JPH10198369A - Electronic musical instrument - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simulate a prior tone wheel organ with a small instrument as accurately as possible, by providing a musical sound generator which newly makes some voice other than a voice which is contributing to output of sound to generate unit sound, and generates and outputs sound required to be generated. SOLUTION: A unit sound determining component 4, according to key-number and pitch of each structural sound which is specified by a structural-sound pitch specifying component 2, determines unit sound corresponding to each of structural sounds which form sound required to be generated. A sound generator 5 increases voice level for unit sound for which generation responsible voice exists of plural unit sounds determined by the unit sound determining component 4, and for unit sound for which no generation responsible voice exists, newly makes one voice other than some voice which is contributing to output of sound assembled from plural unit sound to generate unit sound, and generates and outputs sound required to be generated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to an electronic musical instrument,
In particular, the present invention relates to an electronic musical instrument that simulates the tone of a so-called tone wheel organ.

[0002]

2. Description of the Related Art Conventionally, there has been an electric musical instrument called a so-called tone wheel organ. This tone wheel organ includes, for example, a gear-shaped rotating body (referred to as a wheel), and a sensor for picking up the approaching teeth of the gears of the wheel. The sensor detects the pitch and the rotation speed of the gears of the wheel. A large number of (for example, 91) elements for obtaining a periodic signal corresponding to each chromatic scale are provided so as to obtain a periodic signal corresponding to each chromatic scale, and further, a manipulator of a variable operation amount called a draw bar for adjusting the level of each harmonic. (For example, nine), and each of the drawbars is operated by a desired amount, and when a key on the keyboard is depressed, the drawbar corresponds to each of the depressed keys (key number). Wheels, for example 9 total
An electric signal with a structure that emits the organ sound corresponding to the key depression by amplifying and mixing the periodic signals obtained by the selected wheels to the level corresponding to the operation amount of the corresponding draw bar, and mixing. Musical instrument. Wheels correspond to pipes in classic pipe organs.

On the other hand, in recent years, an electronic musical instrument called a synthesizer that can generate a musical tone having an extremely complicated overtone structure has appeared. In this synthesizer, there is a structure of a concept called a large number (for example, 128) of voices, which is responsible for generating one arbitrarily assigned sound, which is hardware, software, or a combination thereof, and the voice has a concept. As many different musical tones can be produced simultaneously.

[0004]

Here, consider the generation of a musical tone simulating the tone color of a conventional tone wheel organ using a recent synthesizer. Naturally, it is possible to use one voice to generate a tone of one tone having a harmonic structure corresponding to the operation amount of a plurality of (for example, nine) drawbars, that is, a tone of a tone wheel organ tone. is there.

However, if the tone wheel organ is simulated in a simple manner usually used when using a synthesizer, the following problems occur. That is, in the case of a tone wheel organ, for example, when two musical tones corresponding to two keys include harmonics of a common frequency,
One periodic signal obtained by one wheel is used for generation of harmonics of a common frequency of the two tones. That is, the overtones of the same frequency included in the plurality of tones are sounds having the same phase regardless of the timing of the key depression of the plurality of keys corresponding to the plurality of tones. On the other hand, if a synthesizer generates a plurality of tones of the tone wheel organ tone, each tone is generated for each voice at each timing when a plurality of keys are depressed, so that the plurality of tones have the same frequency. When overtones are included, the phases of the overtones of the same frequency are shifted according to a slight shift of the key pressing timing, and the overtones are weakened or weakened according to the shift of the phase. That is, when the amplitudes of the overtones are equal to each other and the phases are shifted by 180 °, the overtone disappears. In other words, even if the player presses the same key every time at the same timing during the operation of the player, a slight timing shift occurs every time because the person plays the music, and the amplitude of the harmonic changes each time. However, even if the key is pressed in the same manner each time, the timbre changes each time, and there is a problem that the timbre cannot be controlled to a constant timbre.

In order to cope with such a problem, a configuration is also conceivable in which the same number of voices as the wheels of the tone wheel organ are provided, and the wheels are simply replaced with voices. There is a problem that the same number of voices are required and the configuration becomes large. Furthermore, in the case of the above-described configuration, each voice simply generates a corresponding periodic signal of the wheel, so that the tone other than the tone wheel organ tone despite the large-scale configuration Cannot be generated.

SUMMARY OF THE INVENTION In view of the above circumstances, an object of the present invention is to provide an electronic musical instrument capable of highly simulating a conventional tone wheel organ with a small configuration. Another object of the present invention is to provide an electronic musical instrument capable of generating musical tones of a tone other than the tone wheel organ tone.

[0008]

FIG. 1 is a diagram illustrating the principle of an electronic musical instrument according to the present invention. The electronic musical instrument of the present invention includes a plurality of voices 1_1, 1_2,..., 1_n, constituent pitch designation means 2, unit sound assignment means 3, unit sound determination means 4, and musical sound generation means 5.

A plurality of voices 1_1, 1_2,..., 1_n
Are responsible for generating one arbitrarily assigned sound. The constituent pitch designation means 2 designates the pitch of each constituent sound of a musical tone composed of a set of a plurality of constituent sounds each having a different pitch. Further, the unit sound assigning means 3 causes a plurality of voices to be in charge of generating unit sounds of different pitches before generating the musical sound, and also controls the level of the generated unit sound to prevent the output of the unit sound. This is to be reduced to a state in which

Further, the unit tone determining means 4 receives a tone generation instruction of a musical tone composed of a set of a plurality of constituent sounds, and designates a key number designated by the tone generation instruction and a sound of each constituent sound designated by the constituent pitch designation means 2. In accordance with the pitch, a unit sound corresponding to each of a plurality of constituent tones constituting the musical tone for which the pronunciation is instructed is determined. Further, the musical sound generating means 5 increases the level of the unit sound including the voice in charge of generation among the plurality of unit sounds determined by the unit sound determination means 4 and is in charge of the generation. For a unit tone that does not have a voice that is present, a voice other than the middle voice that is contributing to the output of a tone consisting of a plurality of unit sounds is newly assigned to generate the unit sound, and a pronunciation instruction is issued. This is to generate and output a musical tone with a certain sound.

In the electronic musical instrument according to the present invention, each unit sound (corresponding to each periodic signal in the tone wheel organ) is generated by a plurality of voices, but the level is lowered so as not to output the unit sound. State, and when a sounding instruction is given, according to the key number of the sounding instruction and the pitch of each constituent sound specified by the constituent pitch specifying means, the level of the unit sound being generated by which voice Is determined, and by adjusting and outputting as described above, a musical tone corresponding to the pronunciation instruction is obtained.There are a plurality of tone generation instructions, and the plurality of musical tones have the same frequency. When overtones are included, one unit sound generated by one voice is used for the same overtone, and therefore, the phases of overtones of the same frequency are always the same regardless of the timing shift of the pronunciation instruction. Always it is possible to output a tone of a certain tone.

Here, the unit sound allocating means 3 determines that there is no voice that is in charge of generating any one of the plurality of unit sounds determined by the unit sound determining means 4. The generation of the unit sound is assigned to any one of the voices except for the middle voice that is contributing to the output of a musical tone composed of a plurality of unit sounds. Basically, each voice generates each unit sound and keeps it idle, but for example, when the number of voices that can be used for generating the unit sound is smaller than the total number of unit sounds, there is a sounding instruction. The above situation occurs when it is necessary to use a unit sound that is not in an idle state for generating a tone. Therefore, in the electronic musical instrument of the present invention, a musical tone can be generated with less voices than the number of unit sounds to be generated, and a small-sized electronic musical instrument can be realized.

Here, the musical sound generating means 5 determines the level of the unit sound of the unit sound for which a voice responsible for generation exists among the plurality of unit sounds determined by the unit sound determining means 4, and outputs the unit sound level. By generating a tone for which a pronunciation is instructed by increasing the level by the level of the component tone corresponding to the unit tone that composes the given tone, and outputting the tone, the tone of the tone being pronounced consisting of a set of a plurality of component tones is generated. When a pronunciation stop instruction is issued, the level of a unit sound generated by a plurality of voices that are responsible for generating a plurality of constituent sounds that make up the tone being sounded is changed to a plurality of units that make up the tone that is sounding. It is preferable that the tone generation is stopped by reducing the tone by the level of the constituent sound.

In this case, the constituent pitch designation means 2 designates the level of each constituent sound in addition to the pitch of each constituent sound, and the musical tone generating means 5 has a sounding instruction. The level of the unit sound constituting the musical tone is increased by the level of the constituent sound corresponding to the sounding instruction specified by the constituent pitch specifying means 2 and the unit sound constituting the musical tone for which the sounding stop instruction has been issued. Is more preferably reduced by the level of the constituent sound corresponding to the sound generation stop instruction specified by the constituent pitch designation means 2.

In this case, the level of each constituent sound can be specified, and the level of each unit sound is increased by the level of the constituent sound corresponding to the sounding instruction and reduced by the level of the structure corresponding to the sounding stop instruction. , A constituent sound of the specified level is generated. Further, in the electronic musical instrument of the present invention, among the unit sounds corresponding to the constituent sounds of the musical tone for which the pronunciation is instructed, the level of the unit sound in which the voice in charge of generation exists is increased to increase one voice. A tone generation instruction of a second form, which is different from the tone of the first form in which a voice is shared by a plurality of pronunciation instructions and uses one voice alone in one pronunciation instruction, is received, The second musical sound generating means 6 for sharing the generation of the musical sound of the second form in which the pronunciation is instructed to any of the voices other than the voice in the first state which is in the middle of the output of the musical sound. Preferably, it is provided.

By providing such a structure, a faithful simulation of the tone wheel organ can be performed, and
The original synthesizer function of generating one musical tone with one voice can be used at the same time. By using the original usage as a synthesizer together, you can play the tone wheel organ with the orchestra backed by the concert with the tone wheel organ and other instruments, or at the start of sounding with the tone wheel organ, By generating a noise sound generated at the end of sounding, the tone wheel organ can be more realistically simulated.

Here, in the case where the second tone generating means 6 is provided, if the second tone generating means 6 has a voice in the second state which is not currently in charge of sound generation, In the second state, when any one of the voices in the second state is in charge of generating a tone in the second form in which the voice is instructed, and there is no voice in the second state, the unit sound assigning means 3 The level of the generated unit sound is reduced to a state where the output of the unit sound is blocked, excluding the voice in the first state among the voices to which the generation of the unit sound is assigned. Of the voices in the second state and the voices in the third state. Second form when no voice exists For any one of the voices in the fourth state in charge of generating the musical tone, the generation of the musical tone of the second form being generated by the voice is stopped, and the voice generation instruction of the second form is stopped. It is preferable to be in charge of generating the second form of musical sound.

By allocating the usage of the voices in this manner, the generation of the first form of musical sound (for example, a musical sound as a tone wheel organ) and the second form of the musical sound (for example, a noise sound or a musical sound as a back orchestra) ), And effective use of voice is achieved. Further, in this case, when the generation of the tone of the second form in the voice in charge of the generation of the tone of the second form is completed, the unit sound assigning means 3 assigns the voice in charge of the generation of the tone of the second form. If there is a unit tone that has not been assigned, generation of one of the unit tones to which the voice responsible for generation is not assigned to the voice for which generation of the tone of the second form has been completed. It is preferable to be responsible for

By doing so, it is possible to put as many unit sounds as possible in the idling state.
As a whole, the response to the pronunciation instruction becomes faster. In this case, when there is no voice that is in charge of generating any one of the plurality of unit sounds determined by the unit sound determining means 4, the musical sound generating means 5 The level of the generated unit sound is reduced to a state where the output of the unit sound is blocked, excluding the voice in the first state among the voices to which the generation of the unit sound is assigned by the means 3. When there is a voice in the third state, one of the voices in the third state is in charge of generating a plurality of unit sounds determined by the unit sound determining means 4. When the voice in the third state is not present, the voice in the fourth state is in charge of generating the tone in the second form. The voice to one of the voices The generation of the musical tone of the second mode being generated is stopped, and the generation of the unit sound for which there is no voice in charge of generation among the plurality of unit sounds determined by the unit sound determination means 4 is performed. It is preferable to be responsible for

By doing so, it is possible to maximize the effective use of the voice while giving priority to the tone generation of the first mode (for example, the tone of the tone wheel organ).

[0021]

Embodiments of the present invention will be described below. FIG. 2 is a block diagram illustrating a hardware configuration of the electronic musical instrument according to the embodiment of the present invention. The electronic musical instrument 10 shown in FIG.
3, a keyboard 14, a panel switch 15, and a tone generator 16 are provided, and are connected to each other via a bus 17. A sound system 20 is connected to the tone generator 16.

The CPU 11 is a central processing unit that reads out and executes a program stored in the ROM 12, and controls the entire electronic musical instrument 10.
The ROM 12 is a read-only memory.
2 stores programs executed by the CPU 11 and various constants that do not need to be changed. RAM 13
It is a readable and writable memory, used as a work area when the program is executed by the CPU 11, or
It is used as a storage area for various parameters and the like set by a panel switch 15 described later.

The keyboard 14 has a number of performance operators (also referred to as keys) designed similarly to a number of keys arranged on an acoustic piano or organ keyboard. It is configured so that a performance operation can be performed with a feeling similar to that of the above. Each key is assigned a key number (key number) whose value increases one by one from the bass side to the treble side.

The panel switch 15 is usually composed of a plurality of switches arranged on the same panel as keys constituting a keyboard, an operation element with a variable operation amount (this is called a draw bar, and details will be described later), and the like. This electronic musical instrument 10
Is used to set various parameters and the like. The musical sound generating section 16 is a section that generates various musical sounds (including noise sounds and the like) based on instructions from the CPU 11, and the generated musical sounds are input to a sound system 20 including an amplifier and a speaker. Sound is emitted into space. Here, not only the musical sound emitted into the space but also the state of the electric signal before the sound is emitted into the space is referred to as a musical sound.

FIG. 3 is a panel diagram showing a part of the keyboard 14 and panel switch 15 shown by blocks in FIG.
Here, a part arranged on the leftmost side of a large number of white keys 141 and black keys 142 constituting the keyboard 14 and a part of the panel switch 15 shown by a block in FIG. I have. Note that the panel switch 15 as a whole also includes a number of switches and operators in addition to those shown in FIG. 3, but only those related to the characteristic portions of the present invention are shown here. ing.

FIG. 3 shows nine drawbars 151.
, Six synth tone color switches 152, and six synth tone color selection LEDs 153 are shown. In the present embodiment, the performance of the tone wheel organ tone (hereinafter referred to as the organ tone and the tone of the organ tone as the organ tone) by the player using the keyboard and the synth tone selection button 152 By pressing any one of these keys, the automatic performance is performed simultaneously with the automatic performance of the selected tone (hereinafter referred to as a synth tone, and the tone of the synth tone is referred to as a synth tone).

In the present embodiment, performance data corresponding to key depression and key release for a keyboard for automatic performance is created in advance and stored in the ROM 12 or the RAM 13.
The nine drawbars 151 are used to determine the organ tone, and can be lowered to an arbitrary position toward the lower side of FIG. 3, and the larger the amount of reduction, the higher the harmonic of the drawbar 151 corresponding to the lowered drawbar 151. Level increases. Each of the drawbars 151 has "16", "5 and 1/3", "8", "4",
“2 and 2/3”, “2”, “1 and 3/5”, “1 and 1 /
3 "and" 1 "are displayed. This is a remnant of a classic pipe organ and is a number that represents the length of the pipe of the pipe organ.

When any one of the keys constituting the keyboard 14 is pressed, a musical tone having a pitch corresponding to each drawbar is simultaneously generated. That is, each drawbar sets the pitch of the corresponding overtone fixedly and sets the level of the overtone arbitrarily. The foot-displayed value of each drawbar 151 is inversely related to the frequency of the harmonic corresponding to that drawbar, for example, a 16-foot drawbar,
The four-foot drawbar corresponds to a pitch of a half frequency and a double frequency of the pitch of the eight-foot drawbar, respectively. The 8-foot drawbar corresponds to the fundamental tone.

As described above, each overtone is assigned to each drawbar 151, and the draw-out amount of each drawbar 151 indicates the level of the corresponding overtone. Thus, in this electronic musical instrument, nine drawbars 15 are used.
The organ tone color is adjusted by adjusting the amount of each drawer. In this electronic musical instrument, in addition to an organ tone that is adjusted by pulling out nine drawbars 151, six synth tones corresponding to six synth tone switches 152 are prepared, and six synth tone colors are prepared. By pressing one of the synth tone switches of the switch 152, the synth tone corresponding to the pressed synth tone switch is selected. The six synth tone LEDs 153 are connected to the six synth tone switches 1.
When one of the synth tone switches is pressed and a synth tone corresponding to the selected synth tone switch is selected, a synth tone LED corresponding to the depressed synth tone switch is arranged.
Lights up so that it is possible to visually recognize which synth tone is currently selected.

FIG. 4 is a conceptual block diagram of the tone generator 16 shown by the block in FIG. Here, there are a total of 128 voices named voice 0 to voice 127. Each voice is responsible for generating one sound, and each voice includes a wave generator WG, a time variant filter TVF, and a time variant amplifier TVA.

The WG generates a waveform that is the basis of a certain sound, the TVF filters the waveform to adjust harmonic components, and the TVA adjusts the amplitude of the waveform.
Here, WG is capable of generating a specified waveform at a specified pitch, and when generation of an organ sound is specified, a wheel corresponding to the specified pitch of the tone wheel organ to be simulated. Generate the waveform generated by
When generation of a synth tone is designated, a waveform corresponding to the synth tone selected by the synth tone switch 152 is generated at the designated pitch. Further, TVF and TVA can adjust their cutoff frequency and amplification factor over time even during generation of one sound, so that the timbre and amplitude can be temporally enveloped. Is given. This means that, for example, the tone and amplitude of a musical tone of a piano change as time elapses after a key is pressed, and a TVF or TVA is required to simulate such a musical tone that changes with time.

The sound generated by each voice is transmitted to the mixer MI.
They are mixed with each other at X and sent to the sound system 20 (see FIG. 2). Hereinafter, the program executed by the CPU 11 shown in FIG. 2 will be described. FIG. 5 is a flowchart of the initialization routine. When power is supplied to the electronic musical instrument, or when initialization is instructed by an operator (not shown), the initialization routine is executed.

First, a variable V is set to V = 0 (step 5).
— 1), V is substituted for the wheel voice [V] (Step 5_2). Here, the wheel voice [V] is V = 0.
And a register array for storing voice numbers of voices that generate a waveform corresponding to a periodic signal generated by a V-th wheel (referred to as a wheel V) of the tone wheel organ, counting the 0th wheel as a 0th wheel. Wheel voice [V] = V indicates that the numerical value V is substituted into the V-th register (wheel voice [V]) in the register array. Here, a configuration corresponding to a wheel of a tone wheel organ is conceptually referred to as a “wheel”. That is, the wheel voice [V] = V means that the periodic signal of the wheel V is generated by the V-th voice (referred to as voice V).

In step 5_2, not only is V substituted for the wheel voice [V], but also an operation of causing the voice V to generate a periodic signal of the wheel V is performed. However, here, the TVA of the voice V (see FIG. 4) has an amplification factor of 0 so that the periodic signal generated by the voice V is not output to the mixer MIX. This state is called that the wheel voice [V] is in an idle state.

In step 5_3, 0 is substituted into the wheel reference counter [V]. The wheel reference counter [V] is an array of registers provided corresponding to the wheel voice [V].
One of the numerical values 2, 3, ... is stored. The wheel reference counter [V] indicates the corresponding wheel voice [V].
(Hereinafter referred to as "voice designated by wheel voice [V]") having a voice number (here, V) stored in the "" Specifically, the wheel reference counter [V] = 0 corresponds to the wheel voice [V].
Indicates that the voice specified by the wheel reference counter is in an idle state, and the wheel reference counter [V] = 1 is a certain 1
One of a plurality of periodic signals forming an organ sound corresponding to one depressed key is a periodic signal generated by a voice designated by a wheel voice [V], and refers to a wheel reference. The counter [V] = 2 indicates that the periodic signal generated by the voice designated by the wheel voice [V] is used in both of the two organ sounds corresponding to the two depressed keys. , Wheel reference counter [V] = 3, 4,... Further, the wheel reference counter [V] =-1 is the wheel voice [V].
Indicates that the voice specified by the does not generate a periodic signal corresponding to the wheel. This state occurs when the voice is used for generating a synth sound or when generating a periodic signal corresponding to another wheel, as described later.

In step 5_3, 0 is set to the wheel reference counter [V]. That is, it is set that the voice specified by the wheel voice [V] is in an idle state as an initial state. In step 5_4, V
<90, and if V <90, step 5
At _5, V is incremented, and the process returns to step 5_2. That is, in this electronic musical instrument, 91 wheels are conceived, and periodic signals of wheels 0 to 90 are generated by voices designated by wheel voices [0] to [90], respectively. Not meant to be idle.

If it is determined in step 5_4 that V ≧ 90, the process proceeds to step 5_6, and 0 is set in the number of loans.
As described above, the periodic signal of each wheel is generated by 91 voices corresponding to each of the 91 wheels for the organ sound. Since this electronic musical instrument has 128 voices (voice 0 to voice 127), even if a periodic signal of each wheel is generated by 91 voices corresponding to 91 wheels, 128−91 = 37. Although there is room for more voices, the number of simultaneous sounds of synth sounds increases,
It is expected that this will be exceeded. At that time, as described later, the generation of the periodic signal in the voice in the idle state for generating the organ sound is stopped, and the synth sound is generated in the voice. In other words, this is considered as “renting” voices from organ sounds to synth sounds, and
Indicates the number of voices lent from the organ sound to the synth sound. Setting the number of lending to 0 in step 5_6 means that the number of lending from the organ sound to the synth sound is 0 in the initial state.

In step 5_7, 91 is set to the number of idle wheels. The number of idle wheels is the number of voices in the idle state, and is 9 in the initial state.
All 91 voices corresponding to one wheel are idle. The number of idle wheels is smaller than 91 when the organ sound is generated and output, and when a voice is lent from the organ sound to the synth sound to generate the synth sound.

In step 5_8, the variable V is set to an initial value 91
And connect V to the free link (step 5_
9), it is determined whether or not V <127 (step 5_1)
0), if V <127, V is incremented in step 5_11 and the process returns to step 5_9. Free links are not used to sound organ sounds or synth sounds, and are not idle for organ sound sounds. A data structure in which the voice numbers of the existing voices are connected to each other.
The voice numbers of a total of 37 voices from voice 91 to voice 127 are connected to the free link, excluding the voices in an idle state for sounding a total of 91 organ sounds.

FIG. 6 is a flowchart of an organ key pressing routine that is executed in response to a sound generation instruction for generating an organ sound. In the present embodiment, the sounding instruction for generating the organ sound is generated by depressing each key of the keyboard 14 shown in FIGS. First, at step 6_1, an initial value 0 is substituted for a variable drb. In step 6_2, it is determined whether or not the draw amount of the drawbar No. drb among the nine drawbars 151 (see FIG. 3) is 0. When the draw amount of the drawbar No. drb is 0, the drawbar related to the drawbar is determined. Since processing is unnecessary, the process directly proceeds to step 6_9, and d
It is determined whether or not rb <8, that is, whether or not the processing has been completed for all of the nine drawbars. If unprocessed drawbars remain, the process proceeds to step 6_10 and d is performed.
rb is incremented, and the process returns to step 6_2.

If it is determined in step 6_2 that the drawbar of the drb number is slightly lowered (the amount of drawing is not 0), the process proceeds to step 6_3, and the keyboard 14
According to the key (key) pressed this time and the variable drb,
One pitch (wheel) for generating an organ sound corresponding to the key pressed this time is determined and stored in a variable W.

That is, the sum of the foot offset [drb] obtained using the variable drb as an argument, the key number of the pressed key, and a predetermined constant is stored in the variable W as a wheel number. The foot offset [] is an array in which numerical values indicating how many pitches the pitch corresponding to each drawbar is based on an 8-foot drawbar are fixedly stored in advance, and are “16”, “5 and 1/3”. "," 8 ",
"4", "2 and 2/3", "2", "1 and 3/5",
-12,7,0,12,19,2 corresponding to the draw bar of each foot display of "1 and 1/3" and "1"
Numerical values of 4, 28, 31, and 36 are stored.

In step 6_4, it is determined whether or not the wheel number W determined in step 6_3 exceeds 90. There are 91 wheels and each wheel has 0-
One of the numbers 90 is assigned (see steps 5_1 and 5_4 in FIG. 5), and there is no wheel with a wheel number exceeding 90. On the other hand, in step 6_3, since the wheel number is mechanically calculated from the key number of the pressed key and the variable drb, a wheel number exceeding 90 may be obtained depending on the pressed key.
Therefore, it is determined in step 6_4 whether the wheel number exceeds 90. If the wheel number exceeds 90, the process proceeds to step 6_5, and 12 is subtracted from the wheel number obtained in step 6_3. The reduced wheel number is stored again in W, and again in step 6_
Return to 4. A wheel is prepared for each semitone, so subtracting 12 from the wheel number means using a wheel with a pitch one octave lower.

In step 6_6, it is determined whether the wheel number calculated in step 6_3 is a negative number. Since the wheel number starts from 0 and there is no wheel with a negative number, if the wheel number is negative, the process proceeds to step 6_7, where 12 is added to the wheel number and stored again in W, and
Return to _6. Adding 12 to the wheel number means returning to a wheel with a pitch one octave higher.

When the wheel number enters any one of the numbers 0 to 90 in this way, the routine proceeds to step 6_8, where an organ voice assignment routine is executed. In the organ voice assignment routine, an operation of outputting a periodic signal corresponding to the wheel number W obtained this time with an amplitude corresponding to the pull amount of the corresponding drawbar is performed.
Details will be described later.

In step 6_9, it is determined whether drb <8 or not.
That is, as described above, it is determined whether or not the above processing has been completed for all nine drawbars. If there are unprocessed drawbars, the process proceeds to step 6_10, drb is incremented, and the process returns to step 6_2.
When the processes for all the drawbars have been completed, the key pressing routine of this organ is completed.

FIG. 7 is a flowchart of an organ voice assigning routine executed in step 6_8 of the organ key pressing routine shown in FIG. Step 7_1
Then, it is determined whether or not the wheel reference counter [W] corresponding to the wheel number W obtained up to step 6_8 in the organ key pressing routine shown in FIG. 6 is -1. The wheel reference counter [W] =-1 means that a voice is lent out for synth sound generation, and a voice that generates a periodic signal of the wheel (wheel W) of the wheel number W exists. Not mean.

In this case, that is, when the wheel reference counter [W] = − 1, the process proceeds to step 7_2, where 1 is set to the wheel reference counter [W]. Here, the generation of the organ sound is prioritized over the generation of the synth sound. Therefore, even when the corresponding voice does not exist, the voice is always secured as described below. The wheel reference counter [W] is set to 1 indicating that the corresponding voice is being used for one organ sound. However, at the stage of this step 7_2, there is no voice generating the corresponding periodic signal. The corresponding voice is determined in the steps described below.
In step 7_3, it is determined whether or not the idle number> 0. The number of idles represents the number of voices that are not currently used to produce organ sounds, but are idle for future organ sounds. If the number of idles> 0, the process proceeds to step 7_4, in which the number of idles is decremented. Further, the process proceeds to step 7_5, a wheel whose wheel reference counter value is 0 is found, and the number thereof is set in a variable tmp.
Set to _W. That is, for example, the wheel reference counter [tmp_W] = 0 is found while incrementing tmp_W sequentially from 0. In step 7_6, by setting the found wheel reference counter [tmp_W] to −1, it is determined that the voice corresponding to the wheel reference counter [tmp_W] has not generated a periodic signal corresponding to the wheel. The wheel voice [tmp_W] is read out and set to a variable V (step 7_7), a periodic signal of the wheel W is generated in the WG of the voice V, and the TVF
Is set to the maximum value, and the amplification factor of the TVA is set to 0 to make the voice V idle at the pitch of the wheel W (step 7_8), and the volume of the voice V is reduced by the amount of the drb-th drawbar. (Step 7_9).

If it is determined in step 7_3 that the number of idles is not greater than 0, that is, that there is no voice that is currently in an idle state, the process proceeds to step 7_10, in which one voice is acquired from the history link and its voice number is obtained. Is substituted for V. Here, the history link has a data structure in which the voice numbers of the voices currently producing the synth sound are connected in order from the voice that started to produce sound earlier in time. In step 7_10, the voices arranged in the history link are connected. The voice number at the head of the numbers, that is, the voice number of the voice that started sounding at the oldest time is removed from the history link, and the voice number is substituted for the variable V.

In step 7_11, the synth sound being produced by the voice V is truncated (silence processing), in step 7_12, the number of rents from the organ sound to the synth sound is decremented, and in step 7_8, the synth sound is truncated in this manner. Step 7_9, the voice V secured by the operation is idle at the pitch of the wheel W.
In this case, the volume of the voice V is increased in accordance with the pull amount of the drb-th drawbar.

If it is determined in step 7_1 that the wheel reference counter [W] is not -1, step 7_1
3 and the wheel reference counter [W] is now 0
Is determined. The wheel reference counter [W] = 0 means that there is a voice that is idle at the pitch of the wheel W,
In this case, the process proceeds to step 7_14, where the idle number is decremented, and further proceeds to step 7_15, where the wheel reference counter [W] is incremented. When the process proceeds to step 7_15 via step 7_14, the wheel reference counter [W] becomes 1 as a result of incrementing the wheel reference counter [W] in step 7_15. This means that the voice V generating the pitch periodic signal is being used for one organ sound. Step 7_
At 16, the value of the wheel voice [W] is read and the variable V
To step 7_9, and the volume of the voice V is increased in accordance with the pulling amount of the drb-th drawbar.

When it is determined in step 7_13 that the wheel reference counter [W] is not 0, that is, when it is determined that the wheel reference counter [W]> 0, based on the determination in step 7_1 as well, Since the pitch signal of the pitch W is already being used in another organ sound and is not already in the idle state, the process directly proceeds to step 7_15 without passing through step 14. Subsequent processing is the same as the processing through step 7_14, and a duplicate description will be omitted.

In this way, an organ sound corresponding to a key depression is generated and output. FIG. 8 is a flowchart of an organ key release routine executed in response to an organ sound mute instruction. As described above, in the present embodiment, an organ sound is generated by depressing the keyboard 14, and when the key being depressed is released correspondingly, an organ sound muting instruction is generated, and the organ sound shown in FIG. A sound key release routine is executed.

All other steps 8_1 to 8_7 except for the step 8_8 of the organ key release routine shown in FIG.
Steps 8_9 and 8_10 are the same as steps 6_1 to 6_7, 6_9 and 6_10, respectively, except for step 6_8 in the organ key pressing routine shown in FIG. That is, in the organ key pressing routine (FIG. 6), the organ voice assignment routine is executed via the steps shown in FIG. 6, but in the organ key releasing routine shown in FIG. 8, the same steps are executed. Then, an organ voice release routine is executed. A more detailed invention of the organ key release routine shown in FIG.
The description of the key pressing routine of the organ shown in FIG.

FIG. 9 is a flowchart of an organ voice release routine executed in step 8_8 of the organ key release routine shown in FIG. In step 9_1, the wheel reference counter [W] is decremented. That is, the number of organ sounds being generated is decremented using the pitch signal of the wheel W.

In step 9_2, it is determined whether the wheel reference counter [W] has become 0 as a result of decrementing the wheel reference counter [W] in step 9_1. When the wheel reference counter [W] is not 0,
That is, here, the wheel reference counter [W] = − 1
Is impossible, when the wheel reference counter [W]> 0, in other words, when there is another organ sound being generated using the periodic signal of the pitch of the wheel W, Proceeding to step 9_4, the wheel voice [W] is read out and set to the variable V, and in step 9_5, the volume of the voice V is reduced by an amount corresponding to the pulling amount of the drb-th drawbar.

If the wheel reference counter [W] is decremented in step 9_1 and it is determined in step 9_2 that the wheel reference counter [W] = 0, the process proceeds to step 9_3, where the idle number is incremented, and so on. Then, the process proceeds to steps 9_4 and 9_5. Here, when the wheel reference counter [W] = 0, it means that there is no other organ sound being produced using the periodic signal of the pitch of the wheel W, and accordingly, In step 9_5, the volume of the voice V is reduced to 0, and the voice V enters an idle state.

FIG. 10 is a flowchart of a synth voice key pressing routine. In the present embodiment, the key press and key release related to the synth sound are stored in the ROM 12 or RAM 1.
3 is stored in advance as performance data, and the performance data is sequentially read out according to the time schedule. If the read performance data is performance data corresponding to key depression, a synth voice shown in FIG. Is executed, and if the read performance data is performance data corresponding to key release, a key release routine of a synth voice shown in FIG. 11, which will be described later, is executed.

When the execution of the key press routine of the synth voice shown in FIG. 10 is started, first, in step 10_1, it is determined whether or not there is a voice whose voice number is connected to the free voice link. As mentioned earlier, the free voice link lists the voice numbers of free voices that are not currently used for organ sounds or synth sounds and are not idle for organ sounds. Link.

If a voice number has been connected to the free voice link, the process proceeds to step 10_2, where one voice number is removed from the free voice link, and the voice number is substituted for a variable V. Thereafter, the process proceeds to step 10_3, in which a variable (voice number) V is added to the end of the history link (indicating that the latest sounding has started). A synth sound corresponding to the corresponding performance data is generated by the voice V. In generating the synth sound in step 10_4, a parameter WG for selecting a waveform generated by the WG of the voice V (see FIG. 4)
Various parameters, such as a parameter for pitch conversion in the above, a parameter representing an initial value of the cut-off frequency of the TVF, and a parameter representing the initial value of the amplification factor of the TVA, are set in the voice V. A sounding instruction is issued, and the sounding of the synth sound is started by the voice V in this manner. After the sounding of the synth sound is started in the voice V, as will be described later (see FIG. 12), parameters such as the cut-off frequency of the TVF and the amplification factor of the TVA are updated at predetermined intervals (for example, every 10 msec). . This parameter update process is called a voice service. Since the processing and voice service for starting the sounding of the synth sound are conventionally known and are not the subject of the present invention, further detailed description thereof will be omitted.

If it is determined in step 10_1 that there is no voice number connected to the free voice link, the process proceeds to step 10_5. In step 10_5,
It is determined whether or not the number of idles is 0, that is, whether or not there is a voice that is not currently contributing to the actual sounding of the organ sound but is in an idle state. Number of idles is 0
If not, that is, it is determined that there is an idle voice, the process proceeds to step 10_6, the number of idles is decremented, and further proceeds to step 10_7, where the wheel reference counter [W] = 0 (ie, idle) The wheel number W (corresponding to the voice in the state) is found, and its wheel reference counter [W] is set to -1 indicating lending to the synth sound (step 10_8), and the lending number is incremented (step 10_9). The wheel voice [W] is read and assigned to a variable (voice number) V (step 10_1)
0). Thereafter, similarly to the above, the process proceeds to step 10_3, a variable (voice number) V is added to the end of the history link, and the process proceeds to step 10_4, where the voice V is started to produce a synth sound.

If it is determined in step 10_5 that the number of idles is 0, that is, that there is no idle voice for generating the organ sound, step 10_1
Go to 1 and select the voice number in the front row of the history link (of the voices currently producing synth sounds,
The voice number of the voice that started generating the synth sound at the oldest time is removed from the history link, and the voice number is substituted for the variable V. Step 10_12
Then, the synth sound being generated in the voice V is truncated, and in step 10_3, the variable (voice number) V is set so as to start generating a new synth sound in the voice V.
Is added again to the history link, but this time at the end, and the process proceeds to step 10_4, where the voice V starts generating a new synth sound.

In the case of the above-mentioned organ sound, if there is no vacant voice or an idle voice, a voice for the organ sound is secured even if the synth sound is truncated. As shown in FIG. 10, when there are no free voices or idle voices, the organ sound is left untouched.
Truncate the old synth sound with the oldest sounding start time to produce a new synth sound. As described above, here, the sound of the organ sound is given priority over the sound of the synth sound.

FIG. 11 is a flowchart of a synth voice key release routine. As described above, the key release routine of the synth voice is executed when the performance data corresponding to the key release is read. In the key release routine of the synth voice shown in FIG. 11, the tone that is being sounded by the voice that has started sounding because the released key is pressed at a timing before the key release is replaced with the released tone. Is changed to In the case of a synth sound, for example, even if a piano key is depressed and then immediately released, the musical tone may be maintained for a while after the key is released as if the musical tone was connected. This tone generation state after the key release is called a release state. When the state shifts to the release state, the state immediately or later slowly goes to the mute state depending on the nature of the musical tone being sounded.

FIG. 12 is a flowchart of the timer routine. This timer routine is executed, for example, for 10 msec.
It is executed each time in response to a timer interrupt generated at a short time interval such as an interval. When this timer routine is started, first, a variable V is set to an initial value 0 (step 12_).
1) It is determined whether or not the synth sound is currently being generated by the voice V (step 12_2). If the synth sound is not currently sounded in voice V, step 12
_6, it is determined whether or not V> 126, that is, whether or not processing has been completed for all voices. If there are unprocessed voices, the flow proceeds to step 12_7, where the variable (voice number) V is incremented. Been step 1
Return to 2_2.

If it is determined in step 12_2 that the synth sound is currently being generated in voice V, step 12
The process proceeds to _3, and the voice service (update processing of parameters such as TVF and TVA (see FIG. 4)) for the voice V is performed. Next, at step 12_4, whether the level of the voice V is 0, that is, in the past, the synth sound which is being sounded by the voice V is shifted to the release state by the corresponding key release, and further thereafter, at 10 seconds.
Every msec, the amplification rate in the TVA is gradually reduced in step 12_3, and the level of the synth sound generated in the voice V at the time when this timer routine is executed becomes zero.
(Mute state) is determined. When the level has not yet reached 0, that is, when the sound is still being generated, the process proceeds to step 12_6.

On the other hand, when the level becomes 0, the process proceeds to step 12_5, where a synth voice release routine for releasing the voice V from the generation of the synth sound is executed, and thereafter, the process proceeds to step 12_6. The above processing is performed for voice numbers 0
127 for all voices (step 1
2_6, 12_7).

FIG. 13 is a flowchart of a synth voice release routine executed in step 12_5 of the timer routine shown in FIG. In step 13_1,
Voice number V is deleted from the history link. Here, the voice numbers arranged at any positions of the history link can be deleted.

At step 13_2, it is determined whether or not the number of loans from the organ sound to the synth sound is zero. When the number of rentals is 0, the process proceeds to step 13_3, and the voice V is inserted into a free voice link which is a link of a voice in an empty state. In step 13_2, the number of loans is 0
If it is determined that the voice V has not reached this level, a process of shifting the voice V whose synth sound level has become 0 to the idle state for the organ sound is performed after step 13_4.

That is, the lending number is decremented in step 13_4, the wheel number W of the wheel reference counter [W] = − 1 is searched in step 13_5, the idle number is incremented in step 13_6, and the wheel voice is incremented in step 13_7. [W]
Is set to the variable (voice number) V,
At 8, the wheel reference counter [W] is set to 0, and the voice V is put into the idle state by the pitch signal of the pitch of the wheel W.

Here, in the above embodiment, the synth sound is abstractly represented as a synth sound, but this synth sound is played by playing the tone wheel organ, which is generated by playing the keyboard 14. The musical sound of another instrument that is being played may be performed, and the key press processing and the key release processing of the synth sound are executed in accordance with the key press and the key release of the keyboard 14 similarly to the organ sound, and the synth sound is used. When you play the keyboard 14 and play the tone wheel organ, you may simulate the noise sound of the organ sound accompanying key press and key release, and use the synth sound in the same way as with a conventional synthesizer. There may be.

In the above embodiment, the organ sound is played on the keyboard. However, the organ sound may be held as performance data, and the keyboard may be used for key press and release of the synth sound. Alternatively, both the organ sound and the synth sound may be included as performance data. In that case, the keyboard is not always necessary. In addition to instructing the generation of an organ sound or synth sound and the instruction to stop sounding in accordance with the operation of the keyboard or performance data stored inside the electronic musical instrument, the instruction is also supplied from outside the electronic musical instrument through communication means such as MIDI. May be performed according to the performance data.

Further, in the above-described embodiment, an example has been described in which the level of each harmonic of the organ sound is set continuously or in three or more stages according to the draw bar drawing amount. However, the level of each harmonic is set in two stages. May be set. That is, on /
The state of each overtone may be selected as sounding or non-sounding according to the operation of the off switch. In addition to setting the level of each overtone in accordance with the operation of the operating element, the level of each overtone is also determined according to a parameter value stored inside the electronic musical instrument or a parameter value supplied from outside the electronic musical instrument through a communication means such as MIDI. May be set.

Further, in the above-described embodiment, detailed description has not been given with respect to the volume change of the volume of the voice that generates the organ sound in response to the organ sound generation instruction and the sound generation stop instruction. . The volume may be changed by an amount corresponding to the draw-out amount of the drawbar related to the sounding instruction and the sounding stop instruction when the sounding instruction and the sounding stop instruction are performed. Since the volume of the voice is determined in proportion to the number of keys, the volume may be unnaturally increased. In such a case, the amount of change in the volume of the voice may be determined depending on the number of key presses related to the sounding of one voice. For example, when increasing the volume of a voice in an idle state in response to a first sounding instruction, increasing the volume by the draw bar withdrawal amount, and further increasing the volume of the voice in response to a second sounding instruction. When the volume of the voice is further increased in accordance with the third sounding instruction, the volume is increased by １ ／ of the drawbar withdrawal amount. Conversely, in response to the sounding stop instruction, the volume is decreased by 1/4, 1/2, or 1 of the draw bar withdrawal amount. Alternatively, regarding the sounding instructions after the second sounding instruction, the sound volume of the voice may not be changed even if there is a sounding instruction. Even in this case, it is rare that the volume of the voice corresponding to all the drawbars does not change, and in most cases, the voice corresponding to any of the drawbars is newly pronounced from the idle state. Can feel that a new musical tone has been pronounced in response to the pronunciation instruction.

In the above embodiment, the number of wheels (91) is used to generate synth sounds in addition to organ sounds.
The above-mentioned voices (128 voices) are provided. However, when only organ sounds are generated, even if there are only a few voices less than the number of wheels of the wheel organ to be simulated, there is a considerable number of voices. No inconveniences occur. That is, the number of fingers of the hand is ten, but in a normal performance, fewer keys are pressed simultaneously. Therefore, even if it has nine drawbars, the number of harmonics specified to be generated at the same time is at most about 90. In the present invention, since one voice is shared by a plurality of pronunciation instructions, a plurality of overtones are generated by the same voice, and the number of required voices is smaller than the number of overtones to be generated. I'm done. The number of wheels of the wheel organ to be simulated as the substantial number is about 2 /
It is enough to have about 3 voices.

In the above embodiment, the tone color of the tone wheel organ is simulated. However, the present invention can also be applied to a tone model of a musical instrument other than the tone wheel organ such as a pipe organ. When simulating the tone of a pipe organ, a waveform corresponding to each pipe is generated for each voice. Further, the present invention is not limited to this, and can be applied to a case where the timbre of a specific musical instrument is not simulated. That is, the present invention is to generate a plurality of musical tones of different pitches in response to one sounding instruction, and can be applied to a case where one voice is shared by a plurality of sounding instructions, and is also applicable to a so-called sine wave synthesis method. it can.

In the above-described embodiment, the synth sound is generated by using only one voice according to one sounding instruction. However, the musical sound is generated by using a plurality of voices according to one sounding instruction. You may make it.

[0078]

As described above, according to the electronic musical instrument of the present invention, a tone wheel organ can be simulated very well.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating the principle of an electronic musical instrument according to the present invention.

FIG. 2 is a block diagram illustrating a hardware configuration of the electronic musical instrument according to the embodiment of the present invention.

FIG. 3 is a panel diagram showing a part of a keyboard and panel switches.

FIG. 4 is a conceptual configuration diagram of a tone generator.

FIG. 5 is a flowchart of an initialization routine.

FIG. 6 is a flowchart of an organ key pressing routine that is executed in response to a sound generation instruction for generating an organ sound.

FIG. 7 is a flowchart of an organ voice assignment routine.

FIG. 8 is a flowchart of an organ key release routine executed in response to an organ sound mute instruction.

FIG. 9 is a flowchart of an organ voice release routine.

FIG. 10 is a flowchart of a synth voice key pressing routine.

FIG. 11 is a flowchart of a synth voice key release routine.

FIG. 12 is a flowchart of a timer routine.

FIG. 13 is a flowchart of a synth voice release routine.

[Explanation of symbols]

 1_1, 1_2,..., 1_n Voices 2_1, 2_2,. RAM 14 keyboard 15 panel switch 16 tone generator 17 bus 20 sound system 141 white key 142 black key 151 drawbar 152 tone switch 153 synth tone selection LED

Claims (7)

[Claims] 1. A constituent sound for specifying a pitch of each constituent sound of a musical tone composed of a set of a plurality of voices and a plurality of constituent sounds each having a different pitch for generating one arbitrarily assigned sound. Pitch designation means, prior to musical tone generation, causes a plurality of voices to be in charge of generating unit sounds of different pitches, and reduces the level of the generated unit sound to a state where output of the unit sound is blocked. A unit sound allocating unit to be reduced, which receives a sounding instruction of a musical tone composed of a set of a plurality of constituent sounds, and generates a key number designated by the sounding instruction and a pitch of each constituent sound designated by the constituent pitch specifying means. In response, a unit sound determining unit that determines a unit sound corresponding to each of a plurality of constituent sounds constituting a musical tone for which a pronunciation instruction has been given, and is responsible for generating among a plurality of unit sounds determined by the unit sound determining unit. ing For a unit sound where a voice exists, the level of the voice is increased, and for a unit sound for which no voice is in charge of generation, a unit sound which contributes to the output of a musical tone composed of a set of a plurality of unit sounds. An electronic musical instrument comprising: a tone generating means for causing any one of the voices other than the voice to newly generate the unit sound, and generating and outputting a tone for which the pronunciation is instructed. 2. The tone generator according to claim 1, wherein the unit sound includes a unit sound for which a voice responsible for generation is present among the plurality of unit sounds determined by the unit sound determiner. By increasing the level by the level of the constituent sound corresponding to the unit sound that composes the given musical tone, a musical tone whose pronunciation is instructed is generated and output. When a pronunciation stop instruction is issued, the level of a unit sound generated by a plurality of voices that are responsible for generating a plurality of constituent tones constituting the musical tone being produced constitutes the musical tone being produced. 2. The electronic musical instrument according to claim 1, wherein the generation of the tone is stopped by reducing the level of each of the plurality of constituent sounds. 3. The constituent pitch specifying means specifies the level of each constituent sound in addition to the pitch of each constituent sound, and the musical tone generating means configures a musical tone for which a pronunciation instruction has been given. The level of the unit sound is increased by the level of the constituent sound corresponding to the sounding instruction specified by the constituent pitch specifying means, and the level of the unit sound constituting the musical sound for which the sound generation stop instruction is given is increased. 3. The electronic musical instrument according to claim 2, wherein the musical tone is reduced by the level of the constituent sound corresponding to the sound generation stop instruction specified by the constituent pitch specifying means. 4. A unit voice corresponding to a tone which is instructed to be pronounced, and a unit voice in which a voice responsible for generation is present is increased by increasing the level of the voice, so that one voice is divided into a plurality of sounds. Receiving a sounding instruction of a second form of musical tone, which is different from the first form of musical sound shared by the instruction and uses one voice solely in one sounding instruction,
A second mode in which any of the voices except for the voice in the first state which is in the middle of contributing to the output of the first mode of musical tone is responsible for generating the second mode of musical tone instructed to sound. 2. The electronic musical instrument according to claim 1, further comprising: 5. If there is a voice in a second state which is not currently in charge of generating a sound, the second tone generating means may include one of the voices in the second state. Is in charge of the generation of the second form of the musical tone that has been instructed to sound, and when there is no voice in the second state, of the voices to which the generation of the unit sound is allocated by the unit sound allocation means. Except for the voice in the first state, any of the voices in the third state in which the level of the generated unit sound has been reduced to a state in which the output of the unit sound is blocked is generated. When the voice of the second state and the voice of the third state are not present, the second form of the musical tone is instructed. The fourth state in charge of generating musical sounds of the form And stops the generation of the second form of musical tone being generated by the voice of any of the voices, and is in charge of the generation of the second form of musical tone for which a new pronunciation instruction has been given. 5. The electronic musical instrument according to claim 4, wherein: 6. A method according to claim 1, wherein said unit tone allocating means assigns a voice in charge of generation of said second form of musical tone when said generation of said second form of musical tone in said voice in charge of said second form of musical tone has been completed. If there is a unit tone that has not been assigned, generation of any one of unit sounds to which the voice responsible for generation is not assigned to the voice for which the generation of the tone of the second form has been completed. 5. The electronic musical instrument according to claim 4, wherein the electronic musical instrument is assigned to an electronic musical instrument. 7. A method according to claim 1, wherein said tone generating means does not include any voice which is in charge of generating any one of the plurality of unit sounds determined by said unit sound determining means. Excluding the voice in the first state among the voices to which the generation of the unit sound is allocated by the sound allocating unit,
If there is a voice in the third state in which the level of the generated unit sound has been reduced to a state in which output of the unit sound is blocked, any of the voices in the third state A voice is made to generate a unit sound for which no voice responsible for generation exists among a plurality of unit sounds determined by the unit sound determination means, and no voice in the third state exists. In this case, any one of the voices in the fourth state which is in charge of the generation of the tone of the second form stops generating the tone of the second form being generated by the voice. 7. The apparatus according to claim 6, wherein said unit tone generation unit is in charge of generating a unit sound for which no voice responsible for generation exists among a plurality of unit sounds determined by said unit sound determination unit. Electronic musical instrument.