the input MIX unit 6 and the non-linear operation unit 1
01 and each signal q _i , S ₁ ,
S ₂ , S _L , v and S _{R P} are provided. These signals are summed after each coefficient MIXD ₁ ~MIXD ₇ have been multiplied,
The result is supplied to the pitch changer 7. In the pitch changer 7, a pitch change is given to the output signal of the input MIX unit 6 (details will be described later), and the result is supplied to the output MIX unit 8.

In the output MIX unit 8, the output signals of the pitch changer 7, the signals q _i , S ₁ , S ₂ , S _L , supplied from the non-linear operation unit 101 and the linear operation unit 102, are output.
v and S _RP , and the output signal of the adder circuit 5,
Each of the coefficients MIXE _{0 to} MIXE ₇ is multiplied. The signals multiplied by the coefficients are summed and supplied to the resonance system 107.

A-6. Configuration of Pitch Changer 7 Next, the configuration of the pitch changer 7 will be described with reference to FIGS. In the figure, reference numeral 50 denotes an LPF. When an output signal of the input MIX unit 6 is supplied, the LPF removes a high-frequency component from the signal and outputs the signal. Reference numeral 51 denotes a delay circuit including a memory or the like, which sequentially stores output signals of the LPF 50 (hereinafter, referred to as pitch-changed signals). That is, the write address when the pitch-changed signal is written into the delay circuit 51 is initially set to the delay circuit 5
The write address is decremented by "1" every sampling period. When the write address reaches the start address of the delay circuit 51, the write address is set to the last address again at the next sampling time, and the same operation is repeated thereafter.

Here, the principle of operation of the pitch changer 7 will be briefly described. First, as described above, since the output signal of the LPF 50 is sequentially written into the delay circuit 51, the pitch-changed signal is reproduced by reading out the written pitch-changed signal. Here, if the read address at the time of reading the delay circuit 51 is given by the absolute read address pointer ix and the absolute read address pointer ix is decremented by "1" every sampling period, the pitch-changed signal is deformed. Read without

On the other hand, when the decrement amount of the absolute read address pointer ix is set to a value other than "1", it is possible to change the pitch of the pitch change signal. For example, if the decrement amount is set to “2”,
The read pitch-changed signal is converted into a signal having a double pitch, that is, a signal one octave higher. If the decrement amount is set to “0.5”, the decimal part rx of the decremented value (integer part ix + decimal part rx) is rounded, and the result is supplied to the delay circuit 51 as an absolute read address pointer ix. The pitch change signal is converted to a signal one octave lower.

By the way, simply constructing the pitch changer 7 based on the above-described principle causes various problems. First, since the decrement of the write address of the pitch-changed signal is different from the decrement of the absolute read address pointer ix, one may catch up with the other. In this case, the phase of the pitch-changed signal read from the delay circuit 51 is shifted, which adversely affects the tone. Further, when the absolute read address pointer includes a decimal part, if the decimal part is simply removed by rounding, there is a problem that the fidelity of the read pitch-changed signal is reduced.

Therefore, in the pitch changer 7 according to the present embodiment, the following solution is adopted for such a problem. That is, together with the absolute read address pointer ix, another absolute read address pointer iy
Is supplied to the delay circuit 51, and cross-fading is applied to the signals read by the two pointers, thereby coping with the phase shift of the pitch-changed signal.

When both address pointers ix and iy include decimal parts rx and ry, ix + rx and iy
Addresses ix, ix + 1, iy and i before and after + ry
By reading out y + 1 and performing an interpolation operation using the fractional parts rx and ry, the decrease in the fidelity is dealt with. The above is the outline of the pitch changer 7, and the details will be described below.

In the figure, reference numeral 52 denotes a multiplier. When a pitch parameter PITC1 for determining the amount of change in pitch is supplied from the control unit 106, the multiplier 52 multiplies this by a predetermined coefficient and outputs it as an increment value PIC1I. Next, 5
Reference numeral 3 denotes a "16" bit counter, which counts an increment value PIC1I at each sampling cycle, and outputs the result as a count value ct10T.

That is, the initial value of the count value ct10T is “0000H” in hexadecimal notation, and
Incremented by PIC1I. And the count value ct10T
Exceeds "FFFFH" in hexadecimal notation, the overflowed digits are discarded and the counting operation is continued. In addition,
In the following description, unless otherwise noted, numbers suffixed with "H" are in hexadecimal notation.

Here, the time t is plotted on the horizontal axis, and the count value ct10T
When a graph having the vertical axis as is created, the waveform of the count value ct10T has a sawtooth shape as shown in FIG. Assuming that the count value ct10T is a hexadecimal number represented by a complement expression of “1”, the count value ct10T is “−8000H”.
To “7FFFFH”, and can be considered to be a value that returns to “−8000H” when the value exceeds “7FFFFH”.

Next, the count value ct10T is multiplied by “１ ／” through the multiplier 55, and “4000H” is added through the adder 57. As a result, the output signal wp10T of the adder 57 is set to a value as shown in FIG. On the other hand, the count value ct10T is calculated by the adder 54 as “8
000H "is added, and the addition result is output as the count value ct11T. As a result, the count value ct11T becomes a signal advanced by a half cycle with respect to the count value ct10T.

Like the count value ct10T, the count value ct11T is multiplied by “１ ／” via the multiplier 56, and “4000H” is added via the adder 58. Therefore, the adder 5
As shown in FIG. 16 (c), the signal wp11T output from the output signal 8 has the same waveform as the output signal wp10T, and has a phase of 180.
° Advanced signal.

Next, the output signals wp10T and wp11T are multiplied by the coefficient WPAP1 via multipliers 59 and 60, respectively, and the multiplication results are output as relative read address pointers ixs and iys. Here, the coefficient WPAP1 may be a negative value or may have a decimal part. That is, the adder 5
The output signals wp10T and wp11T of 7, 58 are represented by integer values, but the relative read address pointers ixs, iys are 20
The bits are represented by a fixed-point representation, that is, represented by 10-bit integer part and 10-bit fraction part.

Next, reference numeral 63 denotes a 10-bit counter, which counts the value "-1" at each sampling period, and supplies the counting result to the adders 61 and 62. Then, the counting result and the relative read address pointers ixs, iys are added to obtain the absolute read address pointer i.
x, iy and their fractional parts rx, ry are output.

When both address pointers ix and iy are supplied to the delay circuit 51, the address i in the delay circuit 51
Data D (ix), D (ix + 1), D (iy) and D (i) stored in x, ix + 1, iy and iy + 1
y + 1) is read out and supplied to multipliers 70 to 73, respectively. In the multipliers 70 to 73, interpolation counts 1-rx, rx, 1-
ry and ry are multiplied. And multipliers 70 and 7
The 1 output signal is added in the adder 74, and the addition result is output as a pitch change signal X. Similarly, the output signals of multipliers 72 and 73 are added in adder 75, and the addition result is output as pitch change signal Y. That is, the pitch change signals X and Y are expressed by the following equation (A
2) and (A3).

X = (1−rx) · D (ix) + rx · D (ix + 1) (A2) Y = (1−ry) · D (iy) + ry · D (iy + 1) (A3)

On the other hand, the count value ct output from the adder 54
11T is supplied to the absolute value conversion circuit 76, and the absolute value env10T of the count value ct11T is output. Accordingly, the waveform of the absolute value env10T is as shown in FIG. Next, reference numeral 77 denotes a level shift circuit which subtracts the shift value EVOF1 from the absolute value env10T, and outputs the subtraction result as a value ev10T. However, if the subtraction result is a value less than “0”, the value ev10
T is set to “0”. Therefore, the waveform of the value ev10T is as shown in FIG.

Next, a multiplier 78 multiplies the value ev10T by a predetermined coefficient EVAP1, and outputs the result of the multiplication as a cross-fade coefficient evx. However, when an overflow occurs in the multiplication result, the maximum value “7FFFFH” is set in the crossfade coefficient evx. Therefore,
The crossfade coefficient evx is, as shown in FIG.
It has a trapezoidal waveform. Next, 79 is a subtractor,
The crossfade coefficient evx is subtracted from “7FFFH”, and the subtraction result is output as the crossfade coefficient evy. Accordingly, the waveform of the crossfade coefficient evy is shown in FIG.
(g).

The two crossfade coefficients evx and evy are supplied to multipliers 80 and 81, respectively, and are multiplied by the pitch change signals X and Y. Then, these multiplication results are added in the adder 82 and output as a signal opc1T. That is, as is clear from comparison of FIGS. 16F and 16G, it is understood that the cross fade is applied to the pitch change signals X and Y. After the high frequency component is removed from the signal opc1T through the LPF 83, the output MIX unit 8 (FIG. 14)
).

B. Next, the operation of this embodiment will be described with reference to FIG. First, when the blowing pressure signal pre and the embouchure signal emb are supplied to the non-linear operation unit 101, a traveling wave pressure signal q ₀ is generated based on these, and supplied to the linear operation unit 102.
In the linear operation unit 102, the traveling wave pressure signal q ₀
Is propagated while being delayed and attenuated, and is fed back to the nonlinear operation unit 101 as a reflected wave pressure signal q _i . As described above, the signals are exchanged between the non-linear operation unit 101 and the linear operation unit 102, whereby the propagation of the pressure traveling wave / reflected wave in the wind instrument is simulated.

Next, the respective signals q _i , S ₁ , S ₂ , S _L , v and S _RP are extracted from the non-linear operation part 101 and the linear operation part 102 and supplied to the sound making effect device 108.
In the sound making effect device 108, the carrier signal and the modulation signal are synthesized based on the supplied signals,
The carrier signal is subjected to FM modulation and AM modulation by the modulation signal. Further, the output signal of the carrier MIX unit 1 and A
The output signals from the M modulator 4 are added in an adder 5, and the addition result is supplied to a pitch changer 7 via an input MIX 6. Then, the addition result of the addition circuit 5 and the output signal of the pitch changer 7 subjected to the pitch change are synthesized in the output MIX unit 8, and the result is output as a tone signal via the resonance system 107 and the DAC 103.

On the other hand, the keyboard and various controls are operated by the player on the controls 105, and the operation information is supplied to the control unit 106 as MIDI signals. In the control unit 106, various control signals are output based on the supplied operation information, and the sound creation effect device 10
Various parameters in 8 are set. As described above, in the present embodiment, the sound making effect device 108 is provided outside the loop in which the traveling wave pressure signal q ₀ and the reflected wave pressure signal q _i are propagated, and the sound is made here. Therefore, even when the parameters are arbitrarily changed, it is possible to prevent problems such as stopping oscillation or shifting the pitch.

In this embodiment, the sound can be created by using parameters such as a variable C_DEPTH specifying the depth of FM modulation and a variable P_DEPTH specifying the depth of AM modulation. It is easy to grasp how the musical tone changes when the music is moved, and it is possible to easily create a desired sound. further,
In the present embodiment, since the non-linear operation unit 101 and the linear operation unit 102 are provided, characteristics unique to the physical model sound source, for example, the blowing pressure signal pre or the embouchure signal e
It also has a feature that music can be controlled by mb etc.

Here, the setting of the coefficients of each part in the sound making effect device 108 will be described. As described above, in the sound making effect device 108, the signals S _L , v and S supplied from the non-linear operation unit 101 are output.
_RP and q _i , S ₁ and S ₂ supplied from the linear operation unit 102 are used after being appropriately mixed. The signal supplied from the non-linear operation unit 101 is a signal simulating the behavior of the lead of a natural wind instrument, and thus has a hard tone. On the other hand, since the signal supplied from the linear operation unit 102 is a signal simulating the wind of the wind instrument, it has a soft tone. Therefore, by appropriately setting the mixing ratio of these signals in the carrier MIX unit 1 or the output MIX unit 8 or the like, the timbre hardness can be adjusted.

[0063] Further, in the input MIX section 6 of FIG. 14, it is possible to set the mixing ratio of the pitch change signal supplied to the pitch changer 7 I'm setting the coefficients MIXD ₁ ~MIXD ₇ appropriately. Here, according to the experiment of the present inventor, the pitch change signal is a signal supplied from the non-linear operation unit 101 or a linear operation unit 102.
It has been found that it is preferable to select only one of the signals supplied from. For example, the linear operation unit 10
When 2 is selected, the coefficients MIXD _{1 to} MIXD ₃ are preferably set to “0” and the coefficients MIXD ₄ to MIXD ₆ are set to values other than “0”. On the other hand, when the non-linear operation unit 101 is selected, the coefficients MIXD _{1 to} MIXD ₃ are set to values other than “0” and the coefficients MIXD _{4 to} M
IXD ₆ should be set to “0”.

By adopting such a setting method, a suitable synthesized sound close to a natural musical instrument was obtained. On the other hand, if a selection other than the above, for example, a selection is made such that the coefficients MIXD ₁ and MIXD ₄ are set to values other than “0” and the other coefficients are set to “0”, as far as the present inventor has performed the experiment, Did not give good results. The reason why this phenomenon occurs is difficult to explain with the sound theory of natural wind instruments that is currently elucidated, and further theoretical elucidation is awaited.

[0065] Therefore, the embodiments of the present invention, For example <br/>, tone modification means, generated by the signal extracted from one of said first signal the linear operation means or said non-linear operation means In addition, it is considered that the second signal is generated by a signal extracted from the other of the linear operation means or the non-linear operation means.
Can be

As described above, according to the tone synthesizer of the present invention, the linear operation means or the non-linear operation is performed based on the signal extracted from the linear operation means or the non-linear operation means.
The first signal extracted from one of the
After performing the pitch change, the second signal extracted from the other is added, and the addition result is output as a tone signal.
A tone signal can be formed without affecting the traveling wave signal at all. Therefore, problems such as stop of oscillation or fluctuation of pitch between the non-linear operation means and the linear operation means are prevented beforehand, and sound can be freely and easily made. As a result, the extracted
The signal is used as a first signal, which is subjected to pitch fluctuation,
The second signal extracted from the non-linear operation means is added
When a tone signal is formed, the nonlinearly transformed progress
Rich overtone components and complex subtle pitch changes in wave signals
It is possible to obtain musical tones with emphasized dynamics,
The signal extracted from the arithmetic means is used as a first signal and
The pitch variation is applied, and the second
Are added together to form a tone signal,
Mainly tones simulating various vibrations in the sound tube,
The complex doubling included in the nonlinearly converted traveling wave signal
It can be added while controlling sound components, fluctuations, and fluctuation components
It becomes possible.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an electronic musical instrument according to an embodiment.

FIG. 2 is a block diagram of a sound making effect device 108.

FIG. 3 is a block diagram of a carrier MIX unit 1.

FIG. 4 is a block diagram of an FM modulator 3;

FIG. 5 is an explanatory diagram of the operation of the FM modulation unit 3;

FIG. 6 is a block diagram of an AM modulation unit 4.

FIG. 7 is a block diagram of a conventional physical model sound source.

FIG. 8 is a block diagram of a linear operation unit 102.

FIG. 9 is a block diagram of the nonlinear operation unit 101.

FIG. 10 is a diagram showing an analog output signal of the DAC 103.

11 is a diagram showing an opening area signal S _L.

FIG. 12 is a diagram showing an air velocity signal v.

FIG. 13 is a diagram showing a chatter signal S _RP .

FIG. 14 is a block diagram of an input MIX unit 6 and an output MIX unit 8;

FIG. 15 is a block diagram of a pitch changer 7;

16 is a waveform chart of each part of the pitch changer 7. FIG.

[Explanation of symbols]

 Reference Signs List 101 Nonlinear operation unit (nonlinear operation means) 102 Linear operation unit (linear operation means) 108 Sound creation effect device (musical sound modification means)

Claims (1)

(57) [Claims] When a traveling wave signal is input, a linear operation means for appropriately propagating and delaying the traveling wave signal and outputting the signal, and an output signal of the linear operation means based on musical tone control information.
Non- linear operation means for performing non-linear conversion and outputting as the traveling wave signal; extracting a plurality of signals generated in the linear operation means or the non-linear operation means ;
Either the linear operation means or the non-linear operation means
After performing pitch fluctuation on the first signal extracted from
And a tone modifying means for adding the second signals extracted from the two signals and outputting the result of the addition as a tone signal.