JPS5846038B2 - electronic musical instruments - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an electronic musical instrument that forms musical tones using a frequency modulation method.

Various proposals have been made in the past as methods for forming musical tones.

For example, musical sound waveforms that produce a certain tone are stored in a memory, and the musical sound waveforms are sequentially read out from this memory, or a sound source waveform containing abundant harmonic components is passed through a filter to attenuate appropriate components to create a tone. Typical examples include a method in which each harmonic is generated separately, and a method in which the amplitude of each harmonic is separately controlled to achieve a desired tone color.

However, all of these conventional methods have a limit to the degree of freedom in changing the timbre, and cannot be used for musical sounds that contain integer or non-integer order harmonic components in complex ratios and whose content changes in a complex manner over time. was extremely difficult to form.

The present invention forms musical tones using a method completely different from the conventional method described above, and easily obtains musical tones that contain harmonic components of integer order or non-integer order in a complex manner, and in which the content rate changes in a complex manner. It aims to provide electronic musical instruments.

According to the present invention, the first means detects a pressed key and assigns the sound of the key to any of a specific number of channels;
a second means for generating phase information whose value sequentially changes at a speed corresponding to the musical tone frequency of the key assigned to each channel for each channel according to the channel assignment of the means; A carrier wave phase component and a modulated wave phase component are formed for each channel based on the phase information of each channel, and a frequency-modulated signal wave is generated for each channel based on these phase components and an arbitrary modulation index. The third part sends out the sound as a musical tone signal.
means, so that a plurality of tones can be generated simultaneously in response to musical tone signals sent from the third means according to the assignment.

First, the principle of forming musical tones by the frequency modulation method used in this invention will be explained.

A musical tone formation method using a frequency modulation method is that since the frequency modulated signal wave contains many sidebands, the frequency modulation signal with these sidebands superimposed and many harmonic components are superimposed. This method focuses on the commonality with the musical tone signal that has been created, and forms a musical tone signal (base) by performing frequency modulation in the audio frequency range.

The general formula for the frequency-modulated signal (modulated signal wave) e is expressed as shown in equation (1).

e = As1n (at + I sinβt)...
・・・・・・・・・・・・・・・(1) Here, α is the angular frequency of the carrier wave, β is the angular frequency of the modulated wave, ■ is the modulation index, A
is the maximum amplitude and t is the time.

Expanding equation (1), the signal e has angular frequencies α±β and α±2β.

It can be seen that it consists of a large number of sidebands α±3β, . . . .

Bessel function J of modulation index ■. (■). Jl(■), J
2(■), J3(■), .

Bessel functions J of the carrier wave and the first to fifth order sidebands.

The curve (■) to J5 (I) is shown in FIG. 1 as an example, and as is clear from this graph, it is a Bessel function J.

(I) to Jn(I) are always positive until the modulation index ■ is about 2.5, but when ■ becomes larger than about 2.5, they can take either positive or negative values.

Also, as can be seen from equation (2), the Bessel function J.

(I: Regardless of whether Jn(I) is positive or negative, the positive or negative signs of the odd-order sidebands are different.

Therefore, in the modulated signal wave e expressed by equation (2) (ie, equation (1)), phase inversion occurs in various sidebands.

For example, when the modulation index I = 4, the amplitude coefficient of each order sideband is Jo(■)≠-0.4, Jl(I)≠-0.05゜J
2(I)7=0.35, J3(I)pu0.42,
J4(I)ge0.3°J5(I)=o, 15, and the frequency spectrum of each sideband is as shown in FIG.

C represents the frequency of the carrier wave, and m represents the frequency of the modulated wave.

Frequencies with negative amplitude coefficients simply have inverted phases, and do not have much significance unless the same frequencies coexist.

However, at the same frequency whose phase is shifted by 180°, the amplitudes are canceled by each other due to phase inversion, which has an important meaning.

The fact that frequencies whose phases are shifted by 1800 are mixed in one modulated signal wave e is explained by "reflection of side waves."

"Reflection of side waves" is caused by the presence of side waves in the negative region of OHz or less in the sideband frequency spectrum.

The side waves in the negative region are actually folded back (reflected) and appear in the positive region.

That is, the negative frequency "-8in(1)t" of the angular frequency ω
'' is sin ωt=sin(-ωt), and it can be seen that this is a signal with the phase of the frequency ``sin ωt'' in the positive region inverted.

In this way, the side waves in the negative region have their phases reversed and are reflected back to the positive region.

This reflection causes the side waves in the negative region to be mixed with the side wave components in the positive region.

Due to this mixing, the frequency relationship in the modulated signal wave e becomes extremely diverse.

In this regard, the carrier wave C is 100 Hz, the modulating wave m is 1
An example in which the frequency is 00 Hz and the modulation index is I-4 will be explained.

Since the general frequency spectrum at I-4 is as shown in FIG. 2, in the above example, the frequency spectrum appears as shown in FIG. 3a.

That is, the low-frequency primary sidewave C-m is OHz, and the low-frequency secondary and higher-order sidewaves fall into the negative region.

These negative region side waves C-2m, C-3m...
... is inverted in phase, as shown in Figure 3, and O
It is folded back into the positive region around Hz.

The folded side waves are the side waves C, C+m, and C+2m in the positive region.
・・・・・・・・・is added mathematically.

The folded side wave C-2m is added to the carrier wave C, the side wave C-3m is added to the side wave C+m, and so on.

By addition, frequencies with different signs are canceled, and frequencies with the same sign are increased in amplitude.

Therefore, if the amplitude in FIG. 3 is expressed in terms of absolute amplitude, it will be as shown in FIG. 3C.

According to FIG. 3C, it can be seen that the frequency spectrum of the modulated signal e consists of overtones C, 2C, 3C, . . . of the carrier wave C.

From the above, it has become clear that frequency modulation creates a signal containing harmonic components similar to the frequency components of musical tones.

By the way, the spectral components of a musical tone signal (modulated signal wave e) obtained by frequency modulation are composed of a carrier wave C and a modulated wave m.
depends on the frequency ratio of and the value of the modulation index ■.

That is, the frequency ratio C//rrI determines the positional relationship of harmonic components, and the modulation index ■ determines the bandwidth of the modulated signal e, so it defines the number of harmonics that have a substantial amplitude. , is already known.

To explain this point in more detail, a harmonic spectrum is obtained when the frequency ratio C/m is an integer ratio.

In other words, when C7m is reduced, N1 and N2
When becomes an integer, a harmonic spectrum occurs.

Since N1/N2 is an irreducible fraction, the fundamental frequency (first harmonic) fo of the modulated signal wave e is expressed by the following relational expression.

It is known that in a harmonic spectrum, the position of a harmonic component is determined by the following relational expression.

However, n is the order of the side wave, and n=0, 1, 2°3...
It takes the value...

K is the harmonic order.

Needless to say, all harmonic components in the harmonic spectrum are of integer order, and as is clear from the above equation (3), the carrier wave C is always an N1-order harmonic.

If N2-1, the spectrum of the modulated signal wave e contains all integer-order harmonics (as long as the modulation index () allows), and the modulated wave m is the fundamental wave f.

becomes. As a further example, if N2 is an even number, the spectrum will contain odd harmonics, and if N2=3, the third harmonic will be absent from the spectrum.

In addition to the harmonic spectrum as described above, a non-harmonic spectrum can also be obtained.

An anharmonic spectrum appears when the frequency ratio C/m is not an integer ratio.

That is, since C7m is a non-integer ratio, folding of side waves in the negative region appears between side wave components in the positive region, thereby making the spectrum an anharmonic series.

The non-harmonic spectrum contains non-harmonic waves, and in this invention, the non-harmonic waves are referred to as non-integer harmonics.

Note that the modulated signal wave e regardless of the harmonic or aharmonic spectrum
Defining the fundamental frequency in , it can be said to be the lowest frequency component in the positive region spectrum including the return from the negative region.

By specifying the fundamental wave frequency and obtaining the modulated signal e by frequency modulation, it is possible to form a musical tone signal with a predetermined pitch.

In this invention, the fundamental frequency can be specified using the keyboard.

As is clear from the above, if the carrier wave C or the modulated wave m is changed, the frequency ratio C/m also changes, and the spectral components can be changed freely.

Further, as is clear from FIG. 1, the amplitude and harmonic number of each spectral component can be changed by changing the modulation index (2).

In the present invention, this property is utilized to create a desired timbre and to change the timbre of musical sounds over time.

By the way, the fundamental wave may be lost in the modulated signal wave e depending on the folding position of the side wave in the negative region or the value of the modulation index {circle around (2)}.

For example, if the side waves in the negative region are folded back to the position of the fundamental wave with the same amplitude and inverted phase with respect to the fundamental wave, or when the carrier wave C becomes the fundamental wave, the modulation index 1 corresponds to the amplitude J of the carrier wave.

(Fig. 1) takes a value such as O, the amplitude of the fundamental wave becomes 0, and the fundamental wave is lost.

Furthermore, even in other cases, the amplitude of the fundamental wave may become considerably small in the harmonic spectrum.

If the modulated signal wave e loses its fundamental wave, or if the amplitude of the fundamental wave becomes extremely small, it will no longer be valid as a musical tone, so such disadvantages must be overcome.

Therefore, by superimposing a special fundamental wave frequency on the modulated signal wave e, it is possible to reliably generate a musical tone signal.

That is, as the basic equation for the musical tone signal E, the following is used: the above-mentioned equation (1) plus the fundamental wave component "a sin γt".

Here, a is the maximum amplitude of the fundamental wave component, and γ is the angular frequency of the fundamental wave.

Note that the general formula for frequency modulation shown in equation (1) above is
Furthermore, it can be extended to various frequency modulation function expressions.

For example, when two modulated waves compete to modulate one carrier wave, the modulated signal wave e1 becomes el - Asin (αt+11sinAt+I2s
inβ2t )−(5).

Here, β1. β2 is the angular frequency of the two modulated waves, ■1,
1□ are the two modulation indices, and α is the angular frequency of the carrier wave.

As is clear from expanding Equation (5), the signal e1 consists of a large number of complex sidebands.

These sidebands are: ■1. ■ Bessel function of 2 Jo(1
1), J, (It), Jn
(It), JO(I2).

The amplitude is determined by Jl(I2), Jn(I2).

Suppose that the ratio of carrier wave and modulated wave is α:β. :β2=1:
When the ratio is set to 0.1:1, the spectrum of the signal e1 becomes as shown in FIG.

That is, each harmonic f1. f2. This results in a complex spectrum in which side waves exist at intervals of β1 on both sides of f3°I4, . . . .

At this time, the magnitude of the harmonic is J.

(11) and J. (I2) to Jn(I2), and the magnitude of the side wave is determined by the product of Jo(11) to Jn(1).
1) and Jo(I2) to Jn(I2).

Also, if you modulate the same carrier wave separately with two modulated waves,
The modulated signal wave e2 is as follows.

The signal e2 obtained by equation (6) is
It is the same as the superposition of two different signals e obtained by .

Moreover, the carrier wave is set to two different angular frequencies α1. When combined with α2 and modulated with one modulated wave, the modulated signal e3 becomes as follows.

Musical tones can also be formed using complex frequency modulation as shown in equations (5) to (7) above.

Hereinafter, one embodiment of the present invention will be described in detail with reference to the accompanying drawings.

In the embodiment shown in FIG. 5, the musical tone signal e(t) is obtained according to the following formula.

Equation (8) is substantially equivalent to Equation (4), and the difference is that the amplitude, carrier wave, modulation wave, modulation index, etc. change as a function of time.

To explain Equation (8) with reference to Equation (4), the value qR
represents the phase γt of the fundamental wave, and the value A1(t 2 ) is the maximum amplitude of the specially provided fundamental wave component sin q R as a function of time t.

The phase αt of the carrier wave is given by the value 1(t)qR, and is obtained by multiplying the phase qR of the fundamental wave by a function of time/(t)G.

The phase βt of the modulated wave is given by the value m(t)qR, and is obtained by multiplying the phase qR of the fundamental wave by a time function m(tm).

The modulation index (2) is a function of time I(t) and can be changed over time.

The value A2(t) is such that the maximum amplitude of the modulated signal portion is given as a function of time t.

The value R is a numerical value related to the fundamental frequency of the musical tone to be generated, and is a numerical value proportional to the phase of the fundamental frequency during a fixed sample period of the waveform amplitude.

The value q increases sequentially as the sample point advances, 1°2.3...
It is a variable which increases as
, 3, etc., and the phase advances.

Time-division allocation operation for producing multiple tones In response to the suppression of keys on the keyboard 1, the calculation of equation (8) is executed, and a musical tone signal e(t) is generated in real time.

Moreover, in accordance with the pronunciation assignment in the key assigner 2, the calculation of equation (8) above is executed in a time-sharing manner for a plurality of tones.

The keyboard 1 includes three types of keyboards, for example, an upper keyboard, a lower keyboard, and a pedal keyboard, and key switches are arranged corresponding to the truncated cones.

The key assigner 2 is connected to the keyboard 1 as shown schematically in FIG.
A key press detection circuit 21 that detects the on/off operation of each key switch in the key press detection circuit 21, and an assignment that allocates the information of the pressed key to one of each channel corresponding to the maximum number of simultaneous polyphonic sounds based on the detection result of the key press detection circuit 21. It includes a circuit 22.

The information on each pressed key sequentially output from the pressed key detection circuit 21 is represented by, for example, a multi-bit code signal (key code KC) representing the name of the suppressed key, and has different contents for each key. .

The allocation circuit 22 includes a code storage circuit 221 with a specific number of storage circuits corresponding to each channel.

When the key code KC from the key press detection circuit 21 is stored in one of the key code storage circuits 221, that key code is assigned to the channel corresponding to that storage circuit.

As is well known, the conditions for this allocation operation are as follows: (A) Allocate to a memory circuit (blank channel) that has not yet been stored.

(B) Assign so that the same key code is not stored in multiple memory circuits redundantly.

That's true.

It is convenient to configure the key code storage circuit 221 with a circular shift register that includes a gate on the input side.Assuming that the total number of channels is 12 and the number of bits of the key code KC is 9, it can store 12 words (9 bits per word). ) shift register is used, and the stored (already assigned) key code KC* is fed back to the input side.

This shift register 221 is sequentially shifted in accordance with the main clock pulse φ1, and as a result of this shift, the memory key code K of each channel is output from the final stage in a time-divisional manner.
C* is used for generating tone waveforms.

That is, the key code KC* becomes address data for reading out the frequency information storage device 3, which will be described later.

The main clock pulse .phi.1 is generated at intervals of 1 .mu.S, for example, as shown in FIG. 7a.

The 1 μs time slot formed by this main clock pulse φ1 is used sequentially for information processing from the 1st channel to the 12th channel, as shown in FIG. Let's call each channel time.

Therefore, the channel times from the first channel to the twelfth channel are cyclical, and each device of the present invention is dynamically arranged to operate in synchronization with each channel time.

The storage key code KC* of each channel is output in a time-division manner in synchronization with this channel time.

In the assignment circuit 22, a key code comparison circuit 222 compares the contents of the input key code KC and the stored key code KC*, and outputs a comparison result depending on whether they match or do not match.

This comparison confirms whether condition mB) of the allocation is satisfied.

The manual key code KC from the key press detection circuit 21 is continuously supplied only during the period when the stored key codes KC* of all channels circulate twice.

The above comparison is made in one cycle period in the first half. This comparison result is translated into the comparison result storage circuit 223 and output during one circulation period in the latter half.

The above assignment clause mA); Well, it can be known by detecting the presence or absence of the memory key code KC*.

The memory key code detection circuit 224 outputs signal 1 at the channel time when the key code KC* exists, and the key code KC*
The signal O is output during the channel time when there is no channel (blank channel).

The output of the detection circuit 224 is used for musical tone control as an attack start signal AS indicating that a key has been pressed (indicating that a key code has been stored in that channel and a sound generation assignment has been made).

Further, the output of the detection circuit 224 is used for determining the condition (A) ¥11.

The set/reset signal generation circuit 225 confirms that both conditions (A) and (B) are satisfied based on the outputs of the comparison result storage circuit 223 and the stored key code detection circuit 224, and generates a new key code KC. A set signal S and a reset signal C are generated for one channel time to be allocated.

This set signal S1 and reset signal C are applied to the gate of the key code storage circuit 221, control the gate, reset the feedback input side, and at the same time input a new input key code KC.
is stored in the first stage of the shift register 221.

In this way, the key code KC is assigned to the channel corresponding to the channel time.

Further, in order to detect a key-off, the key press detection circuit 21 sends a start code indicating the start point of key-off determination almost regularly in place of a key code representing a key switch.

The detection circuit 226 detects that this start code has been sent.
Detects this and generates a forced off signal X.

The key-on 1 time memory circuit 227 has a memory stage corresponding to each channel, and the key code KC is stored in each channel.
When the set signal S is generated to store the channel, the signal 1 is stored in the stage corresponding to the channel.

This memory is forcibly reset by the forced off signal Store signal 1 again.

The key-off memory circuit 228 also has a memory stage corresponding to each channel, and when the forced-off signal The human power for the key code KC assigned to is already canceled, that is, the key associated with the key code has already been released (pressed/released).
It is determined that the key-off memory circuit 228 has a signal 1 in the memory stage of the channel concerned.

This signal 1 becomes a signal DS representing key release.

This signal DS is used as a decay start signal for musical tone control, which will be described later.

It should be noted that the amplitude values A1(t) and A2(t) determine the amplitude of the musical tone signal e(t) obtained by calculating the equation (8). ) becomes 0, respectively, are generated from amplitude information generating circuits 7 and 18, respectively, as will be described later.

The fact that the amplitude values A, (t) and A2(t) both become O indicates that the sounding of musical tone e(t) has ended.

Therefore, the AND circuit 23 detects that both the decay end signals DF1 and DF2 have become signal 1, and thereby it is possible to know that both the values A1(t) and A2(t) have become O, that is, the end of sound generation. can.

The output << 177 of the AND circuit 23 is applied to the OR circuit 24 as a sound generation end signal (total decay end signal) DF.

The reset signal C is also applied to the OR circuit 24, and the output of the circuit 24 is output as a counter signal CC.
Used to reset various counters, memories, etc.

By the way, as mentioned above, the keyboard 1 in this embodiment is composed of three types of keys.

Assuming that the key code KC (KC*) is a 9-bit code signal, the maximum 16 combinations of 4-bit codes are the 12 note names C, C", D, ......AΦ, B
Up to 8 combinations of 3-bit chords can be used to differentiate between octave ranges within one keyboard, and up to 4 combinations of the remaining 2-bit chords can be used to differentiate between octave ranges within one keyboard. It can be used to distinguish between three types of keyboards.

Therefore, among the stored key codes KC*, the 2-bit code representing the keyboard type is 1. 2 is added to the decoder 229 to detect the keyboard type to which the key specified by the key code KC* belongs.

If it is an upper keyboard, an upper keyboard signal UE is output, if it is a lower keyboard, a lower keyboard signal LE is output, and if it is a pedal keyboard, a pedal keyboard signal PE is output.

Each keyboard signal UE-PE is used for musical tone control for each keyboard.

In addition, the input/output signal of the assignment circuit 22 (input key code K
Each signal except C: KC*, AS, DS, CC.

DF1, DF2, etc.) are generated time-divisionally in synchronization with the time of each channel.

The structure of the key assigner 2 is not limited to that shown in FIG. 6, but any structure can be used as long as it can allocate information about pressed keys to each channel.

For example, patent application No. 47-125514, which is already publicly known.
It is also possible to adopt a configuration as disclosed in the specification of No. 49-84216 (Japanese Unexamined Patent Publication No. 49-84216) with the title of the invention "Key Assigner".

Generation of phase information qR The key codes KC* assigned to each channel, which are output in a time-divisional manner from the key code storage device 221 of the key assigner 2, are sequentially added to the frequency information storage device 3.

The frequency information storage device 3 stores in advance the value R of the formula (8) (hereinafter referred to as frequency information) corresponding to the frustum frequency represented by the key code KC* at the address corresponding to the key code. When a key code is input, the frequency information R stored at the address specified by the key code is read out.

The frequency information R is, for example, 15-bit binary data, where a total of 14 bits of data from the lowest digit to the 14th bit represent the value of the decimal part, and 1 bit data of the 15th bit represents the integer part. represents the value of

The frequency information R read out sequentially from the frequency information storage device 3 in a time-division manner is added to a 12-word (1 word - 21 bit) cyclic counter 4, and is time-divisionally read out every fixed time (every 12 channel times). It is cumulatively counted.

In the counter 4, a total of 7 bits of data from the 15th bit to the 21st bit (most significant digit) are treated as integer part data.

The counter 4 is composed of an adder 41 with a capacity of 21 bits and a shift register 42 with 12 words (1 word = 21 bits), and is shifted in accordance with the clock φ1.
After the channel time, the data output from the final stage of the shift register 42 is fed back to the adder 41 and added to the output R of the frequency information storage device 3.

Thus, the value R increases sequentially every 12 channel times: R, 2R, 3R.

It takes a value of 4R... (=qR).

In this way, the phase information qR of the sound assigned to each channel is output from the counter 4 in a time-division manner in synchronization with the time of each channel.

By the way, if the 12 channel time is 12 μS,
The number of times the value R is accumulated in one second by counter 4 is 12 ×
106 times.

Therefore, the number q increases as the phase of the fundamental wave 1 wave advances.
When the fundamental wave frequency is f (Hz), one sine wave waveform for creating the waveform sin q R of the fundamental wave component is stored in the sine wave waveform storage device 5 as the sampling number (
Assuming that 64 addresses (number of addresses) are stored, when the phase information qR reads the final address, qR=64.

Therefore, the value of frequency information H (in decimal notation) to be stored in the frequency information storage device 3 is R=12X64XfX10-
It becomes 6.

Frequency information R given by the above formula is stored in the storage device 3 in binary format in correspondence with each key code.

The phase information qR output from the musical tone signal generation counter 4 is supplied to three processing series A, B, and C, respectively.

Each processing series A, B, and C is a circuit that executes the calculation of each term on the right side of equation (8).
1(t)sin q RJ is calculated, and sequences B and C are frequency modulation terms ``A2(t)sin(l(t)qR+I(t)s
in (m(t)qR), calculate lj.

Therefore, in series A, phase information qR is literally used as a signal corresponding to the phase of the musical tone signal, but in series B and C, it is used as basic data for introducing the phase elements of the carrier wave and modulated wave in the frequency modulation formula. used as.

Incidentally, a sufficient effect can be obtained by using only the integer part of the upper 7 bits of the 21 bits of the phase information qR in the series A-C.

First, to explain the calculation of the fundamental wave component, the sine wave waveform storage device 5 is configured by, for example, a read-only memory, and divides one period of the sine wave waveform into an appropriate number of samplings (for example, 64), and calculates the amplitude value at each sampling point. are stored in advance in correspondence with respective predetermined addresses.

The sine wave waveform storage device 5 receives the phase information qR as an address input and reads out the amplitude value of the waveform sample point corresponding to the address.

In this way, the sine wave waveform amplitude value corresponding to the phase at each point in time is outputted in real time, and the sine wave 5inqR is output from the storage device 5.
will be output.

This sine wave signal (fundamental wave signal) sin qR is applied to a multiplier 6.

Other inputs of the multiplier 6 include maximum amplitude information A of the fundamental wave component;
(t) is added, and the multiplication in the multiplier 6 outputs the calculation result of the fundamental wave component element A1(t)sinqRJ.

Needless to say, such calculation of the fundamental wave component is performed on a time-division basis for each channel.

The maximum amplitude information A1(t) is generated from the amplitude information generating circuit 7 in a time-division manner for each channel in synchronization with the channel time.

The amplitude information A1(t) changes over time, and is a so-called envelope waveform that is generated in response to a key press and attenuates as the key is released.

Therefore, for the amplitude information generating circuit 7, a circuit already known as an envelope generator can be used.

FIG. 8 shows an example of the amplitude information generating circuit 7, which operates in response to the attack start signal As and the decay start signal DS supplied from the key assigner 2, and generates amplitude information as shown in FIG. Digitally generate the envelope of A1(t).

When the attack start signal AS is applied to the AND circuit 71, the attack clock pulse AC output from the attack clock oscillator 75 is applied to the AND gate group 79 via the AND circuit 71 and the OR circuit 14.

The AND gate group 79 receives the signal 1 via the inverter 60.
is added, the 1 addition data P1 is selected by the gate group 79 and output in synchronization with the attack clock AC.

1 addition data P1 is n-bit data, the least significant bit (first bit) is "1", and the other bits (second to nth bits) are all "0".

The 1-added data P output from the AND gate group 79 is input to the n-bit adder 62 via the OR gate group 61.

The output of the adder 62 is 12 through the AND gate group 63.
The word (1 word - n bits) is input to the shift register 64, delayed by 12 channel times according to the clock φ, and then output from the aperture register 64.

The output of the register 64 is fed back to the other input side of the adder 62 and added to the data from the OR gate group 61.

Therefore, the data of the channel held in the register 64 increases by 1 in accordance with the attack clock AC.

The output of the shift register 64 is sent to the attack curve memory 65.
and a decay curve memory 66, the memory 6
This is an address signal for reading out the attack curve or attack curve stored in 5.66.

At the time of attack, only the attack curve memory 65 can be read, and the decay curve memory 66 does not operate.

Therefore, as the output of the register 64 increases sequentially during the channel time, an attack curve as shown in the attack portion of FIG. 9 is sequentially read out.

When all n-bit outputs of the shift register 64 become '1', the maximum value of the attack curve is read out, and this is detected by the AND circuit 67.

When the readout of the attack curve is completed, the AND circuit 67
The output becomes signal 1 and is stored in a 12-bit shift register 69 via an OR circuit 68.

The signal 1 stored in the shift register 69 is output at the time slot of the channel after 12 channel time, and is self-held in the register 69 via the AND circuit 50.

The output of the register 69 is a signal AF representing the end of the attack, and when this signal AP becomes <11'', the AND circuit 71 is inhibited and the AND circuit 72 becomes operable.

Also, since the attack end signal AF is input to the inverter 60, the AND gate group 79 is input to the output of the inverter 60? (Inhibited by 0''.

This time, the AND gate group 78 to which the signal AF is applied becomes operational, and the first decay clock oscillator 76
The first decay clock DC1, which is oscillated from is input to the adder 62 via the OR circuit 61.

The 1-subtraction data M1 is n-bit data, and all bits are "1".

Therefore, by adding 1 subtraction data M1 to the contents of the channel of the shift register 64 that had the maximum value (all n bits are signal 1), the register 6
The contents of 4 are subtracted by 1 in synchronization with the first decay clock DC1.

That is, all carry data above the n-th bit overflows, and subtraction is essentially performed.

. When the attack end signal AF becomes ``1'', the output of the inverter 51 becomes the signal O, so the attack curve memory 65 is prohibited from operating, and the decay curve memory 66 is enabled instead.

A decay curve as shown in the first decay portion of FIG. 9 is read out from the decay curve memory 66 in response to address data that is outputted from the shift register 64 and decreases sequentially.

Attack curve memory 65 and decay curve memory 66
The outputs of the multiplier 6 are combined by an OR gate group 52 and
Therefore, amplitude information A1(t) continuous from the attack to the first decay as shown in FIG. 9 can be obtained.

The sustain level SUL shown in FIG. 9 is set by the sustain 1 novel setter 53 at a value corresponding to the at 1/s of the I/bevel UL.

1/bell SUL and 1/jistaro 4 set with setting device 53
The comparator 1 noter 54 detects a match between the outputs (at1nos of the memory 66), and the match detection output a 1 n is sent to the 12-bit shift 1/register 5 via the OR circuit 55.
6 to be memorized.

The output of the 1 register 56 is input to the AND circuit 73 as the first decay end signal IDF, inhibits the AND circuit 72, and is held in the 1 register 56 via the AND circuit 57.

In response to the signal IDF, the feeding of the first decay clock DC1 is stopped, and the count contents of the corresponding channel in the shift 1/distaro 4 are held at a constant value.

Therefore, the readout output of the decay curve memory 66 is also constant, and as shown in FIG. 9, the sustain 1 novel SUL is maintained until the key is released.

When the key is released, the decay start signal DS is generated from the key assigner 2, so the AND circuit 73 becomes operational, and the second decay clock DC2 oscillated from the second decay clock oscillator 77 is transmitted to the AND circuit 73 and the OR circuit 74. It joins the AND gate group 78 via .

Therefore, the 1 subtraction data M1 is the second decay clock DC.
It is input to the adder 62 in synchronization with 2, and subtraction from the contents held in the 1 nodisk 64 is restarted.

Thus, the address for reading out the memory 66, which was temporarily interrupted at the sustain level SUL, is further advanced, and the decay curve of the second decay portion as shown in FIG. 9 is read out.

When the subtraction progresses and the content held in the corresponding channel of the 1-no register 64 becomes O, reading out the decay curve is completed.

When the NOR circuit 58 detects that all n bits output from the register 64 have become 110'', the end of the decay is detected.

AND circuit 5 into which the attack end signal AF is input
The reason why the output signal 1 of the NOR circuit 58 is taken out through the NOR circuit 9 and used as the decay end signal DF1 is to generate the decay end signal DF1 only after the attack is completed.

This decay end signal DF1 is supplied to the AND circuit 23 of the key assigner 2.

When the counter clear signal CC is supplied from the key assigner 2, all contents held in the relevant channel of each shift l/distal 4, 69.56 are reset to O.

Thus, a digital envelope waveform as shown in FIG. 9 is input to the multiplier 6 as time-varying amplitude information A1(t).

The manner in which the amplitude information A1(t) changes is determined by each oscillator 75 to 77.
This can be changed freely by appropriately changing the oscillation clock of , or by changing the settings of the sustain 1 novel setter 53.

Since the adder 62 and the shift 1/distal 4 configured as a counter are shared by each channel in a time-division manner, the amplitude information A1(t) is generated in a time-division manner for each channel.

Next, calculation of the frequency modulation part will be explained.

In processing sequence B, phase information qR is applied to multiplier 8.

In this multiplier 8, phase information qR is multiplied with time-varying coefficient information J(t)#:(, and phase information "J(t)qRJ" of the carrier component is calculated.

Coefficient information 7(t) is generated from carrier wave control signal generation circuit 9.

This coefficient information? According to (1), the frequency of the carrier wave is changed as appropriate.

In processing sequence C, phase information qR is applied to multiplier 10.

In the multiplier 10, the phase information qR is multiplied by the coefficient information m(t) that changes over time, and the phase information of the modulated wave component is
m(t)qRJ is calculated.

Coefficient information m(t) is generated from the modulated wave control signal generation circuit 11.

Using this coefficient information m(t), the frequency of the modulated wave is changed as appropriate.

Phase information m(t)q of the modulated wave output from the multiplier 10
R is added to the sine wave waveform storage device 12, and the amplitude value of the sine wave waveform sample point corresponding to the phase value m(t)qR at that time is read out.

The storage device 12 has a similar configuration to the sinusoidal waveform storage device 5 described above.

The modulated wave signal r output from the sine wave waveform storage device 12
sin Gn(t) qRJ is input to the multiplier 13,
Multiplied by modulation index information I(t).

The modulation index information I(t) can be changed over time, and the modulation index control signal generation circuit 14
generated from.

The value "I(t) sin(m(t)qRJ" output from the multiplier 13 is added to the adder 15, and the value r?(t)qRJ supplied from the multiplier 8
is added.

Therefore, from the adder 15, a value that determines the phase of the entire frequency-modulated signal wave, r J (t) qR + I (t) sin (m (t)
qRj is output.

The output of this adder 15 is input to a sine wave waveform storage device 16, and the amplitude value of each sample point of the sine wave waveform is arbitrarily read out.

The storage device 16 has the same configuration as the sinusoidal waveform storage devices 5 and 12 described above.

Modulated signal wave s output from the sine wave waveform storage device 16
in (, ff(t) qR+ I(t) sin (m
(t) qR) ) is added to the multiplier 17 and multiplied by maximum amplitude information A2(t) of the frequency modulated wave component.

The maximum amplitude information A2(t) is generated from the amplitude information generating circuit 18.

This generating circuit 18 can use the same configuration as the amplitude information generating circuit 7 shown in FIG. 8, and the envelope waveform corresponding to the key press and key release as shown in FIG. (t) is supplied to the multiplier 17.

Therefore, the envelopes of the fundamental wave component and the frequency modulated wave component are separately and independently supported by the amplitude information A1(t) and A2(t).

As a result of the multiplication, the multiplier 17 outputs an amplitude-controlled modulated signal wave, A2(t) sin (J(t) qR+ I(t
) sin (m(t) qR) ) is output.

By the way, as mentioned above, the frequency ratio C7m of the carrier wave and the modulated wave determines the positional relationship of harmonics in the modulated signal wave, and the modulation index ■ defines the number of harmonics, so the coefficient information? The positional relationship of harmonics is determined according to the values of (1) and m(t), and the number of harmonics is changed according to the value of modulation index information I(t).

Therefore, each information? By appropriately setting and changing (t), m(t), I(t)E, any desired tone can be obtained and complex temporal changes in tone can be easily simulated. be able to.

The signal generation circuits 9, 11.14 that generate each information 1(t), m(t), I() generate each information ?(t), m(t), The value of I(t) and its temporal change can be freely programmed.

This programming can be easily performed using operators such as switches without using complicated software.

FIG. 10 shows an example of the carrier wave control signal generation circuit 9, the modulated wave control signal generation circuit 11, or the modulation index control signal generation circuit 14.

The signal generating circuit shown in FIG. 10 has a similar configuration to the amplitude information generating circuit 7 of FIG. 8, and for a detailed explanation, please refer to the explanation section of FIG. 8.

To explain the outline of FIG. 10, first, attack start signal A
When S is supplied from the key assigner 2, the attack clock pulse AP from the variable clock oscillator 85 is sent to the AND gate group 8 via the AND circuit 81 and the OR circuit 84.
Open gate 9.

The n-bit 1 addition data P1 is output from the gate group 89 in synchronization with the attack clock AP, and is applied to the n-bit adder 91 via the OR gate group 90.

An adder 91, an AND gate group 92, and a cyclic shift 1/register 93 of 12 words (1 word - n bits) constitute a counter, which is time-divisionally shared by 12 channels in accordance with the main clock φ1.

In this way, 1 is added sequentially according to the attack clock AP, and the integration result is accumulated in the 1-no register 93.

The output of the shift 1/registor 93 is coefficient information 1(t) or m(t), or modulation index information ■(t).
from each generating circuit 9 or 11 or 14 to each multiplier 8
or 10 or 13.

Therefore, each piece of information z(tL m(t), I(t), that is, the output of the 1/jister 93 is typically a value that increases sequentially from the beginning of the key press to the attack portion as shown in FIG. be.

The output of the shift 1 register 93 is input to a comparator 94 and compared with the attack 1 novel ATL preset by the attack level setter 95.

When they match, the output of the comparator 1 notor 94 becomes signal 1, which is stored in the 12-bit circular shift 1/register 96.
It is stored and held via an AND circuit 97 and an OR circuit 98.

The output of the register 96 serves as an attack end signal AP' to enable the AND circuit 82, and the AND circuit 81
inhibit.

Further, the attack end signal AF' closes the AND gate group 89 via the inverter 99, and the AND gate group 89 is closed.
3 to be ready for operation.

Therefore, the gate group 88 is opened in synchronization with the first decay clock pulse DP1 from the variable clock oscillator 86.
The 1-subtraction data M, in which all n bits are signals 1, is input to the adder 91.

In response to the input of the 1-subtraction data M1, the accumulated value stored in the register 93 of the channel is sequentially subtracted by 1, and the information/(1) is obtained as shown in the first decay portion of FIG.

m(tL I(t)#(decreases.

The output of the shift register 93 is also applied to the comparator 1/-31 and compared with the sustain level SUL/ set in advance by the sustain 1 novel setter 32.

When a match occurs, the 12-bit circular shift 1/register 33
The signal 1 is stored in the corresponding channel of the AND circuit 34.
, is held via the OR circuit 35.

The output of the register 33 is the first decay end signal I DF/
It is added to the AND circuit 83 as a signal and inhibits the AND circuit 82.

As a result, the clock transmission is temporarily stopped, and as shown in FIG. 11, the output of the shift 17 register 93 (information J (
t).

m(t), I(t)) is a constant sustain level SU
Continue L'.

When the decay start signal DS is generated from the key assigner 2, the AND circuit 83 becomes operational, and the second decay clock DP2 from the variable clock oscillator 87 is applied to the gate group 88.

Therefore, the accumulated value stored in the 1-noister 93 is the clock DP2.
is sequentially subtracted by 1 according to
.

Obtain I(t).

The sound generation of the relevant channel is completed, and the counter clear signal C
When C is generated, the stored contents of the corresponding channel of No. 1 register 93, 96, 33 are cleared.

In each signal generation circuit 9, 11.14, each clock A
P, DPl, DP2, each 17 bells ATL, SUL' can be changed separately, so each information 7 (tL
m(t) and I(t) can be freely programmed.

Since the sustain level SUL' takes a constant value, the carrier wave, modulation wave, or modulation index does not change and remains constant.

Therefore, a steady tone is achieved.

However, during attack or decay, the tone changes in a complex manner.

As a result, it is possible to obtain a timbre effect similar to a change in the complex harmonic components of a natural musical instrument sound at the beginning or attenuation of the sound.

Each signal generating circuit 9, 11, 14 is not limited to the example shown in FIG. 10, but is configured to generate various types of information so as to imitate temporal changes in the frequency spectrum of various natural musical instrument sounds. (1).

It may be configured such that changes in m(t) and I(t) are stored in advance in a memory and read out in response to key presses and key releases.

Now, the fundamental wave component signal output from the multiplier 6, A1(t)sin qR and the frequency modulated signal wave output from the multiplier 17. are input to the adder 43 and added.

All calculations in each processing sequence A, B, and C are performed digitally and time-divisionally for each channel time.

Therefore, the adder 43 outputs the waveform amplitude value of the musical tone signal e(t) at the relevant time point, which is the calculation result of the equation (8), as a digital signal.

Therefore, the output of the adder 43 is converted to the digital-to-analog converter 4.
In addition to 4, it is converted into an analog amplitude value.

Since the analog musical tone signal e(t) of the tone assigned to each channel is output from the digital-to-analog converter 44 in a time-divisional manner, these outputs are converted into an analog game.
The tone signals are supplied to tone circuits 45, 46 and 47, respectively, and musical tone signals are distributed according to the type of keyboard.

The decoder 229 (see FIG. 6) of the key assigner 2 generates signals UE, LE, and PE indicating to which keyboard the tone assigned to each channel belongs in synchronization with the channel time.

Moreover, the upper keyboard signal UE is applied to the gate circuit 45, and the gate circuit 45
is opened to conduct the musical tone signal e(t) from the converter 44.

Similarly, the lower keyboard signal LE is applied to the gate circuit 46 to make only the musical tone signal of the lower keyboard conductive.

Also, by adding the pedal keyboard signal PE to the gate circuit 47,
Only the musical tone signal of the pedal keyboard is conducted.

The musical tone signals for each keyboard outputted from each of the gate circuits 45 to 47 are individually volume-controlled by variable resistors VR1t, VH2t, and VH2.

Thereafter, the upper keyboard sound and the lower keyboard sound are balanced and volume-adjusted by a variable resistor VR4, and mixed with the pedal keyboard sound via resistors R1t and R2.

In this way, the musical tone signal whose volume is adjusted for each keyboard is output from the speaker 49 via the audio device 48.

FIG. 12 is a block diagram showing another embodiment of the electronic musical instrument of the present invention.

The device shown in FIG. 5 above is an electronic musical instrument constructed based on the basic frequency modulation method shown in equation (1) above. If you configure an electronic musical instrument using a complex frequency modulation method,
It becomes possible to obtain more complex timbre changes.

In the electronic musical instrument shown in FIG. 12, musical tones are formed using the frequency modulation method of equation (5), and the musical tone signal is obtained as follows.

e(t) follows the following equation (9). Equation (9) is the frequency modulation term that is substantially equivalent to the above equation (5), and is the sum of the fundamental wave component term. It is.

In equation (9), the value qR is the phase of the fundamental wave, and the value A1(t) is the maximum amplitude of the fundamental wave component expressed as a function of time t.

Comparing Equations (5) and (9), we find that the carrier wave phase α
t is given by 1(t)qR and is obtained by multiplying the phase qR of the fundamental wave by a time function 2(1).

The phase β1t of the first modulated wave is given by the value m(t)qR, and is obtained by multiplying the phase qR of the fundamental wave by a time function m(t)G.

The phase β2t of the second modulated wave is given by the value n(t)qR, and is obtained by multiplying the phase qR of the fundamental wave by the time function n(t).

The first modulation index ■1 is expressed by a time function 11 (t), and the second modulation index ■2 is expressed by a time function ■2 (t), and each can be changed over time. It looks like this.

The value A2(t) is the maximum amplitude of the frequency modulated signal and is expressed as a function of time t, so that the amplitude changes over time.

The electronic musical instrument shown in FIG. 12 can use almost the same configuration as the electronic musical instrument shown in FIG. 5, and only requires adding some necessary circuits to the device shown in FIG.

Therefore, for convenience of explanation, circuits having the same configuration as those in FIG. 5 are given the same reference numerals, and detailed explanation will be omitted.

However, it goes without saying that the devices shown in FIG. 5 and FIG. 12 are separate devices.

As described above, the key assigner 2, the frequency information storage device 3, and the counter 4 operate in response to the suppression of keys on the keyboard 1, and time-divisionally generate phase information qR of the sound assigned to each channel.

This phase information qR is supplied to processing series A, B, C, and D, respectively.

Processing sequence A calculates the fundamental wave component A1 sin qR, as in the embodiment shown in FIG.

Frequency modulation calculations are performed in processing sequences B, C, and D.

The difference from the electronic musical instrument shown in FIG. 5 is that a processing sequence is added.

In the processing sequence, a modulated wave control signal generation circuit 11
The coefficient information n(t) generated from 0 and the phase information qR are multiplied by a multiplier 100, and the output of this multiplier 100 is n(t).
The second modulated wave signal sin (n(t) qR) is read from the sinusoidal waveform storage device 120 in accordance with qR.

The second signal generated from the modulation index control signal generation circuit 140
modulation index information 2(t) and the signal s of the second modulated wave
in (n(t) q R) is multiplied by multiplier 130 and the signal 12(t) sin (n(t) qR) is provided to adder 150.

The detailed circuit configuration of each of the circuits 100 to 140 in the processing series can be the same as the configuration of each of the circuits 10 to 14 in the processing series C.

In the processing sequence C in FIG.
)qR) is output.

The adder 150 receives phase information of the carrier wave output from the multiplier 8. (t) qR, output of multiplier 13, multiplier 130
It is designed to add up the output of , and the power of three people.

According to the output of this adder 150, the sine wave waveform storage device 1
6 is read out and multiplied by the amplitude information A2(t) in the multiplier 17 to obtain a frequency modulated signal.

This frequency-modulated signal is converted to the fundamental wave component signal A1(t)sin output from the multiplier 6 by the adder 43.
qR to obtain a musical tone signal e(t) which is the calculation result of the preceding equation (9).

This musical tone signal e(t) is processed through circuits 44 to 48 in the same manner as described above, and is output from a speaker 49.

Harmonic Limiting By the way, when forming a frequency signal by sampling, as known from the sampling theorem,
Harmonic components higher than 100 frequencies of the sampling frequency are folded back into the audible range, resulting in subharmonics.

To prevent this, signal components higher than 100 sampling frequencies must be removed.

In the case of the above embodiment, the frequency of the main clock φ1 is IMHz
Since the waveform is created by 12-tone time division, the sampling frequency of one waveform is as follows.

Therefore, it is necessary to remove signals above about 40 kHz.

Generally, the frequency bandwidth BW by the frequency modulation method is and ■ Since d/m, It is known that

Here, BW is the entire bandwidth, but since only the high band side needs to be considered, the bandwidth on one side BWp is given by:

m is the modulation wave frequency, and ■ is the modulation index.

Therefore, in a frequency modulated signal, the highest frequency among the frequency components with substantial amplitude is:

C is the frequency of the carrier wave. This frequency is 40 kHz
If it is smaller, subharmonics will not occur, so the basic condition for harmonic limiting is .

Carrier wave C and limit frequency 4 several intervals, Frequency between 0kHz The maximum value M of the number of side waves included in is, It is.

Therefore, if the bandwidth BWp on the high frequency side is smaller than the value Mm, subharmonics will not occur.

Therefore, the basic conditions of equation 00) can be simplified as follows.

becomes.

Therefore, if the modulation index (2) is determined within a range that satisfies the relationship in equation 11), subharmonics can be prevented.

In the embodiment of FIG. 5 or FIG. 12, the network 11
) A special harmonic limit circuit (not shown) can be provided to determine the equation.

In this harmonic limit circuit, frequency information R read out from the frequency information storage device 3 and coefficient information 1(t), m(t) generated from each signal generation circuit 9, 11, 110, 14°140.

Based on n(t), I(t), 11(t), and 12(t), determine the frequency of the carrier wave C1 modulated wave m, calculate the value M, and determine if the equation (11) is satisfied. Determine whether or not.

If equation (11) is not satisfied, modulation index information I(
tL ■1 (t), ■2 (Limit the tth value to a small value,
1) Appropriate measures may be taken such as changing the equation so that the equation is satisfied.

Note that the present invention is not limited to the above-mentioned embodiments, and musical tone signals can also be generated using other complicated frequency modulation methods (for example, the above-mentioned equations (6) and (7)).

Such various modifications can be realized by simply slightly modifying the circuit shown in FIG. 5 and adding necessary calculation processing sequences to it.

Furthermore, if a waveform storage device containing many harmonic components such as a sawtooth wave, triangular wave, or rectangular wave is used instead of the sine wave waveform storage devices 5, 12, and 120, harmonic waveforms can be used as the fundamental wave component or modulation wave component. A waveform containing many waves can be used, and a musical tone containing more complex harmonic components can be obtained.

As described above, according to the present invention, in an electronic musical instrument using a frequency modulation method, it is possible to synthesize a plurality of tones in real time in accordance with the suppression of a plurality of keys and simultaneously produce them.

[Brief explanation of drawings]

Figure 1 is a graph illustrating the curve of the Bessel function, Figure 2 is a graph showing the side wave spectrum when the modulation index is I-4, Figure 3 a''-c is a graph explaining side wave reflection, Fig. 4 is a graph showing an example of a sidewave spectrum due to a complicated frequency modulation method, Fig. 5 is a block diagram showing an embodiment of the electronic musical instrument of the present invention, and Fig. 6 is a diagram showing the usurpation of the key assigner in the same embodiment. Illustrative block diagram, Figure 7 a, b
is a graph showing the timing of the main clock and each channel time used in the same embodiment, FIG. 8 is a block diagram showing an example of an amplitude information generating circuit in the same embodiment, and FIG. 9 is a graph showing the timing of the amplitude information generation circuit used in the same embodiment. FIG. 10 is a graph showing a typical envelope of amplitude information, FIG. 10 is a block diagram showing an example of various control signal generation circuits in the same embodiment, and FIG.
FIG. 12 is a graph showing typical envelopes of control signals generated in FIG. 12, and FIG. 12 is a block diagram showing another embodiment of the present invention. 1...keyboard, 2...key assigner, 3
...Frequency information storage device, 4...Counter, A, B, C. D... Processing series, 5, 12, 16, 120...
... Sine wave waveform storage device, 7, 18 ... Amplitude information generation circuit, 9, 11, 14, 110, 140 ...
...Control signal generation circuit.

Claims (1)

[Claims] 1. A first means for detecting a pressed key and assigning the sound of the key to one of a specific number of channels; a second means for generating phase information whose value changes sequentially at a corresponding speed;
A carrier wave phase component and a modulated wave phase component are respectively formed for each channel based on the phase information of each channel generated from the means, and a signal wave frequency modulated based on these phase components and an arbitrary modulation index is generated for each channel. Create,
and a third means for transmitting this signal wave as a musical tone signal, the electronic musical instrument being capable of simultaneously generating a plurality of tones in response to the musical tone signal transmitted from the third means according to the previous assignment. .