JP2836028B2 - Automatic performance device - Google Patents

Description

The present invention relates to an automatic performance device such as a sequencer of an electronic musical instrument.

[Conventional technology]

In an electronic musical instrument using computer technology, a microcomputer (microcomputer; hereinafter the same) executes a sound source processing program stored in a ROM or the like to set various switches such as a tone setting switch and a keyboard portion. The tone generation operation is performed on the basis of the performance information output from the device.

In this case, according to the program, the microcomputer scans which key is turned on or off, scans which tone setting switch or the like is pressed, and emits or silences a tone corresponding to the processing result. The processing and the like are executed sequentially and repeatedly. Instructions that are repeatedly executed in this manner are generally called a main routine. The time required for one round of processing of the main routine varies depending on the number of key-on / key-off keys or the number of operations of various switches, but is at most about 70 msec (millisecond).

Accordingly, the microcomputer can execute the above-described series of sound source processes while sufficiently following the operation of the player operating the keyboard instrument or the like.

Recently, many automatic keyboard instruments having an automatic performance or an automatic performance function have been realized. The function of automatic performance is to store performance data (hereinafter referred to as event data) for generating an event such as note-on / note-off stored in a commercially available ROM or the like or stored in a RAM or the like in advance by a player. , And is sequentially read out at a time timing corresponding to the time data stored in the pair, and the sound generation / mute processing is executed based on the event data. The time data in this case is the note length of each note to be automatically played (whole note, half note, quarter note,
8th note, 16th note, 32nd note, etc.). The above-described reproduction processing of the automatic performance data (event data and time data) is also realized as a part of the main routine.

Here, the process of measuring the time timing compares the time data set each time the event data is read with a timer value that is counted up independently of the operation of the main routine by a timer interrupt from a hardware timer. This is a process for determining whether or not the time corresponding to each time data has elapsed. This processing is executed with the maximum time required for one round of the main routine being about 70 msec as the longest interval.

Now, consider a case where a musical tone signal corresponding to a sixteenth note is automatically played continuously. In this case, the time data stored as the automatic performance data is defined according to the type of note. For example, "1" for a 16th note, "2" for an eighth note, "4" for a quarter note, "2" A minute note is like "8".

In the above case, if the automatic performance is performed at a tempo in which quarter notes are emitted 60 times per minute, for example, as shown in FIGS. 8 (a) and 8 (c), each note F _4, A _4, G _4, etc., (60 seconds / 60 times) × (4/16) = 250ms
pronounced at ec intervals. At this time, the timer interrupt occurs once every 259 msec as shown in FIG. 8 (d), and the above-mentioned timer value increases by +1 at that rate.
Then, in the process of measuring the time timing, it is determined whether or not the timer value has been counted by a value “1” corresponding to each time data since the previous sounding, and based on this, the sounding operation based on the next event data is performed. Done. in this case,
Time timing processing is up to 70ms as described above
Since the execution is performed at the ec interval, the actual tone generation timing may be delayed up to about 70 msec from the target timing, but the automatic performance is realized at the target timing substantially within the deviation range. That is, the processing for measuring the time timing in the main routine is executed at the timing indicated by “↑” in FIG. 8 (e). And the eighth
At the timing of “↑” with symbols such as F ₄ , A ₄ , G _{4 in} FIG. 9E, the timer value has been counted by the value “1” corresponding to each time data since the previous tone generation. Is detected,
A tone corresponding to the event data is generated.

[Problems to be Solved by the Invention] On the other hand, in the above-mentioned conventional example, when the reproduction tempo of the automatically played musical tone is increased, the following problems occur.

That is, the automatic performance is performed, for example, quarter notes are 25 minutes per minute.
If performed at a tempo that is pronounced five times, as shown in FIGS. 9 (a) and 9 (c), each of the sixteenth notes
F ₄ , A ₄ , G ₄ , ... are (60 seconds / 255 times) x (4/16) ≒ 59
It is pronounced at msec intervals. At this time, the timer interrupt occurs once every 59 msec as shown in FIG. 9 (d), and the above-mentioned timer value increases by +1 at that rate. However, as described above, since the processing for measuring the time timing is executed at intervals of 70 msec at the longest, it often occurs that the processing cannot catch up. As can be seen from a comparison between FIGS. There is a problem that the sounding timing of the event data read out from the program accumulatively lags behind the respective target sounding timings, resulting in a performance that is unbearable for listening.

An object of the present invention is to realize an automatic performance with less listening difficulty even if the reproduction tempo is increased.

[Means for solving the problem]

According to the present invention, event data for controlling a tone generation operation of a musical tone in a sound source is sequentially read from an automatic performance storage means at a timing according to the time data while reading a time table stored therewith. It is assumed that an automatic performance device that causes a sound source to perform an automatic performance by repeatedly executing an automatic performance control process of instructing a sound source based on data as a part of a main routine consisting of a series of processes.

First, there is provided a timer value storing means for storing a timer value that counts up at a cycle based on an arbitrarily set tempo from the start of the automatic performance. This timer value is incremented by +1 by, for example, an interrupt from a hardware timer that operates at a cycle based on an arbitrarily set tempo.
Counts up by one.

Further, the automatic performance control process includes the following first to first steps.
It consists of a sixth process.

That is, the automatic performance control process determines whether or not the above-described timer value is equal to or greater than the time data currently read from the automatic performance storage means. If the determination is true, the following second to sixth steps are performed. Is performed, and if the determination is false, a first process for ending the process at the current timing is performed.

The automatic performance control processing is executed when the determination of the first processing is true, and the second processing is performed in which a value obtained by subtracting the time data from the timer value is set as a new timer value. I do.

The automatic performance control process is executed when the determination of the first process is true, and executes a third process of reading out new event data from the automatic performance storage means.

The automatic performance control process is executed when the determination of the first process is true, and executes a fourth process of instructing the sound source to generate sound based on the read event data.

Further, in the automatic performance control process, after the end of the fourth process, a fifth process of repeating the above-described first process for a new timer value based on the second process and new event data based on the third process. Execute

Then, the automatic performance control processing is performed when the event data instructing the sounding of an arbitrary musical tone and the event data instructing the silencing of the musical tone are successively read from the automatic performance storing means at the current timing. A sixth process for controlling the sound source so as not to issue a tone generation control instruction to the sound source without executing the fourth process of the above is executed.

[Action]

The automatic performance control process is repeatedly executed as a part of a main routine consisting of a series of processes,
In an automatic performance device of a type in which a sound source is automatically performed, if the user sets the tempo to be faster, the repetition of the automatic performance control process may not be able to catch up with the tempo.

In such a case, the timer value in the timer value storage means is
The value advances according to the tempo from the current time data value.

On the other hand, in the automatic performance control processing, the current time data is subtracted from the timer value every time the second processing is executed, and the next timing control is performed based on the new timer value, thereby performing the automatic performance control processing. The timing of the automatic performance is advanced by an amount corresponding to the delay of the repetition of the processing from the tempo.

In the automatic performance control process, if the above-mentioned new timer value is still equal to or longer than the newly read time data after one event data process at the repetition timing in this main routine, The process of reading event data from the performance storage means is repeated. In this case, the automatic performance control processing is performed when the event data instructing the sounding of an arbitrary musical tone and the event data instructing the silencing of the musical tone are successively read from the automatic performance storing means at the current timing. Then, control is performed so that the tone generation control instruction for the musical tone is not issued to the sound source. By such control processing, the event data delayed from the tempo is appropriately canceled, and the delayed sound is not generated.

As described above, even if the reproduction tempo is increased, an automatic performance with less listening difficulty is realized.

〔Example〕

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

Configuration FIG. 1 is a block diagram of an embodiment of the present invention.

The performer performs a performance using the keyboard 104, and the performance information generated from the keyboard 104 by the performance is the microcomputer 1
Captured in 01. The performer designates a tone with the tone key 108, gives an instruction to start recording of the automatic performance with the recording key 109, stops recording of the automatic performance with the start / stop key 110 (when recording), and plays the automatic performance. An instruction to start (when not playing) or stop playing (when playing) is given. Further, the player operates the tempo up key 111 or the tempo down key 112 to increase or decrease the tempo of the automatic performance.

For each of the above-mentioned operations, the microcomputer 101 executes the sound source processing program stored in the ROM 102 with the RAM 105 as a work memory, thereby identifying the operation state of each part, and instructing the sound source 106 to perform note-on / note-off and various other operations. Is set in the sound source 106. Also, the microcomputer 101
Plays back the automatic performance data stored in the RAM 105 and
Are sequentially set.

In response to the above operation of the microcomputer 101, the sound source 106
A corresponding musical tone is generated via the speaker 107 or the like.

Operation of Main Routine Next, an operation flowchart of the main routine of the sound source processing program executed by Myco 101 in FIG. 1 is shown.

First, when a sound source switch (not shown in FIG. 1) is turned on, a work area on the RAM 105 used for sound source processing is initialized ( _S201 ).

Next, the following processing S ₂₀₂ is manipulation-back portion of the main routine, by processing represented by this part is repeatedly performed by the microcomputer 101, the tone generation control operation based on the keyboard performance or automatic performance of the player Be executed. The time required for one round of processing of this main routine is 70 msec (millisecond) at the maximum as described in the section of [Prior Art].
It is about.

First, in the keyboard scanning process ( _S202 ), it is detected whether each key of the keyboard 104 has been pressed, and an initial touch, an after touch, and the like of the pressed key are detected.

In the sound production / silence processing (S ₂₀₂ ), based on a key operation performed by the player, processing for instructing the sounding of the musical sound of the note number corresponding to the sound source 106, processing for instructing the silencing of the musical sound of the note number corresponding to the sound source 106, Is performed.

Next, it is determined whether or not the recording key 109 has recently been operated by the player to record the automatic performance data ( _S202 ). If it is determined that recording is in progress, note-on / note-off event data obtained by the player operating the keyboard 104 and time data indicating the occurrence timing are obtained in the recording process ( _S205 ). Are sequentially stored in an area for automatic performance data in the RAM 105. The time data is stored in the timer 103
Is generated based on the timer value obtained from. If it is determined that recording is not being performed, the above-described recording process is not performed.

Subsequently, only the start / stop key 110 is operated by the player from a state where automatic performance has not been performed recently, and it is determined whether or not automatic performance data is currently being reproduced ( _S206 ). If it is determined that the reproduction is being performed, in the reproduction processing ( _S207 ), the time data stored in the RAM 105 is sequentially read, and the read processing of the event data is executed while controlling the timing.
Further, in the sound generation / silence processing ( _S208 ), processing of sound generation or mute of a musical tone based on the read event data is executed. The reproduction processing and the sound generation / mute processing will be described later in detail. If it is not determined that reproduction is being performed, the above-described reproduction processing and sound generation / mute processing are not performed.

In the function key scan process ( _S209 ), the tone key
The operation state of each of the 108, the recording key 109, the start / stop key 110, the tempo up key 111, and the tempo down key 112 is scanned.

As a result, if it is determined that any of the tone key 108 is turned on (the determination in S ₂₁₀ YES), the change instruction to the corresponding tone is made to the sound source 106 (S _211).

Start / stop key 110 is judged to have been turned on (YES determination at S212), further, if it is determined that the state of the currently recording (YES determination at S _213), NO is determined in the aforementioned S ₂₀₄ In particular, the corresponding flag stored in a register or the like (not shown) is reset. On the other hand, is not determined to be the state of the currently recording (determination is NO S _213),
Further, if it is determined that the state is currently being reproduced (S
_If the determination at ₂₁₅ is YES), the corresponding flag stored in a register or the like (not shown) is reset to make the determination at _S206 described above NO. If it is determined that the state is not the current playback state (the determination in _S215 is NO), a corresponding flag stored in a register (not shown) is set in order to set the determination in _S206 to YES. Is done.

If a recording key 109 is judged to have been turned on (YES determination at S _218), the corresponding flag is set in particular is stored in the register (not shown) or the like in order to YES determination of the above S ₂₀₄ .

If it is determined that the tempo up key 111 has been turned on (S
The determination at ₂₂₀ is YES), and the timer period (described later) is reduced by a predetermined amount.

If it is determined that the tempo down key 112 has been turned on (S
_If the determination at ₂₂₁ is YES), the timer period is extended by a predetermined amount.

In the processing of the main routine described above, the reproduction processing of the automatic performance data in _S207 is a particularly characteristic processing part in the present embodiment. Here, as described above, the time data stored in the RAM 105 is sequentially read out, and the event data reading process is executed while the timing is controlled, and the time data is read for the next sound generation / silence processing ( _S208 ). The sound source 106 is instructed to sound or mute at the time of automatic performance based on the issued event data. In this case, when the playback tempo of the automatic performance is not so fast, the playback operation of the normal automatic performance is performed. However, when the playback tempo of the automatic performance becomes faster, the event data is appropriately canceled. By doing so, even if the re-tempo is raised, automatic performance with less listening difficulty is realized.

Hereinafter, after explaining the automatic performance data, FIG.
Sequentially described in detail pronunciation-mute processing of the playback processing and S ₂₀₈ of the S _207.

Data Configuration of Automatic Performance Data First, in the automatic performance data stored in a specific area of the RAM 105 (FIG. 1), as shown in FIG. 5A, event data and time data are stored in pairs. Here, when the value of the event data is “0” to “87”, note-off of the key code of that value is instructed, and when the value is “88” to “175”,
If the value of {event data-88} is a key code note-on instruction, and if the value is "182", an instruction to stop automatic performance is issued. The time data indicates the timing from when the immediately preceding event data is read to when the immediately following event data is read. In the case of the timing corresponding to the sixteenth note, "1". “2” has a value of “4” for a quarter note, and “8” for a half note. In addition,
When the minimum note unit of the sequencer is expressed more finely than the 16th note, for example, up to 32 notes, the timing corresponding to the 32nd note is "1", the 16th note is "2", etc. Is set.

FIG. 5B shows a specific example of the automatic performance data. In this case, each note F ₄ , F ₄ , A ₄ , A ₄ , G ₄ , G ₄ , F ₄ , each of 16th notes (time data is “1”)
・ The automatic performance of each musical tone corresponding to is designated.

Then, around the regeneration process in Figure 2 S _207, described in detail in conjunction with sound-muffling process S _208. FIG. 3 is a detailed operation flowchart of the reproduction processing of _S206 in FIG. 2, and FIG.
It is a detailed operation | movement flowchart of the sound production | generation / silence processing of _S207 . In the following explanation, the automatic performance
In the case of a normal tempo performed at a tempo that is pronounced 60 times,
The following description will be made on two examples of a fast tempo performed at a tempo in which a quarter note is generated 255 times per minute.

First, the player performs a tempo-up key so that the automatic performance is performed at a tempo in which quarter notes are pronounced 60 times per minute.
An example in which the user operates the 111 or the tempo down key 112 (FIG. 1) will be described. In this case, as shown in FIGS. 8A and 8C, the notes F ₄ , A ₄ , G _4, etc. of the sixteenth note are (60 seconds)
/ 60 times) x (4/16) = 250 msec interval. At this time, the microcomputer 101, in S ₂₂₁ or S ₂₂₃ in the main routine in FIG. 2, sets a period corresponding to the performer of the above tempo set for timer 103. This allows the timer
The timer interrupt by 103 occurs once every 250 msec as shown in FIG. 8 (d). On the contrary,
The reproduction process ( _{S207 in} FIG. 2) in the main routine is executed at the timing indicated by “↑” in FIG. 8 (f). Then, a musical tone corresponding to the event data is generated at the timing of “↑” to which symbols such as F ₄ , A ₄ and G _{4 in} FIG. 8 (f) are added. In this case, since the reproduction process is executed at intervals of up to about 70 msec as described above, the actual tone generation timing is also up to 70 msec from the target timing.
Although there may be some delay, automatic performance is realized at substantially the target timing within the range of the shift.

The above-described operation example of FIG. 8 will be specifically described with reference to the operation flowchart of the reproduction process of FIG. 3 and the sound generation / muting process of FIG. 4 and the operation explanatory diagram of FIG.

First, various control variables for reproduction processing are stored in the RAM 105 of FIG. On pointer onpt indicates the number obtained by adding 1 to the number of the musical sound to be sounded in one sound-muffling process (FIG. 2 S _208). On-buffer onbuf
This is array data for storing key codes to be sounded in the mute processing. At the end of the reproduction processing ( _{S207 in} FIG. 2),
Valid key codes are stored in the on-buffers onbuf (1) to onbuf (onpt1-1). Off pointer offpt
Indicates the number obtained by adding 1 to the number of musical tones to be muted in one sounding / muting process. The off-buffer offbuf is array data for storing key codes to be silenced in the sound generation / mute processing.
(1) -offbuf (offpt-1) stores a valid key code. The timer value timplay is timer data that is counted up at regular time intervals.
Each time a timer interrupt occurs from 103, the microcomputer 1
01 executes the interrupt program shown in the operation flowchart of FIG. 3 (b), and sets the contents of the timer value timplay to +
It is incremented by one ( _S327 ). Time data segw
In ait, time data sequentially read from the RAM 105 is stored. The variable n will be described later.

When the player operates the start / stop key 110, the reproduction start process ( _S217 ) of the main routine shown in FIG. 2 is executed. Thereby, the timer 103 (first
Is reset, and the timer value timplay and the time data segwait are cleared to zero.

Then, at time t ₀ of Figure 8 (C), FIG. 3 (a) reproducing process of the main routine shown in (FIG. 2 S ₂₀₆₎ is executed.

First, as in FIG. 6, Offpt and onpt is cleared (S _301, S _302).

Next, the contents of timplay and segwait are compared ( _S303 ). If both contents are determined to be equal, timpla
segwait is the content of the subtracted new timplay from y (S _304).

Next, as shown in Figure 6, the first event data "132" indicating the ON key code F ₄ from RAM105 of contents shown in FIG. 5 (b) is read out, onbuf (onpt) = o
The key code value indicating the key code F ₄ to nbuf (0) "132-8
8 "=" 44 "is stored ( _S305 → _S306 → _S307 → _S308 ).
Here, “88” is subtracted from “8” in the event data indicating “ON” in order to distinguish it from the event data indicating “OFF”.
Since an 8 "offset is added, the offset is subtracted to return to the true key code.
When a value is stored in f, o is used to advance the contents of the array.
the contents of the npt is +1 (S _309).

Subsequently, F ₄ from RAM105 of contents shown in FIG. 5 (b)
The time data “1” next to the event data indicating ON is read and set to segwait (S ₃₁₀ → S ₃₁₁ ).

Thereafter, the process returns to _S303 via _S312 and _S313 (described later), and the determination is NO, ending the current reproduction process.

Above reproduction processing result onbuf (0) valid key in only the code F ₄ is set.

After the above-described regeneration process, sound-muffling the main routine (FIG. 2 S ₂₀₈₎ is executed. The details of this process
Shown in the figure.

First, the value of offpt is 0 and no valid data is stored in offbuf, so the determination in _S403 is NO. Next, since the variable n = 0 (S ₄₀₅ ) is smaller than onpt = 1, in S ₄₀₇ , the tone of the note No. of onbuf (0), that is, the tone of the key code F ₄ , is sounded by the sound source 106 (FIG. 1). ).

Thereafter, the determination of S ₄₀₆ is NO, ends the sound-muffling process.

By the above operation, as shown in FIG. 8 (c) and (f), the tone of the key code F ₄ at timing t ₀ is pronounced.

Thereafter, at each of the timings t ₁ , t ₂ , t ₃ , and t ₄ in FIG. 8C, the reproduction process of the main routine (S _{207 in} FIG. 2)
Is repeatedly executed, during which no timer interrupt from timing 103 (FIG. 1) occurs and timplay
Is not more than the content of segwait, the determination in _S303 is NO, and the reproduction process is not executed as shown in FIG.

At timing t ₅ of FIG. 8 (c), the timer interrupt from the timer 103 occurs, the interrupt program shown in the operation flowchart in FIG. 3 (b) is executed, the contents of timplay is +1 (S ₃₂₇ ).

Immediately after that, when the reproduction process is executed, _{S303 in} FIG.
Is YES, and as shown in FIG. 6, timplay
After segwait is the contents of the subtracted new timplay (S _304), the event data "44" indicating the off-key code F ₄ from RAM105 of contents shown in FIG. 5 (b) is read from, the data is set to offbuf (offpt) = offbuf (0 ) (S 305 → S 306 → S 316 → S 317 → S 318). offb
When the value is stored in ut, to advance the contents of the array by one
The contents of offpt is +1 (S _319).

Subsequently, F ⁴ from RAM105 of contents shown in FIG. 5 (b)
The time data “0” next to the event data indicating OFF is read out and set to segwait (S ₃₁₀ → S ₁₁₁ ).

Thereafter, the process returns to _S303 via _S312 and _S313 (described later). In this case, although timplay is returned to "0" at S _304, since even Segwait "0" is set in the S _311, S
_As soon as the determination at ₃₀₃ becomes YES again, the next event data is reproduced. Thus, as shown in Figure 6, the event data "136" indicating the ON key code A ₄ from RAM105 of contents shown in FIG. 5 (b) is read out, the key code A ₄ "136 -88 "=" 48 "is onbuf (onpt)
= Onbuf (0) is set (S ₃₀₄ → S ₃₀₅ → S ₃₀₆ → S
₃₀₇ → _S308 ). The contents of onbuf is +1 (S _309).

Subsequently, A _{4 is obtained} from the contents of the RAM 105 shown in FIG.
The time data “1” next to the event data indicating ON is read out and set to segwait (S ₃₁₀ → S ₃₁₁ ),
Returning to S ₃₀₃ through S _312, S _313, the judgment is to end the current reproduction process becomes NO.

As a result of the above reproduction process, offbut (0) to a valid key code F ₄ is valid key code A ₄ is set to onbuf (0).

After the above-described regeneration process, the fourth sound-silencing process of the main routine shown in FIG. (FIG. 2 S ₂₀₈₎ is executed. Here, the processing of S ₄₀₃ is executed once offbuf (0) of the note No. That mute instruction tone key code F ₄ sound source 106
Then, the process of _S407 is executed once and
onbuf (0) of the note No. sounding instruction of musical sounds i.e. key code A ₄ of is made to the sound source 106.

By the above operation, as shown in FIG. 8 (c) and (f), the tone of the key code A ₄ at the timing immediately after the t ₅ is pronounced.

Thereafter, the same processing is repeated, so that the timer 10
At a timing almost following the timer interrupt generated once every 3 to 250 msec, a sounding operation based on the event data is executed, and the automatic performance is performed at a tempo in which quarter notes are sounded 60 times per minute. .

The sound source 106 can simultaneously generate or mute 16 musical tones. Accordingly, in step _S312 of the reproduction process shown in FIG.
And in S _313, the contents of offpt and onpt reaches 16, to cross the number of sound or mute 16 voice polyphony, more regeneration process is not performed unconditionally.

Next, the player performs a tempo-up key so that the automatic performance is performed at a tempo in which quarter notes are pronounced 255 times per minute.
An example in which the user operates the 111 or the tempo down key 112 (FIG. 1) will be described. In this case, as shown in FIGS. 9A and 9C, the notes F ₄ , A _4, etc. of the sixteenth note are (60 seconds / 2
(55 times) x (4/16) = 250 msec interval. At this time, the timer interrupt by the timer 103 occurs once every about 59 msec as shown in FIG. 9 (d). On the other hand, the reproduction processing in the main routine (FIG. 2)
_S207 ) is executed at the timing indicated by "で" in FIG. 9 (f). Then, at the timing of “↑” to which symbols such as F ₄ and A _{4 in} FIG. 9 (f) are given, a musical tone corresponding to the event data is generated. In this case, since the playback process is executed at intervals of about 70 msec at the maximum as described above,
The actual tone generation timing is also delayed by up to 70 msec from the target timing. Therefore, during the delay, the next timer interrupt may occur. For example, after a timer interrupt at a timing t ₄ as FIG. 9 (d) has occurred, before the reproducing process at the timing t ₆ is executed, when again the timer interrupt timing t ₅ is as occurs is there. In such a case, the key code G ₄ to be pronounced at the target timing t ₄
Is canceled, and the next target timing t ₅
In pronunciation of a tone of a key code F ₄ to be sound is performed. As a result, even if the tempo is raised again, automatic performance with less listening difficulty is realized.

The above-described operation example of FIG. 9 will be specifically described with reference to the operation flowchart of the reproduction processing and the sound generation / mute processing of FIG. 4 and the operation explanatory diagrams of FIGS. 7 (a) and 7 (b). I do.

First, the reproduction process at the timing t ₀ is similar to that of the normal tempo of the foregoing, as FIG. 7 (a) onbu
The key code of F ₄ is set in f (0), and based on the key code, in the sound generation / silence processing (S _{208 in} FIG. 2), the tone generation instruction of the tone of the key code F _{4 to} the sound source 106 (FIG. 1) is issued. Done.

Thereafter, at the timing t ₁ of FIG. 9 (c), although the reproduction process of the main routine (FIG. 2 S ₂₀₇₎ is executed,
Timer interrupts between t ₀ and t ₁ from the timer 103 (FIG. 1) is not generated, because the contents of timplay is not a more content of Segwait, determination of S ₃₀₃ of FIG. 3 (a) is NO The reproduction process is not executed as shown in FIG.

At timing t ₂ of FIG. 9 (c), the timer interrupt from the timer 103 occurs, the interrupt program shown in the operation flowchart in FIG. 3 (b) is executed, the contents of timplay is +1 (S ₃₂₇ ).

The reproduction process at the timing t ₃ immediately after is the same as the playback processing just after t ₅ when the normal tempo of the foregoing, the result of the processing, valid key code offbuf (0)
F ₄ is, onbuf (0) valid key code A ₄ is set to, based on their, sound-silencing process (FIG. 2 S ₂₀₈₎
In, mute instruction tone key code F _4, the key code
Sounding instruction tone of A ₄ is performed for the sound source 106.

Subsequently, the timer at the timing t ₄ of FIG. 9 (c)
When the timer interrupt from 103 occurs, the contents of timplay is being +1 "1" (FIG. 3 (b) S _327). Here, it takes time to process, such as function keys scan in the main routine of FIG. 2, before reproduction process corresponding to the timing t ₄ described above is executed at the timing t ₅ of FIG. 9 (c) When the timer 103 generates a timer interrupt, the content of timplay is incremented by one again to "2" as shown in FIG. 7 (a).

When the reproduction process at the timing t ₆ immediately following is executed, so the determination of S ₃₀₃ of FIG. 3 YES, and as shown in FIG. 7 (b), segwa from the contents of timplay "2"
is subtracted the content "1" of it, after the contents "1" of the new timplay (S _304), the event data indicating an off-key code A ₄ from RAM105 of contents shown in FIG. 5 (b) “48”
Is read, and the data is stored as offbuf (offpt) = offbuf
Set to (0) (S ₃₀₅ → S ₃₀₆ → S ₃₁₆ → S ₃₁₇ →
S ₃₁₈ ). The contents of offpt is +1 (S _319).

Subsequently, A _{4 is obtained} from the contents of the RAM 105 shown in FIG.
The time data “0” next to the event data indicating OFF is read out and set to segwait (S ₃₁₀ → S ₃₁₁ ).

Thereafter, the flow returns to S ₃₀₃ through S _312, S _313. In this case, Timplay is a 1 in S _304, since the zero segwait is set in S _311, reproduction of the next event data determination as soon becomes YES again S ₃₀₃ is performed. Thus, as FIG. 7 (b) is shown, the event data "134" is read out indicating the ON key code G ₄ from RAM105 of contents shown in FIG. 5 (b), the A ₄ Key code "134-88" = "46" is onbuf (oupt) = onbut
Set to (0) (S ₃₀₄ → S ₃₀₅ → S ₃₀₆ → S ₃₀₇ →
_S308 ). The contents of onbuf is +1 (S _309).

Subsequently, G _{4 is obtained} from the contents of the RAM 105 shown in FIG.
The time data “1” next to the event data indicating ON is read out and set to segwait (S ₃₁₀ → S ₃₁₁ ),
Back to S ₃₀₃ through S _312, S _313. 1 for segwait like this
“1” corresponding to the timing of a sixth note is set, but the content of timplay is “1”, so the next event data is reproduced immediately after the determination in S ₃₀₃ becomes YES again. Will be This is before the reproduction process corresponding to the timing t ₄ of FIG. 9 (c) is performed at the timing t ₅ of FIG. 9 (c), the timer interrupt from the timer 103 occurs, the timplay This is because the content is incremented by 1 to become "2". The rest, the processing for canceling the onbuf the set value of the key code G ₄ in (0) "46".

That is, the determination in _S303 in FIG. 3 is YES, and segwait is subtracted from timplay to obtain a new ti as shown in FIG. 7 (b).
After being the contents of mplay (S _304), FIG. 5 (b) event data "46" indicating the off-key code G ₄ from RAM105 contents represented by is read.

The determination of S ₃₀₆ is after reaching YES, and is set to "0" to the variable n in S _316, and the contents of the variable n and onpt in S ₃₁₇ are compared. Now, since the onbuf (0) to a valid key code G ₄ the contents of the stored onpt has a "1", S
The determination at ₃₁₇ is YES.

Furthermore, the contents of the event data and onbuf read by the processing of S ₃₀₅ (n) in S ₃₂₀ are compared.
Now, the event data is the event data "46" indicating the off-key code G _4, the contents of both match. Thus, the determination of S ₃₂₀ is YES.

After this operation, after the contents of the variable n becomes the +1 "1", the contents of content and oupt of this variable n is compared in S _322. In this case, since the contents of both match,
The determination at ₃₂₂ is YES.

As a result of the above determination, the value “0” is stored in onbuf (n−1) = onbuf (0) (S ₃₁₄ ), and the content of onpt is decremented by 1 (S ₃₁₅ ). In this way, the key code G ₄ stored once oubuf (0) is canceled.

Subsequently, G _{4 is obtained} from the contents of the RAM 105 shown in FIG.
The time data “0” next to the event data indicating OFF is read out and set to segwait (S ₃₁₀ → S ₃₁₁ ).

Thereafter, the flow returns to S ₃₀₃ through S _312, S _313. In this case, although timplay is returned to “0” in S ₃₀₄ ,
_Since “0” is also set to segwait in ₃₁₁ , the determination in S ₃₀₃ becomes YES again, and the next event data is reproduced. Thus, as shown in FIG. 7 (b), the event data "132" indicating the ON key code F ₄ from RAM105 of contents shown in FIG. 5 (b) is read out,
Key code "132-88" of F ₄ = "44" is oubuf (onpt) = on
Set to buf (0) (S ₃₀₄ → S ₃₀₅ → S ₃₀₆ → S ₃₀₇ → S
₃₀₈ ). The contents of onbuf is +1 (S _309).

Subsequently, G _{4 is obtained} from the contents of the RAM 105 shown in FIG.
The time data “1” next to the event data indicating ON is read out and set to segwait (S ₃₁₀ → S ₃₁₁ ),
Returning to S ₃₀₃ through S _312, S _313, the judgment is to end the current reproduction process becomes NO.

As a result of the above reproduction process, valid key code A ₄ is set to offbuf (0). In the onbuf (0), the key code G ₄ that are set once is canceled,
Key code F ₄ that follows is set.

After the above-described regeneration process, sound-muffling the main routine (FIG. 2 S ₂₀₈₎ is executed, mute instruction and sounding instruction key code F ₄ key code A ₄ is shown in Figure 4, the sound source 10
6 can be made.

By the above operation, as shown in FIG. 9 (c) and (f), at the timing t ₆ just after t _5, the key code G ₄ corresponding to the timing of t ₄ are not sound is canceled, tone key code F ₄ corresponding to the timing of the subsequent t ₅ is pronounced.

Thereafter, the same processing is repeated, so that the timer 103 outputs 59 msec while the main routine of FIG.
If two or more timer interrupts occur once every two times, the difference between timplay and segwait widens, so that the on / off of the widened event data is canceled in the playback processing. As a result, automatic performance with less listening difficulty becomes possible with quarter notes 255 times a minute.
Performed at a tempo that is pronounced twice.

Here, for example, as the event data corresponding to the timing t ₄ of Figure 9, the key code on the other key code in addition to the ON G ₄ simultaneously (at intervals of time data "0")
If the content of onbuf (0) is canceled in the case where it is designated, S ₃₂₁ → S ₃₂₂ → S ₃₂₃ → S ₃₂₁
In iteration of the loop, Onbuf by treatment S ₃₂₃
(1) The contents of the subsequent array are sequentially shifted forward by the primary array address, and the contents of the last array address onbuf (n-
1) is set to "0".

For other key codes, the time data "0"
If the key code-off event data is specified as the automatic performance data with the intervening, the reproduction process is repeated to be sequentially canceled.

As a result of the contents of onpt is -1 at S _325, S
Because the contents of onbuf (n-1) in ₃₂₄ is not to be subsequently accessed, its contents are always "0" need not be cleared, the processing of S ₃₂₄ is not necessarily performed.

Furthermore, without the contents to be canceled onbuf (0), when such are present in onbuf (1) and later, S ₃₂₀
In the loop of → S ₃₂₆ → S ₃₁₇ → S ₃₂₀ , the contents of the array are searched until the contents to be canceled are found.

On the other hand, for example, when the key code G ₄ is event data corresponding to the timing t ₄ of FIG. 9 was more note length rather than 16-note is read out keycode G ₄ on the event data is the time interval until the event data off of a key code G ₄ is read from the, Timplay and segwait when first entered the reproduction processing timing t ₆
Larger than the time interval "1" corresponding to a difference between, at time t ₆ keycode G ₄ are performed the sounding instruction without being canceled, tone corresponding to it, the time interval between off from the on The sound is produced for the time obtained by subtracting the time interval corresponding to the difference between timplay and segwait.

In the above-described reproduction processing, as shown in FIG. 9, a series of sixteenth notes has been specified as the automatic performance data, and the event data has been specified at intervals of time data "1". Even when the note length is longer than a note, time data corresponding to the note length is sequentially stored in the segwait and compared with the contents of timplay, so that a similar operation is realized at a timing corresponding to each note length.

〔The invention's effect〕

According to the present invention, even when the repetition of the automatic performance control processing cannot keep up with the tempo, the timing of the automatic performance is advanced by an amount that is delayed from the tempo. When the delay becomes large, the process of reading out the event data from the automatic performance storage means at one time is repeated a plurality of times, so that the event data delayed from the tempo is appropriately canceled, and the delayed sound is not generated. Is controlled.

With the above control processing, it is possible to realize an automatic performance with less listening difficulty even if the reproduction tempo is increased.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is an operation flowchart of a main routine in the embodiment, FIG. 3 is an operation flowchart of reproduction, and FIG. Flowchart, FIG. 5 is an explanatory diagram of the automatic performance data, FIG. 6 is an explanatory diagram of the operation when the tempo of the quarter note is 60, and FIG. 7 is an operational diagram when the tempo of the quarter note is 255. FIG. 8 is an operation timing chart when the tempo of the quarter note is 60, and FIG. 9 is an operation timing chart when the tempo of the quarter note is 255. 101: microcomputer, 102: ROM, 103: timer, 105: RAM, 106: sound source, 111: tempo up key, 112: tempo down key.

Claims (2)

(57) [Claims] An event data for controlling a tone generation operation of a tone in a sound source is sequentially read from an automatic performance storage means at a timing corresponding to the time data while reading time data stored in a pair with the event data. An automatic performance device that performs an automatic performance of a sound source by repeatedly executing an automatic performance control process including a process of instructing a sound source based on event data as a part of a main routine consisting of a series of processes. And a timer value storage means for storing a timer value that counts up at a cycle based on an arbitrarily set tempo from the start of the automatic performance, and the automatic performance control process further comprises: It is determined whether or not the time data is equal to or longer than the time data read from the automatic performance storage means. A second process to execute a second process to a sixth process, and when the determination is false, a first process to end the process of the current timing, and a value obtained by subtracting the time data from the timer value as a new timer value A second process for reading out new event data from the automatic performance storage means; and a fourth process for instructing a sound source to generate sound based on the event data read out in the third process. And after the end of the fourth process, the first process is repeated for a new timer value based on the second process and time data corresponding to new event data based on the third process. If the event data for instructing the sounding of an arbitrary musical tone and the event data for instructing the silencing of the musical tone are successively read from the automatic performance storage means at the current timing, The sixth control is performed so that the tone generation control instruction is not issued to the sound source without executing the fourth processing.
And an automatic performance device comprising: 2. Event data for controlling a tone generation operation of a tone in a sound source is sequentially read from an automatic performance storage means at a timing according to the time data while reading time data stored in a pair with the event data. An automatic performance device that performs an automatic performance of a sound source by repeatedly executing an automatic performance control process including a process of instructing a sound source based on event data as a part of a main routine consisting of a series of processes. Each time the automatic performance control process is repeated, a delay detecting means (S304) for detecting how much the repetitive operation is delayed compared to an arbitrarily set tempo; The automatic performance control process is executed while the readout timing of the next event data is advanced according to the delay amount As a result, if the event data instructing the sounding of an arbitrary musical tone and the event data instructing the silencing of the musical tone are successively read out from the automatic musical performance storage means in the automatic performance control processing of this time, And a timing control means for performing control so as not to issue a tone generation control instruction of the musical tone.